Help/Advice with my Line Statistics please in Canadian Broadband

Re: Help/Advice with my Line Statistics please

Sun, 06 Dec 2009 22:51:34 EDT

Bicephale : Yes, there's a humourous note in my answer indeed. On the other
hand, you're well aware that i stopped trusting instant snapshots
a while ago... My reasons should seem obvious when one looks at
some of the "marginal events" i captured during the last months.

For example, one may be witnessing a popcorn-like progression in
terms of SNR Margin vs Error Rate curve and then the MoDem will
reset and the situation changes radically... Depending on when it
was captured, i consider that i just can't be absolutely sure what
happened exactly unless a unit was actually monitored externally.

I see a difference between having a hint and collecting evidence.

:)

Re: Help/Advice with my Line Statistics please

Sun, 06 Dec 2009 19:22:02 EDT

dslrocker3 :

said by Bicephale :

I'll be pleased to take a look at your long-term log, if any, but it
seems i must remind you that 'RouterStats' already plots curves!

Bicephale, my last post was meant as a joke. Take a look at the error totals and the dsl uptime.... also, ask yourself this: what is the difference between this recent error total and the other error totals from previous days and what's the only thing that appears to have caused the difference? (this is a riddle for you)

Based on your laughing icon, I think you already knew that I was joking and I think that you already knew that I remebered that Routerstats will plot the graph for me. The joke was that there's really nothing to graph. The error rate is so rediculously low now that I've switched modems that the graph would be meaningless.

Assuming the modem's counters are accurate, why is it that my TD-8840 v1.3 with a firmware hacked to one that is meant for a different TP-Link modem is so much superior at handling line noise when compared to the Speedtouch 516 v6 (even when using firmware 7.4.3.2)?

Re: Help/Advice with my Line Statistics please

Sun, 06 Dec 2009 19:16:37 EDT

Bicephale : I'll be pleased to take a look at your long-term log, if any, but it
seems i must remind you that 'RouterStats' already plots curves!

Re: Help/Advice with my Line Statistics please

Sun, 06 Dec 2009 17:24:01 EDT

dslrocker3 : Bicephale, as much as a I agree with you that DMT screenshots only show a certain about of information, here's a DMT screenshot for you.

If I had something like Routerstats running and used it to record the frequency and timing of errors, do you think you would have been abel to help me graph the errors that are summarized in this screenshot?

Re: Help/Advice with my Line Statistics please

Sat, 05 Dec 2009 09:07:19 EDT

Bicephale : Euh...

Allow me to remind you that TerTech created 'DSL_StatScope'
specifically because of Thomson's 'DMT'-hostile FirmWare... He
once wrote essentially what you just expressed here, as i recall.

Re: Help/Advice with my Line Statistics please

Sat, 05 Dec 2009 03:48:26 EDT

dslrocker3 : Hahaha! More tones should mean a higher attaianble rate but I'll never know thanks to our friends at Thomson!

!!! No more Diagnosis mode available in DMT 7.35 if using firmware 7.4.3.2.0. :(

Also, I can't get Routerstats to even update read the error coutners properly from the modem page any more (it can read get an initial reading of the error counter but will never update the total for me meaning that the graph just shows a flat line at 0 errors all the time). I sense that this vesion 7 frimware won't be lasting long on my Speedtouch 516.

Re: Help/Advice with my Line Statistics please

Thu, 26 Nov 2009 16:20:30 EDT

Bicephale : Depends on the predictability of the line, i like that.

:)

Re: Help/Advice with my Line Statistics please

Thu, 26 Nov 2009 15:40:20 EDT

dslrocker3 : Perhaps we can say that the usefullness of DMT screenshots can depend on the predictability of the line. A line with constantly changing conditions could render DMT reported error rates to be nothing more than an average and would only show those averages or line conditions at that exact point in time. If my line once again became problematic, I would be more likely to peform more detailed diagnostics. At the moment, the urgency isn't really there. Have I become complacent?

Re: Help/Advice with my Line Statistics please

Thu, 26 Nov 2009 06:15:52 EDT

Bicephale : Hi DSLRocker,

I will not dispute your observations, my
intent is to simply react to a statement:

DR> ...DMT screenshots only give a DR> general idea of what's going on...

This needs to be rephrased in my opinion!

My choice of words would differ. Imagine
a collection of successive 'DMT' captures
taken at a quick pace as when filming, it
is the movement which emerges from them
that gives a general idea of what's going
on... Instant 'DMT' snapshots are similar
to one single frame, it does fit within the
film but it reveals little about the overall
story unless it's a film with no action...

Lets remind the readers that all films are
not as dull as yours, that all phone-lines
are not providing such a stable DSL signal!

In fact, even yours may suffer from a few
hickups which will slip through if you don't
monitor it using more potent resources...

The 'DMT' instant snapshots are handy if
one trusts them to be representative of a
situation and he needs to communicate with
others in a swift/convenient way but they
fail to provide a "general" perspective. My
impression is that you're aware of this, of
course, but the words left the door opened
to some misunderstanding: i find that the
'DMT' snapshots only give a restricted idea
of what's going on during a specific time,
as far as i'm concerned! Don't you?

Look at a recent graph i've published lately:

From the ground up!, Bicephale, 2009-Nov-23

Even a dozen 'DMT' instant snapshots taken
at random intervals would fail to inform an
observer about what was really going on...

All he would see is a variation of the SNR
Margin taking place on some occasions, it's
nothing significant considering that these
could result from marginal events/readings.

Well, i might be tempted to ignore ordinary
3+ dB spikes but certainly not a couple of
those solid boost events when they have a
level of such magnitude! In my case those
anomalies turned out to be the mainstream
story and it's quite a puzzling one which no
single 'DMT' capture can document.

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 20:06:54 EDT

dslrocker3 : I know that these DMT screenshots only give a general idea of what's going on with the line but as I suspected before, the gap between convential wiring and unconventional wiring (better results) has been narrowed.

It used to always be a 3% difference between the two wiring methods but now it only appears to be 2%. The unconventional wiring tests still show an RCO that varries from 55 to 56% just as before. But the conventional wiring which always used to vary between 58 and 59% now varries between 57 and 58%. I know that sounds like a very, very small difference which it is but the fact tht I've been able to repeat it a few times can only lead me to believe that simply restripping one little wire closed the gap a little. This small change in RCO happened as soon as a finished restripping that wire and turned the modem back on.

conventional wiring Nov 24/09
conventional wiring Nov 25/09
unconventional wiring Nov 25/09

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 17:55:32 EDT

Bicephale : Indeed, i see you point! Well, to the very least
you've proven the US 6 dB IKNS bug, i'd say!...

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 16:20:30 EDT

dslrocker3 : Disabling upstream tones didn't seem to open up any additional downstream bandwidth but then again, the IKNS remotes leave such a large gap between the upstream and downstream tones that are used.

I wonder how long I could leave my connection like this before my profile got "improved" for me? hahaha! I think that I better stop making my modem sync at crazy rates like that, at least for a few days.

default settings for reference
upstream ton···ment.txt 10,152 bytes

upstream tones 0 to 15 and 16 to 31 turned off

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 16:14:00 EDT

dslrocker3 : The overall SNRM remains so close in both screenshots that there really was no overall difference (I could have clicked on refresh and got a higher number for the first screenshot if I really wanted to).

For my line, this doesn't accomplish anything but I do see that my "least good" tones were dropped if you look at the SNR per tone graph (the lowest tone SNR is higher than before). I could see how this could possibly allow a user to achieve stability on a higher than "proper" profile, even if the user doesn't connect at the full rate. In this experiment, some of my attainable downstream bandwidth was sacrificed.

defaults for reference purposes
downstream t···ment.txt 15,260 bytes
tones that had less than 5 bits disabled (dmt20091125_1529 tones with below 5 bits from reference sample disabled.png)

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 16:07:29 EDT

Bicephale : If that change of Attenuation level were real i'd
expect this condition to reflect into the graphs.

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 16:01:00 EDT

dslrocker3 : A question was presented earlier about how turning off things such as Annex M and ADSL2 modes helped when the user wasw connecting only in G.dmt mode. My screenshots tend to say that it does not help. The comparisons are almost indentical execpt for one very strange behavior with the attenuation. I can only assume that this is either a firmware bug or a non-repeatable conincidence because it just doesn't make any sense for the attenuation to change like that when the differences in settings should have had zero effect.

TD-8840, All settings at default
All DSL modes enabled

extra dsl modes off, only g.dmt is enabled

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 09:45:58 EDT

Bicephale : Hi AnonDSLGuy,

Let me guess, your 'DMT' capture would be like this:

You know, i think it's great when one can obtain some
expert advice around here, even if the source happens
to be anonymous, but i searched a few threads about
Impulse Noise Protection (INP) and there's no "Before"/
"After" comparison whatsoever:

»Re: DSL overhead/sync
»Re: Obscenely High Ecc (sec)? Low US SNR?
»Re: Speedtouch 516 dying. Replacement modem?
»Re: high profile+modem sync lower OR low profile+modem sync max?

It appears you're using a 2Wire and you don't have to
deal with Bell - lucky you!... I may be wrong but i've
found the "INP" string over there:

»I cant understand my line stats for bell 2wire 2701HG-G

Is that where you became aware of the INP feature
for the 1st time?...

Appart from your GNet basher theories which once
again come with nothing but your anonymous word,
i'd like to read further about "Interleaver Depth"...

Euh... But there again, something tells me i simply
can't hope for "Before"/"After" comparisons figures
anyway. So, what can you really offer other than
remedies which are probably off-topic here and will
be best suited in some AT&T american forum???

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 06:41:31 EDT

AnonDSLGuy :

said by Bicephale :

You seem to miss the action but now that you own a
TP-Link TD884x with bin-per-bin spectral shaping lets
have some fun by simply experimenting with those:

64, 74, 84, 129, 133, 134, 137, 140, 148, 150, 158,
182, 191, 192, 199, 226, 234, 235, 236, 237, 243.

Let DSL do its thing, and don't notch tones! If the bins are noisy less bits (possibly none) will be assigned to them. Noise is transient, and if you notch those tones now, they may be less noisy next time you retrain, so you're pissing away bandwidth for disabling tones you could be using.

Instead complain to Bell as to why they relegate wholesale customers to ADSL, when running everyone on ADSL2+ would save them money from much added stability (ie: SRA, INP). You poor easterner's and your utterly ignorant ILEC =(.

Re: Help/Advice with my Line Statistics please

Wed, 25 Nov 2009 04:34:18 EDT

Bicephale : You seem to miss the action but now that you own a
TP-Link TD884x with bin-per-bin spectral shaping lets
have some fun by simply experimenting with those:

64, 74, 84, 129, 133, 134, 137, 140, 148, 150, 158,
182, 191, 192, 199, 226, 234, 235, 236, 237, 243.

Re: Help/Advice with my Line Statistics please

Tue, 24 Nov 2009 18:46:36 EDT

Mauricio : Temperature variations, amongst others, could be the cause for most of the fluctuations. My connection has never been stable in the Spring/Fall. I would assume, even if you're line is far from being marginal, that it could also be affected by such variables.

Colder weather usually translates into better line conditions. My line attenuation has seen a slight decrease over the last few weeks (67.7 to 67.0).

Can't wait for expertech to finish splicing/activating the RDSLAM that will feed my line.

Re: Help/Advice with my Line Statistics please

Tue, 24 Nov 2009 18:32:19 EDT

dslrocker3 : Yesterday, I restripped one of my wires that goes into the jack that feeds my modem. It might just be a conincidence but it would appear that the performance difference gap between normal wiring and my uncoventional wiring trick appears to be getting smaller. If it's a just a coincidence, that would just mean that my line has got slightly better in the last 24 hour all round and it has nothing to do with me redoing that wire.

Re: Help/Advice with my Line Statistics please

Mon, 23 Nov 2009 15:51:34 EDT

dslrocker3 : I guess I'm trying to say not to believe everything that these manufacturers tell you. If I did that, I would till be stuck with a TP-Link firmware that did not have a PPPoE disconnect/reconnect button.

Re: Help/Advice with my Line Statistics please

Mon, 23 Nov 2009 15:50:18 EDT

Bicephale : I'm listening but i have no clue as to why FirmWares
happen to be cross-compatible or not. At one point
i was under the impression you were about to try ST
FirmWare in order to turn a TP-Link into a Thomson,
actually!... Promise me you won't do that, not even
think of it!!!

Re: Help/Advice with my Line Statistics please

Mon, 23 Nov 2009 13:41:16 EDT

dslrocker3 : Hang on a second!!!! this so-called "23K" firmware for the TD-8840 v4 is the same firmware as the one for the TD-8841 v1 firmware! What types of "games" is TP-Link trying to pull here?

Interestingly, the TD-8840 v4 firmware whose file is starts with the name "TD8811A.....up" works but the TD-8840 v2 firmware that has a file name that starts with "TD8840A...up" won't load onto my TD-8840 v1.3 unit.

Bicephale, do I have your attention yet when it comes to firmware updates if TP-Link decides to make your version of the TD-8841 dissapear from the technical support page?

Re: Help/Advice with my Line Statistics please

Mon, 23 Nov 2009 13:26:38 EDT

dslrocker3 : Hahahahahaha!!!

WARNING, do not try this stuff unless you fully understand that you could destroy your modem by doing what I did!!!!

Now, downloading the firmware for TD-8840 v4 turned my TD-8840 v1.3 into a TD-8810 v4! LOL!

Everything works fine. Since TP-Link's support website denies that TD-8840 v1.3 ever existed, I was forced to take into my own hands to find the most update-to-date firmwares that are unofficially available. So far, I've sucessfully installed the following and everything works fine to my knowledge: TD-8841 v1 frimware dated November 28, 2008 and TD-8840 v4 fimware dated July 12, 2009.

I am becoming increasing suspicious about TP-Link's claims that you cannot use firmwares from different hardware versions. I am very much suspecting that the differences are purely cosmetic and that they are simply saying not to do it to discourage the sale of used units (ie. they will purposely discontinue certain versions and no longer provide more firmware updates)

I wonder if there will be any performance differences between all of these different firmware versions.

Re: Help/Advice with my Line Statistics please

Fri, 20 Nov 2009 16:42:40 EDT

Bicephale : I'm not going to switch MoDems before the week-end
is over but i'll sure keep your stimulating idea in mind.

:D

Re: Help/Advice with my Line Statistics please

Fri, 20 Nov 2009 16:26:00 EDT

Vomio : This has probably been covered before, but with my TD-8841 hooked up to an Alcatel DSLAM, I found an apparent improvement in the reported signal to noise margin and a reduction in errors by turning off modulation protocols that were not in use. See Advanced Setup, DSL Settings.

Of course this may merely be an observed anomaly from a short snapshot of time and possibly pure coincidence.

I'd be interested if anybody else observes this or not.

Marcus

Re: Help/Advice with my Line Statistics please

Fri, 20 Nov 2009 12:53:15 EDT

dslrocker3 : Oh boy, if you thought my error rates were uninteresting with my Speedtouch 516, life is about to get even more uniteresting with my TD-8840. I left the connection up for almost a day and I'm averaging 4 correctable errors per hour. Even if I ran some advanced diagnostic program to monitor line noise and errors, the graphs aren't going to show a heck of lot with that low of error numbers.

I can't speak as to the accuracy of the modem's error counter but I based on these numbers, it is a rather apparent that the TD-8840 (with TD-8841 firmware installed) is much better at handling noise to prevent line errrors than my Speedtouch 516 with firmware 6.1.0.5. To be fair, the Speedtouch modem is running an older firmware but why should I have to lose features to upgrade it?

unconventional wiring, TD-8840
unconventional wiring, speedtouch 516 (from a few days ago)

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 22:01:56 EDT

Mike2009 : I'm on an Alcatel DSLAM.

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 21:44:49 EDT

dslrocker3 :

said by Mike2009 :

I'm curious about this interleaver depth as well. Yours is 4 on the downstream and 1 upstream. Mine is 1 downstream as well as upstream. Anyone understand this?

Based on your 9.0dB upstream signal to noise ratio margin, you are not on an IKNS rdslam. The 1 interleaver depth just means that you are on fast path. This is also confirmed by your modem's adsl statistics page near the top. You are connected to a different type of DSLAM OR RDSLAM than me.

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 21:32:09 EDT

Mike2009 : I'm curious about this interleaver depth as well. Yours is 4 on the downstream and 1 upstream. Mine is 1 downstream as well as upstream. Anyone understand this?

Stats.doc.zip 114,440 bytes

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 21:11:31 EDT

Bicephale : Of course i knew you were serious! So, when do we
get good close-ups taken from the top of that pole?

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 18:48:31 EDT

dslrocker3 :

said by Bicephale :

You bet!

You are laughing but in reality, I think you know that while I was joking, I was actually being serious at the same time.

With this TP-Link TD-8840 (with Td-8841 firmware) modem, I have noticed another trend that holds true when compared to the Speedtouch 516. The lower the Downstream power number, the better my stats get. Nothing has changed on my setup. It was simply a matter of rebooting my modem to get a lower downstream power amount and higher signal to noise ratio margin.

note the lower downstream power output, TD-8840

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 15:37:29 EDT

Bicephale : You bet!

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 14:52:39 EDT

dslrocker3 :

said by Bicephale :

...or if one can climb on top of his pole to install a
MoDem over there!...

Knowing my luck, installing my modem on top fo the pole will probably give me worse results!

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 14:51:10 EDT

Bicephale : ...or if one can climb on top of his pole to install a
MoDem over there!...

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 14:43:23 EDT

dslrocker3 : My screenshots could lead people to believe that DSL works better at longer distances.

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 14:39:06 EDT

Bicephale : I'm sure we both know what it means to talk about
a paradox ("Bridge Tap")! Actually, the word i'm not
happy with is "envelope" as i tried to designate the
3rd curve in grey - which could easily lead to some
real misunderstanding. Sorry!

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 13:35:01 EDT

dslrocker3 : Haha, you don't have an appropriate word any more to describe this crazy paradox.

My TD-8441 (cough, cough. sorry, I mean TD-8840) does pretty much does the same with as the Speedtouch 516 did with "normal" wiring. However, the Speedtouch 516 seemed to enjoy "messed up wiring" more than my tp-link modem.

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 11:08:53 EDT

Bicephale :
More SNR Margin with a lesser "envelope" seems to
confirm the presence of a (...) paradox once again!

DSLRocker - Moving Around (2009-Nov-19)

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 10:35:39 EDT

dslrocker3 : Suddenly, the effects of my strange wiring do seem as great as before. This TD-8840 with TD-8841 firmware installed doesn't seem to care as much and doesn't seem to benefit as much by my strange wiring signal boost as much my Speedtouch 516 did. This leads me to wonder if the Speedtouch 516's attainable rates might have been exhagerated when used in combination with my improvised wiring solution.

normal, conventional wiring, with centralized microfilter in basement
improvised, improved wiring

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 09:15:43 EDT

Bicephale : Oh, and look here for documentation from other clone sources:

»TP-Link TD-8841 MoDem - Exploration & Discoveries Part 2

...especially the TelNet CLI "bibles"!

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 08:54:31 EDT

Bicephale : I believe that the readers who are interested into the
present thread should read this one (and vice-versa):

TP-LINK TD-W8920G toughts and review, cw30751, 2009-Jul-31

Re: Help/Advice with my Line Statistics please

Thu, 19 Nov 2009 00:08:36 EDT

Bicephale : A TP-Link TD-8841 clone for thirty-some dollars?

That's worth a trophee, i just didn't have time to
change the base...

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 23:59:24 EDT

dslrocker3 :

said by dslrocker3 :

Bicephale, I have a little riddle for you.... Can you figure out what the riddle is and what it means?

In case anyone else noticed, I have a TD-8840 v1.3 modem. There are no firmware updates available on the tp-link website. Tp-link uses the same packaging for the v1.3 harware version of the TD-8841 (with only a sticker on the side of the box to tell you which one you have in hand) and they both look identical with the exception of the TD-8840 lacking the USB port.

I went to the tp-link techncial support website and puprosely flashed the TD-8841 v1 firmware to my modem and it works fine. The web interface now seems more snappy and I have gained a quick setup option and a disconnect/connect PPP button. Doing this of course voids the manufacturer's warranty but honestly, who's going to actually pay $10 to $15 to ship a modem back for repair/warranty servicing when they only cost about $20 to $30 to begin with?

After taxes and currency conversion, my unit came out to exactly $34 (canadian dollars), purchased from Gentek on Ebay. »cgi.ebay.ca/ws/eBayISAPI.dll?Vie···MEWAX:IT

That's not exactly the lowest price I have ever seen on a DSL modem but considering that it's the Broadcom chipset (vs. the generally cheaper and more widely available Trend chip TP-Link modems) and also considering that it includes free shipping to anywhere in Canada or the lower U.S., it's not bad at all.

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 23:14:17 EDT

Bicephale : There is the HTTP server and also there is the TelNet one,
my suggestion is that you use both to generate logs that i
can try to combine later...

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 23:11:04 EDT

dslrocker3 :

said by Bicephale :

I'm more curious to
see long-term statistics, is it not possible to collect
DS/US CRC/HEC/FEC Error Rates by combining both
'DMT v8.07' and 'RouterStats', if necessary, as i did
explain in the post i've linked to previously?...

downstream CRC's would be possible as well as up and downstream FEC and HEC. However, in the spot in the modem's interface where it should show the upstream CRC's, it just says "N/A". "N/A" isn't going to help us make graphs for US CRC.

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 23:07:37 EDT

Bicephale : I'll leave the interleaving depth matter to our friends
who are customers outside Bell's territory since we
don't have Interleaving options and hence few of us
here are familiar with this concept of academic value
only, for all practical purposes... I'm more curious to
see long-term statistics, is it not possible to collect
DS/US CRC/HEC/FEC Error Rates by combining both
'DMT v8.07' and 'RouterStats', if necessary, as i did
explain in the post i've linked to previously?...

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 22:34:26 EDT

dslrocker3 : Bicephale, I have a little riddle for you.... Can you figure out what the riddle is and what it means?

DMT 8.07 now reports an interleaving depth for me. However, pings of 15 to 20ms are still possible. :)
DMT now reports downstream interleaving but upstream fast path! LOL!



take note of the different interleaving depths and delays for upstream and downstream

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 21:39:09 EDT

dslrocker3 : Bicephale, it's interesting that you mentioned how DMT 8.07 is missing some of the important error types in diagnosis mode.

I was just about to comment about how I've never seen 1 upstream CRC in any of my DMT 8.07 screenshots. And in the modem's web interface, the modem just says "N/A" for upstream uncorrected errors. So it would appear that the generic broadcoms just aren't capable of reporting upsteram CRC errors. :(

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 21:13:03 EDT

Bicephale : I think you'll want to revisit this post/thread:

»Re: OrbMT on D-Link DSL 2320B

;)

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 15:54:25 EDT

dslrocker3 : Jan 1 00:00:16 syslog emerg BCM96345 started: BusyBox v1.00 (2007.07.30-03:01+0000)
Jan 1 00:00:16 user crit kernel: eth0 Link UP.
Jan 1 00:00:20 user crit kernel: ADSL G.994 training
Jan 1 00:00:25 user crit kernel: ADSL G.992 started
Jan 1 00:00:28 user crit kernel: ADSL G.992 channel analysis
Jan 1 00:00:32 user crit kernel: ADSL G.992 message exchange
Jan 1 00:00:33 user crit kernel: ADSL link up, interleaved, us=800, ds=6016

I did also notice one more thing. This TD-8840 is noticbly faster at syncing. It doesn't always sync this fast but The Speedtouch 516 never syncs that fast (based on log entries). To verify the accuracy of the td-8840's log, I turned off my modem and used a stopwatch. From the time I hit the power button, sync was achieved in 39.59 seconds (log says 33 seconds). I just did this test again, and my td-8840's log said 20 seconds this time ("Jan 1 00:00:20 user crit kernel: ADSL link up, interleaved, us=800, ds=6016"

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 15:23:35 EDT

dslrocker3 : System Log

Date/Time Facility Severity Message
Jan 1 00:00:15 syslog emerg BCM96345 started: BusyBox v1.00 (2007.07.30-03:01+0000)
Jan 1 00:00:16 user crit kernel: eth0 Link UP.
Jan 1 00:01:23 user crit kernel: ADSL G.994 training
Jan 1 00:01:28 user crit kernel: ADSL G.992 started
Jan 1 00:01:30 user crit kernel: ADSL G.992 channel analysis
Jan 1 00:01:35 user crit kernel: ADSL G.992 message exchange
Jan 1 00:01:36 user crit kernel: ADSL link up, interleaved, us=800, ds=6016
Jan 1 00:02:36 daemon crit pppd[233]: PPP server detected.
Jan 1 00:02:36 daemon crit pppd[233]: PPP session established.
Jan 1 00:02:36 daemon crit pppd[233]: PPP LCP UP.
Jan 1 00:02:37 daemon crit pppd[233]: PPP LCP UP.
Jan 1 00:02:43 daemon crit pppd[233]: Clear IP addresses. PPP connection DOWN.
Jan 1 00:02:43 daemon crit pppd[233]: Received valid IP address from server. Connection UP.

This Tp-link definitely has better logs than the Speedtouch 516 with firmware 6.1.0.50. I never could quite understand why the Speedtouch 516 (6.1.0.5.0) has a "clear log" button when it randomly decides to do this on it's own anyway which renders its logs almost useless. Unfortunately, this tp-link model with the stock firmware does not show access conentrator information. The TP-Link model also clears the older log entry but does appear to at least reatain the information longer. Based solely for the reason that the tp-link retains all log entries for a greater length of time, edge to the tp-link td-8840 1.3.

As for attainable sync rates, my speedtouch 516 does better on downstream but when it comes to upstream, the winner is the tp-link td-8840. However, I believe that one of the main reason for having one of these broadcom 6338 based modems is to properly match up with IKNS peers and the crazy upstream values and problems that many other modems have difficulties with. verdict: edge to tp-link (assuming you are on IKNS and it is nearby. if this is the case, you'll probably need more help on the upstream to get full sync rates of your profile)

ease of use: This one clearly goes to the speedtouch 516. It is simply dummy proof with the intuitive set up wizard without the need to use some strange software utility. (see below when it comes to my comments about the web interface)

web interface: The Speedtouch 516's interface is clean and uncluttered. It is also very basic to the point where it provides you with almost no information about your connection. The td-8840's (at least with this frimware version) lack of a dummy-proof setup utility means that if you know absolutely nothing about configuring modems and DSL connections, you might want to go for the Speedtouch model. The TD-8840 is rather cluttered, partially because of the numerous features, but it does provide some nice options and shows much more connection information to the point that many users might not even find the need to use DMT 8.07. verdict: edge to tp-link for at least trying and not dumming it down so much to the point where you lose functionality/features/options. To setup your PPPoE connection, you have to go through mutiple levels of menus on the TD-8840 (advanced setup, wan, add or edit) but once you get familar with this generic interface used in many different modems, it shouldn't pose a problem. Strangely missing in the the stock firmware of the TD-8840 (1.3) is the ability to disconnect and reconnect the PPPoE session. I know that this firmware dependent but unfortunately, it appears that tp-link has abandoned support of this older hardware revision (there might be Telenet command to reconnect PPP).

cosmetics/buid quality: edge to the speedtouch 516. The TD-8840 (version 1.3 hardware) looks rather clunky and big. The plastic also seems less bendable (more sturdy) on the Speedtouch 516, at least in my opinion.

other: My TD-8840 does not exhibit as loud of a high pitch sound as my Speedtouch 516. The TD-8840 does still have a high pitch "hiss" sounds but it far quieter and far less annoying than my Speedtouch 516's even louder high pitch "whining" sound.

DMT (software utility) usability and tweaking: The Speedtouch 516 is able to show two things that the TD-8840 cannot. DSLAM vendor information and signal to noise ratio margin per tone. Both modems will allow tweaking within DMT if the Speedtouch 516 has firmware 6.1.0.5 installed. However, the ability to turn on and off specific tones/bits makes the TD-8440 the winner here.

Cost/pricing: I can see almost no reason to justity the higher price of a Speedtouch 516. If you are computer literate, you should have no problem configuring the TD-8840 for use and be able to tweak to your liking.

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 14:14:53 EDT

Bicephale : What a nice bargain opportunity! Have you started to play with this?:

From the ground up!, Bicephale, 2009-May-22

I believe the range between 10 % and 100 % should be of a particular interest...

Re: Help/Advice with my Line Statistics please

Wed, 18 Nov 2009 12:18:48 EDT

dslrocker3 : Wish me luck at my TP-Link TD-8840(A) version 1.3, firmware 3.06L.03-T1.0a-070730.A2pB021c.d17m vs. Speedtouch 516v6, firmware 6.1.0.5.0 comparison with my current line conditions.

The web interface is identical to the TP-Link TD-8841 version 2.0 so this means that anyone who is having a hard time finding the Td-8841 but doesn't want to pay the full price of a Speedtouch 516 would probably be happy with the TD-8840 version 1.3. And yes, it's comatible with DMT 8.07. Unfortunately, the firmware version is a little dated because of a discontinued hardware version (but not any more dated than my 6.1.0.5.0 Speedtouch 516 firmware).

Gentek is even including free shipping on this model when you purchase from its Ebay listing :)

tone selection option

4 ethernet ports
DMT 8.07 compatibility
connection still worse statistically when using conventional wiring methods

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 21:34:55 EDT

Bicephale : I thought perhaps we can do with noise what we do
when there's a mouse in the vincinity: put a trap to
keep it away from the kitchen - or if you prefer, use
another analogy where the radiator is a sponge and
noise is like excess water... Can you imagine?! Using
a radiator to trap noise! Ha! Ha! I must be tired!...

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 20:07:03 EDT

wayjac : You can't always reach the ideal......you do the best you can
All of my phone cable is what's in the second image

You can buy bulk stranded cat5e and most ethernet patch cables are stranded

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 20:02:17 EDT

dslrocker3 :

said by Bicephale :

I wonder, in the golden years of airborne TV there
were people who experimented with antennas made
from old car radiators... Would you, by any chance,
happen to have one or two of those used radiators
in your basement? I'm just curious!...

I don't think there is. What is this for, Bicephale? Are you looking for some parts to perhaps created some type of antenna at your place that somehow "deflects" away noise!!???? hahahahhaha!!!

I can tell you that I have my dad to blame for that crazy wiring in the basement. Even years ago, I questioned him about why he was doubling it up and he just told me it was a good thing but I never understood why he did it. It just happens to be by chance that what he did there really does help the connection.

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 19:53:01 EDT

Bicephale : I wonder, in the golden years of airborne TV there
were people who experimented with antennas made
from old car radiators... Would you, by any chance,
happen to have one or two of those used radiators
in your basement? I'm just curious!...

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 19:47:54 EDT

dslrocker3 :

said by Bicephale :

You're using two pairs as a single one!?... Well, if it
does you good i won't dispute your observations as
i'm aware of the fact that a "Bridge Tap" can cause
such hard to predict phenomenons... If it does help
then i guess maybe impedance is really affected, it
seems. Would you specify the length please?

:D

it's about 20 feet long but then it gets spliced into some other old phone wiring and insulated with electral tape. I can only assume it is doubled up there as well and the second wire part then runs probably another 20 feet to the jack on the main floor. From there, I daisy chained from that jack with 50 feet of cat5 up to the second floor (used only one wire per ring and tip). Using the end jack of on this run to the basement is the "extension cord trick" in action.

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 19:42:01 EDT

dslrocker3 : Wayjac, your first picture is exactly what I used to plug into my first jack and separate the house wiring. I know it's not ideal but it does work.

»/speak/slidesh···03ODA%3D

The black wire plugged into the RJ-11 is just normal phone wire. But I've found that cat5e ethernet wire is usually in non-solid strands as well.

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 18:56:02 EDT

wayjac : It sounds like you are using cable like this to connect wall jacks
[att=1,l]

Cable like this is normally used to connect wall jacks

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 18:53:09 EDT

Bicephale : You're using two pairs as a single one!?... Well, if it
does you good i won't dispute your observations as
i'm aware of the fact that a "Bridge Tap" can cause
such hard to predict phenomenons... If it does help
then i guess maybe impedance is really affected, it
seems. Would you specify the length please?

:D

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 18:25:46 EDT

dslrocker3 :

said by Bicephale :

You've lost me along the road, can you illustrate this please?

With a drawing, or even a picture maybe!...

I know you aren't really supposed to attach phone wiring like this but it makes it much easier to attach it to the junction because this makes the wire thicker.

On my line, it just happens to also affect the balance of the line and seems to counter the bridge tap and reflections.

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 18:15:26 EDT

Bicephale : You've lost me along the road, can you illustrate this please?

With a drawing, or even a picture maybe!...

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 15:18:51 EDT

dslrocker3 :

said by Oinktastic :

I've experienced a bit of that as well. Those wires are extremely thin. I'm really surprised they even carry voice as well as they do, let alone data.

I had my wires doubled up before I reran the newer ones and the SNRM was actually better back then as well.

I've found that in most cases, using wiring that is made of strands is fine as long as it makes proper contact. But if you don't double up, it's really quite difficult to connect even after twisting the wire. Even when using s single wire, things are usually fine once it's actually connected - it's just more of a pain.

I don't believe that my line was actually helped by the fact that it was easier to connect the wire when it was double up but I think that doing this somehow made a better impedance match.

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 15:13:44 EDT

Oinktastic : I've experienced a bit of that as well. Those wires are extremely thin. I'm really surprised they even carry voice as well as they do, let alone data.

I had my wires doubled up before I reran the newer ones and the SNRM was actually better back then as well.

Re: Help/Advice with my Line Statistics please

Tue, 17 Nov 2009 01:08:44 EDT

dslrocker3 : Tonight, I was concentrating on the run from the basement to the second floor that has non-standard wiring for dsl of daisy chaining 2 jacks together and plugging in that magical extension cord in the middle jack.

The wires in the basment for this run are doubled up on each of ring and tip. When I removed the extra wire on each of the ring and the tip, I didn't notice much of a difference at first. But then, I started noticing that noise margin almost appeared to be "leaking". Within a few minutes of having my modem hooked up with the wiring like this, the signal to noise ratio margin was widely fluctuating anywhere from 20.5 to 25.0dB. As soon as I return the wiring to way it was before, using 2 wires for ring and 2 wires for tip in the basement, the signal to noise ratio has stabilized almost permanently at 25.0dB.

Re: Help/Advice with my Line Statistics please

Sun, 15 Nov 2009 23:36:52 EDT

dslrocker3 : never mind me, I'm just documented my averaged error rate for the last 24 hours for future reference with my wiring in its current state.

It's time for me to think up my next experiment. I think that I know what you want to see, bicepahle (and it has to do with ferrite cores).

(dmt20091115_2333 centralized microfilter in basement, dsl on end of 100ft cat5.p)

Re: Help/Advice with my Line Statistics please

Sun, 15 Nov 2009 19:28:33 EDT

dslrocker3 : those stats and bit loadings look almost identical to my line with the extension cord trick in action - attenuation and power outputs the same too....

here's what it looked like before I changed the basement's main junction point earlier in the week:
»/speak/slidesh···03MjA%3D

With my current main basement junction (the phone jack) and using the extension cord trick, things are statistically about the same but the bit loading looks kind of ugly now (but better for my voice line, less noise on the line).
»/speak/slidesh···03ODA%3D

and without the extension cord boost and with central filtering: »/speak/slidesh···03NjA%3D

... not much difference to me in RCO and noise margins but vastly different bit loading graphs depending on how I wire the basement... There's little doubt to me that the first screenshot in this post looks the best. It's too bad it causes a phone to pick up hum when it's near electronic devices.

Re: Help/Advice with my Line Statistics please

Sun, 15 Nov 2009 18:36:32 EDT

Bicephale : Who knows what's next in it for you, maybe the spectral response will flatten out like this:

Bell found how to throttle MLPPP MultiLink??? Maybe :S, InvalidError, 2009-Jul-22

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 19:17:50 EDT

dslrocker3 :

said by Bicephale :

It's good to be able to tweak in an emergency case...

:p

It's also good to know that I generally have fundamentally sound wiring inside the house ready to go should that dreaded bridge tap ever get chopped off, or at the very least, connected to someone's house and terminated at that end.

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 18:28:20 EDT

Bicephale : It's good to be able to tweak in an emergency case...

:p

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 18:00:55 EDT

dslrocker3 :

said by Bicephale :

Too bad, it could have been useful but i don't
even recall where i found this. Maybe it's from
the 620 series ("business" grade), i wonder!...

:huh:

It would have been great because then a user could just take the higher possible profile and then just set the connection speed accordingly.

Part of this whole thing with my line does bug me how the connection is worse at the NID and part of it intrigues me. Based on signal to noise ratio margins and attainable rates, the defacto standard way of wiring yield lesser results but I'll probably end up just connecting everything like that as it would seem to be the safer route.

This whole thing with me moving my wiring around for the last few days really was just for my own knowledge and so that I could try to better understand how my line worked. I'll admit that this knowledge would have been much more valuable back in the days of the my 1728/640 and 2496/800 profiles where I had to squeeze every last drop out of my line. From 2002 until early 2009, I would have to say that I uneccessarily got stuck on 1728Kbps profiles because back then, I never really examined the different modem and wiring options. This recent understanding of my unique outdoor wiring stuation and how it interacts with my internal wiring provides "emergency" insurance procedures should by line conditions ever worsen.

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 15:14:59 EDT

Bicephale : Too bad, it could have been useful but i don't
even recall where i found this. Maybe it's from
the 620 series ("business" grade), i wonder!...

:huh:

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 14:21:33 EDT

dslrocker3 :

said by Bicephale :

xdsl maxspeed tx=640 rx=6016

I can tell you that for firmware version 6.1.0.5.0, this command does nothing. The modem did accept the command but did not appear to yield any results for me.

I issued the command of
xdsl maxspeed tx=512 rx=5056

After issuing this command, I told the modem to "DSL Resync." (without powering off because that might clear the command from the modem's memory). My modem still synced at 6016/800.

I then wondered if the command just limited the throughput internally while still maintaining the normal sync rate. I ran a speedtest at speakeasy.net/speedtest and got a result of 4995/652 Kbps which is still in line with the 6016/800.

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 13:36:39 EDT

Bicephale : I know that's a modest challenge in comparison
but i thought you could try it, in the meantime:

»Re: Help/Advice with my Line Statistics please

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 13:28:19 EDT

dslrocker3 : I am going to keep opening Bell tickets until they remove the bridge tap! hahahahahaha!!!! Bicephale, do you dare me to?

But really, if I have been able to correct Bell's outside wiring by changing my inside wiring, does it really matter? I'm just doing all this nonsense to try to learn a few things and to see if there's a better way for me to achieve my current connection results.

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 13:04:33 EDT

Bicephale : I don't see how you'd be able to get Bell to remove a
"Bridge Tap" now that they sent out one of their guys
to move your line to an IKNS remote, my bet is you're
better just to accept your destiny!...

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 12:39:08 EDT

dslrocker3 :

said by Bicephale :

What
it does is to show that because of Bell's practice we
may encounter weird situations like yours where our
conventional wiring methods can expose Bell's failure
to use adequate wiring on their own side!

When I followed conventional DSL signal optimization practices, the results were a lot more similiar to the test results when having the modem connected directly to the NID (which were worse).

I'm also not doubting that each one of my small wiring changes could possibly make different modem models act differently as well.

maybe, I should just leave my messed up wiring as it is for now to see what happens.

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 12:30:39 EDT

Bicephale : It seems you are using a "Bridge Tap" to reduce the
effects of another which not only happens to be out
of your reach but that also has a picky behaviour.

You may very well manage to fit all of the unfiltered
wiring into metal cans while using central filtering for
the voice side but even if you don't it doesn't mean
the rest of us shouldn't continue to clean up... What
it does is to show that because of Bell's practice we
may encounter weird situations like yours where our
conventional wiring methods can expose Bell's failure
to use adequate wiring on their own side!

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 12:05:35 EDT

dslrocker3 :

said by Bicephale :

Luckily it's not 30+ dB, 30+ dB would have sounded
obscene... What about making a roll that fits into a
paint can? That way you'd still have the length but
without the antena effect.

as I was saying, all signs point to these conditions for optimal dsl in my house
- a 100 ft cat5 connected in basement
- a chain of multiple jacks, extension cord in middle jack, modem at end jack on a seprate run to the basement (separate from the 100 ft cat5)

Right now, the 100 foot cat5 is run up 2 floors. But I guess I could do the paint can trick but that means I would have to run another cable upstairs to keep the dsl and voice phones on separate jacks/runs to the basement. Plus, who knows, the combination of variables that I just stated might not be all of it. It's possible that some of the other house wiring is also contributing to the signal boost which would then also need to be left unfiltered.

With the need to leave all these wires unfiltered, is there even a point to centralized filtering in my case?

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 12:02:46 EDT

Bicephale : Alright, two empty paint cans!...

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 11:59:23 EDT

dslrocker3 : I just checked one more thing... I can't put things the back the way they were before because that causes voice line interference. So you might be thinking why not only put that chain of jacks and extension cord onto one junction and filter everything else. I already thought of that. I went and filtered that 100 foot white colored cat5 cable that was directly feeding my modem yesterday, and guess what happened? The signal boost dissapeared (modem not connected to filtered jack obviously). This would mean that with my current wiring in my basement, that's now one more variable that has to be in place for me to achive the best possible connection. That 100 foot cable has to be left unfiltered.

To get centralized filtering without the loss of dsl signal quality, it would appear that I now have to go get another 100 foot cat5 cable to feed the phones on the upstairs floor. Then, I would be able to leave the other 100ft cat5, the chain of jacks, and the extension cord unfiltered while filtering everything else that has a phone device attached. This is enough to give me headache!

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 11:59:01 EDT

Bicephale : Luckily it's not 30+ dB, 30+ dB would have sounded
obscene... What about making a roll that fits into a
paint can? That way you'd still have the length but
without the antena effect.

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 11:18:47 EDT

dslrocker3 : Hey Bicephale, look at what resurrected itself! The magical and crazy extension cord signal booster is back. Don't ask me to explain it.

I have found that since yesterday's rewiring, the extension cord "signal booster" only works if 2 things are true:
1. centralized filtering must be disabled in the basement
2. modem has to be hooked up to the very end jack of the daisy chain and the "magic" extension cord has to be attached in the middle jack

If you look at one post before this one, you will see that there is still a boost in signal at only this one jack when disabling centralized filtering. The extension cord boost is no longer as extreme as before but I'm tried a few times at reconnecting at the extension cord games have once again begun!

But look at how the "bridge tap trenches" are still there compared to before. My modem just doesn't seem to care about those any more, based on this screenshot.

So now I'm back to the same dilema: back to centralized filtering and lose 2 to 3% RCO and 2dB of SNRM and leave cetralized filtering disabled and increase attenuation, and use all sorts of strange methods to improve dsl signal.

I think it's a pretty good dilema for me to be in. No matter which choice I make, I've at least made my voice line cleaner sounding.

(dmt20091114_1139 no centralized filter, dsl at end of daisy chain, extension cor)

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 10:44:16 EDT

dslrocker3 : and my strange beast of a line lives!!!!

Now, I have discoverd that the "dsl signal boost" that occurs when I remove central filtering is still possible but only through one of my jacks. The "boost" is still possible but only when connecting my modem to the end of that daisy chained jacks (the middle jack was where the magic extension cord was previously attached). Has my DSL really become that picky or is just that most people don't check this many variables and this type of dsl signal strangeness usually isn't noticed?

(dmt20091114_1037 dsl at end of daisy chained jack, no central filter.png)

Re: Help/Advice with my Line Statistics please

Sat, 14 Nov 2009 10:33:18 EDT

dslrocker3 : Strangely, after changing my house wiring so that the pictured jack is the first junction point, I not longer get a dsl signal boost by attaching all of the voice wiring. The results are now virtually identical when removing the centralized filter with the exception of the attenution being higher.

I'm sure I could get that "dsl signal boost" back but that would mean rewiring the basement the way it was before and it would also mean losing my centralized filter there. I'm not sure about everyone else here but I think losing 1 to 2% of RCO is worth the trade when in return I got a centralized filter, a basement test jack, and my upstairs phone no long humms.

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 21:01:25 EDT

wayjac : Thanks for the image, never seen one of those
The image should be ok as long as its less than 600 pixels wide I don't think file size is a concern any more except for those that don't have a high speed connection

I'm thinking if you have spare unused test jacks you could use one or two inside the house

[att=1,l]

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 20:38:40 EDT

dslrocker3 :

said by wayjac :

said by dslrocker3 :

However, doing that would leave me unable to have a true secondary basement demarcation point.

If you have a nid with unused test jacks, they can be used in the basement as a demarcation point

Can you post a better picture of the "junction jack" that looks like something you didn't have before

Hello Wayjac

I did have that jack connected in the basement before but it was only a secondary junction point that fed a tertiary junction point that fed the house's voice wiring.

Now, the jack serves as a primary junction point.

Yes, I do have an NID test jack outside. It's just nice to be able to separate everything again once inside.

Wayjac, you should like this picture better. The picture before was from my cell phone so it wasn't good quality and plus I resized it smaller so that it wouldn't take up too much space on the dslreports server. I'm not sure what file size this site recommends/enforces for pictures but I'm sure a mod will inform me if my file size is out of line on this picture. It's 1MB right now and I've kept it intact at the original resoluation so that you can get a good look, Wayjac.

crap, looking at the close up of the basement wiring, I just realized that I didn't do a great of job terminating the ring wire on the left secondary junction.

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 20:38:31 EDT

Bicephale : Indeed, but if both carbon protectors happen to be
defective simultaneously who knows what it could
cause some day, when you least expect it?...

Oups! No carbon protectors according to the next
picture! Hummm... Well, what about that bakelite
support itself going berzerk?... Mabe those things
weren't built to last until WW3!!!

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 20:25:07 EDT

dslrocker3 :

said by Bicephale :

usually means beautiful... Why not imitate nature? Get
rid of those ugly museum artifacts which have no other
purpose than to cause bizarre failure modes someday!!!

at least I removed the grounding wire from them months ago and at least I don't still have fuses in mine, unlike our friend Davde in the Teksavvy forum! (davde does have a ground in the outdoor NID)

(davde's indoor junction point)
»/speak/slidesh···YXJ0PTIw

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 20:21:19 EDT

Bicephale : Right now the behaviour which we thought could result
from the presence of a "Bridge Tap" has vanished but i
believe i have one of those too and i couldn't observe a
similar behaviour at my location... Perhaps it will return,
maybe never. Is it possible some Bell worker decided to
recycle that unused length to service a neighbour?...

Too many variables, so few answers! All we know is at
the moment it doesn't really matter how you rewire, as
long as you observe conventional rules, that is. So, my
advice will be to simply do whatever serves you best...

Euh... But in nature a reasonable amount of simplicity
usually means beautiful... Why not imitate nature? Get
rid of those ugly museum artifacts which have no other
purpose than to cause bizarre failure modes someday!!!

For this reason i like WayJac's suggestion even more: a
star topology starting from the NID would look just fine!

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 20:21:03 EDT

wayjac :

said by dslrocker3 :

However, doing that would leave me unable to have a true secondary basement demarcation point.

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 20:02:17 EDT

dslrocker3 : If you look at my above picture of my basement house wiring, you will see that the junction jack feeds secondary junction points, one for dsl and another for voice.

I know a better way would be to attach an RJ11 to that cat5 cable feeding my DSL but I don't have a crimping tool, the rj11 ends, nor the knowledge (yet).

But assuming that I don't attach an RJ11 end to that 100ft cat5e wire, would it have been better to just hook up the DSL's cat5 wiring into the screws of the jack box? I know that would be simpler and would provide the least amount of steps for the signal to be delivered to the modem. However, doing that would leave me unable to have a true secondary basement demarcation point. Opinions? Xbell, are you around perhaps to answer this?

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 18:18:49 EDT

Bicephale : If someday you happen to be intoxicated enough to
want to chase extraterrestrials with your phasor try
putting an A.M. radio on top of your calculator while
you do some financial number crunching or else!...

The buzz man!!! Ask dBSanFTE...

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 18:09:38 EDT

Bicephale : Ma> It is indeed a very complex matter.

But there's only a finite number of turns on a toroïdal
core, fifteen for a phone cord with molded connectors
on a 1½" ferrite core as i recall... If a shunt capacitor
is to be mounted on one of those turns outside of the
core that limits the number of possible connections, a
variable capacitor should simplify the rest of such an
experiment. It might work and happen to be more fun
than what dBSanFTE suggested, not to mention it may
also be accessible to many more persons compared to
the few who chase extraterrestrials with their phasor!

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 18:08:13 EDT

dslrocker3 : my days of 55 and 56% downstream RCO are now over since my extension cord trick no longer works.

Now, my RCO's fluctuate between 58 and 59%. But in the meanwhile I decided to take Xbell's previous advice and have the wire coming into the house first go to the jack as the main house junction. The modem is now connected to one side of this and the rest of the house wiring to the other side. At the very least, the wiring looks more balanced now (splitting now rather than just daisy chaining).

Interestingly, there is a noticeable drop off (but minor) at the basement jack but once again, things return to normal once plugging in the modem to my normal upstairs jack.

On a side note, I did narrow down a "ground hum" sound in one particular phone to when a nearby tv was on or if I placed the phone near my laptop. If I moved the phone away from my laptop or if I turned off the tv, the hum immediately would dissapear. However, after today's wiring change, I cannot reproduce that type of almost "ground humming" sound at all in the upstairs phone. In fact, now it's so clean that I almost can't tell it's off the hook when I listen to it's handset speaker. However, now it is picking up interference that "sounds" like the blinking LEDs and the sound does appear to be in sync with the flashing lights on my modem. I think that people who have had DSL for a while know which type of high pitch "blinking sound" I'm refering to. The sound does appaer to get softer when plugging in more filters (now 2 into this phone and also 2 in the basement). So when it comes to this part, how in the world does a small change in the basement's wiring make a phone now unable to pick up RF interference that sounds like ground hum 2 floors up? (The interfering devices that cause the hum are also 2 floors up).

At the very least, my DSL connection has not been harm at all by this change and my voice line now sounds cleaner on one of my phones. :) Also, I can now disconnect any combination of all house wiring and/or all DSL wiring in the basement without the use of any tools and without the need to go outside.

At basement jack
after today's rewiring, upstairs
to jack first, then secondary junction

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 17:25:47 EDT

Bicephale : If you want to limit your UpStream power try to reduce the
correspondant BitRate... I found this CLI searching around:

xdsl maxspeed tx=640 rx=6016

I think the syntax is for Thomson's later v7.4 FirmWare but
you have nothing to loose by experimenting with it. My ST
isn't on-line at the moment, i could have waited to try this
myself but i thought this was a perfect opportunity...

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 15:42:22 EDT

dslrocker3 : Although this goes completly against "normal" phone wiring theory, I can see absolutely no benefit in my case to having the modem hooked up to the NID. This screenshot once again shows lower noise margins, lower attainable rates, and higher RCO's when connected direclty at the NID test jack. I let it run for a little bit and although I didn't let the conection stay up for very long, I think that I can also safely say that hooking up my modem directly to the NID also does not appear to help the errorr rate and I would classify this amount of errors as not being significantly different than inside the house.

This would tend to tell me that the vast majority of line errors are coming from outdoor sources.

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 15:01:38 EDT

dslrocker3 : Even though it is not shown in this comparson, I believe since being connected to an Ikanos RDSLAM, it's fair to say that my results have always been better when the downstream transmission power is lower. 8.5 downstream transmission power seems to be the magic number for me and 9.5 being the worst possible number. We should keep in mind that I have always got worse results at my NID than inside the house on the upper floor.

Please notice the effect that putting something in between the modem and NID. I'm willing to bet that my impedance inbalance can countered by putting various objects between the NID and modem (ie. extension cords in weird places, filters, jacks, extra long cords attached to modem, etc.).

direct at NID
At NID but through z-blocker splitter

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 14:57:04 EDT

dbsanfte : phasors aren't tricky, they make things easier. In my E&M class we did it all with the exponential definitions of sine/cos, if you want a headache, try that. =P

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 14:51:57 EDT

Mauricio :

said by Bicephale :

Impedance matching is somewhat more complex than
dealing with resistance alone, most unfortunately!...

It is indeed a very complex matter. I study avionics and I had a full semester on AC circuits. It's all about vectors, which makes it tricky ! ;)

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 13:20:49 EDT

Bicephale : Impedance matching is somewhat more complex than
dealing with resistance alone, most unfortunately!...

:o

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 11:38:04 EDT

dslrocker3 :

said by Oinktastic :

said by dslrocker3 :

I'm still mourning the lose of 2dB of signal to noise ratio margin and the death of my magical extension cord. :(

Maybe your cord is all out of electrons :(

very funny.

I just had a thought about impedance matching. I know that this goes completely against the "rules" of phone wiring but who says that I couldn't use more than one pair of inside wiring to provide my DSL signal? Could experimenting with this while purposely and incorrectly conecting multiple pairs change the resistance level?

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 11:23:37 EDT

Oinktastic :

said by dslrocker3 :

I'm still mourning the lose of 2dB of signal to noise ratio margin and the death of my magical extension cord. :(

Maybe your cord is all out of electrons :(

Re: Help/Advice with my Line Statistics please

Fri, 13 Nov 2009 09:56:03 EDT

dslrocker3 : In the end, it would appear that "obsessive compulsive filtering" yielded results that were no better. They weren't any worse either.

I'm still mourning the lose of 2dB of signal to noise ratio margin and the death of my magical extension cord. :(

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 22:20:16 EDT

Bicephale : Yes, low-pass but that may be an accessorial
coincidence only as this is what i had in mind:

An RF Choke looks symetric by nature and not
this schematic representation but that's about
the same design anway:

Help/Advice with my Line Statistics please, Bicephale, 2009-Jul-30

So, i thought if RF Chokes happened to modify
the impedance why not help it further more?...

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 21:43:07 EDT

Oinktastic :

said by Bicephale :

...I know i'll pass!

Speaking of passing. That looks like a low-pass filter. Am I right? Or does it depend on the actual values of inductance and capacitance?

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 20:22:15 EDT

Bicephale : I think a more disciplined method would be to
use a circuit which looks as shown above, the
problem being it must be a nightmare to figure
out what's the magic mix. If you Google a bit
in search for "Transmission Line" theories and
the "Smith Chart" in particular, you shoud get
a slim idea of what's involved. I know i'll pass!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 19:33:50 EDT

wayjac :

said by dslrocker3 :

I get a higher signal to noise ratio margins and attainable rates when I purposely attenuate the line more.

Sometimes being very close to the dslam is as bad as being too far


dsl0: up G.DMT Annex A fast Rate:6016/768 Max:9356/932
dsl0: Margin:15.0/16.0 Atten:16.2/7.5 Power:4.7/10.7
dsl0: Country: {D1} Vendor: {CTNW} Specific: {0}

The modem has a loss of sync about every 45 seconds

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 19:27:44 EDT

dslrocker3 : This diagram represents my current setup. The DSL side is more traditional than before as there is no longer a need to feed the modem through all those daisy chained jacks as the old extension cord has lost it's magical effect.

Removing those microfilters in the basement is what now results in a slight signal to noise ratio improvement.

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 19:00:30 EDT

dslrocker3 :

said by Bicephale :

I think it's not as much a matter of attenuation than of impedance...

So it's probably not the sum of the voice wiring that is causing the slight improvement but rather a particular wire that is providing a balancing effect, just as that extension cord use to.

I'm looking up stuff on Google about impedance matching circutry. »www.patentstorm.us/patents/63696···ion.html

It's so much nicer to think about ways to fix my "tragic" dsl line conditions when I don't have to.

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 18:25:38 EDT

Bicephale : I think it's not as much a matter of attenuation than of impedance...

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 18:18:01 EDT

dslrocker3 :

said by wayjac :

I'm trying to keep track of what you're doing with the filters/wiring
I was thinking with the short loop the inside wiring would return to normal

Hello Wayjac,

The strange behavior continues. I get a higher signal to noise ratio margins and attainable rates when I purposely attenuate the line more. As I was saying before, my extension cord trick no longer works. I'm not sure what that's about. I must have either damaged the wire (don't think I did) when I removed the stapling from the wall or there was a staple in the wire before causing a short that possibly created more of attenuation effect (which might have actually helped).

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 18:14:57 EDT

Bicephale : Well, there's the angel part and then there's the pimp one...

This is how the world works i guess!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 17:59:49 EDT

wayjac :

said by dslrocker3 :

It's funny that Wayjac asked if I was going to filter at the basement and then at the devices too and I said that I wasn't.

I'm trying to keep track of what you're doing with the filters/wiring
I was thinking with the short loop the inside wiring would return to normal

Bicephale is filter pimp #1
dslrocker3 is filter pimp #2
rankings are subject to change at any time

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 17:58:02 EDT

Oinktastic : Thank you Your Excellency. I knew I felt an angel's presence with me today...

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 17:54:02 EDT

Bicephale : Hi Oinktastic,

The Filter bless you my son, may the RF Choke be with you!

The last time i heard of DeadPool relatively to this it wasn't
conclusive. Things have changed since last year so i would
not put too much hope into an intervention but good luck!!!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 17:32:32 EDT

Oinktastic : Greetings Father,

CC: Your Excellency

Dslrocker, are you located anywhere near me by any chance? I'm on the members map, McCowan and Ellesmere, Scarborough. Due to the very noticeable spectrum loss in the AM radio station frequency bands, I'd say we've got approximately the same amount of AM radio interference on our lines.

Though your line is short enough and you're close enough to the remote for it not to trouble your new line, it's still very noticeable. In fact, the holes in the spectrum reported by your modem, which correspond to AM radio stations, are pretty much the only missing spectrum your modem reports.

Also, does anyone know if Deadpool still saves people like us?

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 17:07:00 EDT

Bicephale : I strongly suggest you avoid trusting the Thomson SpeedTouch
5x6 internal counters because these are just unreliable over long
periods, that's one of the reasons why i promote the "Diagnosis"
mode in 'DMT', for example. I've seen some counter values being
silently reset and most especially the Error Rate ones!... When i
do some SpreadSheet work in order to calculate "Delta" values it
makes such resets all too obvious when the rest isn't affected...

Don't worry about being the priest of "Obsessive Filtering" as i've
just checked my installation and it turns out i'm The Pope: i have
two GNet BB0001 micro-filters connected to the Fun-Jin ADS302G
ADSL Filter which came with my TP-Link TD-8841 and that one is
connected to my WilCom PS-15-01 i3S xDSL POTS Splitter so you
can call me "Your Excellency" from now on if discussing in public!!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 16:10:54 EDT

dslrocker3 : I would run Diagnosis mode but even in my noisy periods, the errors aren't significant. I fear running diagnosis mode would just return results that change from 0, 1, or 2 and aren't gong to show much more than an almost flat line. I'm looking more for counter totals but over a 24 hour period.

NOOOOO!!!!! A just got 4 upstream CRC's in the past 35 minutes. lol. For now, I can live with 1.5 CRC's per hour.

about the filtering... what would you call it when someone double filters all of the voice wiring in the basement and then attaches 1 or 2 more filters to each phone device?

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 16:09:19 EDT

Bicephale : "Obsessive compulsive filtering"?...

Beware, this could turn into a religion and make you a priest!!!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 16:07:20 EDT

Bicephale : Ha yes! We both know that of course but i tend to forget...

Thanks for reminding me!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 15:47:11 EDT

dslrocker3 :

said by Bicephale :

If it were under FastPath that would be even more interesting!

You and I both know that the modem will never report fast path with IKNS peers. My first hop today is showing as 14ms. If that's interleaved, then who cares what they call it! :)

We also both know that there were more CRC errors than this before. Even yesterday after the wiring change, my error rate didn't decrease in any way. It's funny that Wayjac asked if I was going to filter at the basement and then at the devices too and I said that I wasn't. But today's CRC magic trick can only be attributed to "obsessive compulsive filtering" if it holds.

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 15:40:44 EDT

Bicephale : If it were under FastPath that would be even more interesting!

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 15:30:04 EDT

dslrocker3 : Could this be? an error rate that is lower but with a lower signal to noise ratio margin than before?

I know it's premature at this point but if this keeps up for more than a day, the CRC error total could be quite "interesting".

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 14:04:28 EDT

Oinktastic : Time for Bell to start thinking about shielding their wires.

Some crazies speak of this absolutely brand-new, revolutionary, saviour-of-all new conduit. The name's a little fuzzy as the prophesy was translated many times before I caught a glimpse of it, but I believe it was called something like: Fye-ber Hopticks.

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 13:42:58 EDT

dslrocker3 :

said by Bicephale :

...and/or use RF Chokes everywhere to see what happens...

If you take a look at that the bit loading graph, most of the "strangeness" happens between 600 and 700 KHz. Coincidently, there are radio stations in Toronto at 590, 640, and 680 KHz.

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 13:13:49 EDT

Bicephale : ...and/or use RF Chokes everywhere to see what happens...

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 13:11:22 EDT

Bicephale : With 'DMT' running in "Diagnosis" mode you could monitor
the SNR Margin level while a friend or relative makes your
phone ring a few times every five minutes for as long as
they can, hoping the log-file can reveal such interaction.

:)

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 12:47:37 EDT

dslrocker3 :

said by Bicephale :

I'd sugest you go for the lower RCO: it's convenient as it
will allow you to move the MoDem around and this is what
Bell opted for when they 1st launched their deployment...

The non-centralized filtering system is only reducing RCO by 1 to 2% and incraeses signal to noise ratio 0.5 to 1.0dB with my current setup.

I think I might take the "safe" way out for now because connecting an disconnecting devices without centralized filtering makes my modem lose sync. If phone devices can do that when plugging them in, what else can they do to the DSL connection when they are in use? I suspect that not using a central filter in the basement may cause the the signal to noise ratio margin to fluctuate more when the phone rings.

For now, I have all voice wiring double filtered in the basement and each device now how an additional 1 or 2 filters plugged directly into them. This results in each phone device effectively being filtered at least 3 to 4 times. LOL. Let's see if that helps any with error rates.

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 12:13:53 EDT

Bicephale : Or do nothing!!! A 22+ dB SNR Margin that's comfortable...

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 12:09:07 EDT

Bicephale : I'd sugest you go for the lower RCO: it's convenient as it
will allow you to move the MoDem around and this is what
Bell opted for when they 1st launched their deployment...

You signal is strong enough to sustain a fair amount of RF
Choking, even if i doubt it would have noticeable effects.

Why not optimize that RCO and RF Choke simultaneously?

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 11:44:39 EDT

dslrocker3 : Unfotunately, no matter what I do, I cannot get that magical extension cord to do anything positve any more....

These screenshots would appear to confirm what I was seeing yesterday. I get slightly higher attainable rates and lower RCO's when I do not use a centralized filter in the basement. I went and tried my modem in the basement and you will see that the best results happen even there when I reconnect all of the house's wiring (not filtered). So the question is this: do I go for slightly lower RCO's and slightly higher attainable rates but not using a filter in the basement or do I lose 1% to 2% in RCO (worse) and take what is probably the "safer" route?

house wiring···fect.zip 292,717 bytes

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 11:18:41 EDT

Bicephale : The humming frequency itself being 60 Hz i doubt it has a
significant effect even if present at the unit's WAN input,
but if it comes from a noise injection site coupled to aerial
power wiring picking up airborne noise on top of the street
then it could be of concern, although unlikely in a case as
yours because the DSL signal happens to be too strong!...

Re: Help/Advice with my Line Statistics please

Thu, 12 Nov 2009 09:57:15 EDT

dslrocker3 : I have what sounds like a ground hum on one particular phone and it only happens on one of my jacks (others phones don't have the humming sound on this jack). it gets worse if i put the phone on top of my laptop. I wonder if the interference my laptop provides could also affect DSL.

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 21:05:10 EDT

Bicephale : Nah! Not that bad, maybe something like this:

»Re: new customer need help interpreting line stats

...but to a lesser degree, hopefully!!!

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 20:58:59 EDT

dslrocker3 :

said by Bicephale :

Most unfortunately it's not close enough to open a ticket...

:p

Yeah sure, I should open a ticket. I just can't wait to see what will happen to my 6016/800 Kbps fast path profile if I do that! Maybe I'll get the famous 64/64 Kbps profile just for being a smart ass with Bell!

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 20:57:35 EDT

Bicephale : Most unfortunately it's not close enough to open a ticket...

:p

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 20:56:08 EDT

dslrocker3 :

said by Bicephale :

With a "Bridge Tap" and an IKNS remote only making echoes
worst my bet is you're condemned to proceed through trial
and error, you can't just trust normal rules anymore. Enjoy,
these could happen to be quite stimulating times!...

I just realized that if there is indeed a bridge tap on my line like the Bell technicians have said before, it must be very nearby because my loop to the remote is already very short. This just means even more craziness with my line!

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 20:40:09 EDT

Bicephale : With a "Bridge Tap" and an IKNS remote only making echoes
worst my bet is you're condemned to proceed through trial
and error, you can't just trust normal rules anymore. Enjoy,
these could happen to be quite stimulating times!...

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 20:32:47 EDT

dslrocker3 :

said by wayjac :

Are you going to fit a filter at each phone and also filter in the basement

nope, I would remove the filter in the basement becaause preliminary results show that my connection is slightly better without one there. I would just attach a filter to each phone directly.

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 20:19:16 EDT

wayjac : Are you going to fit a filter at each phone and also filter in the basement

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 19:45:55 EDT

dslrocker3 : That's a true point that the differences are small and that the screenshots only show a connection in action for a split second. I did repeat the process for a few times and each time, the connection was better in signal to noise ratio margin when not filtering the house's wiring. Time will tell and maybe after a few days, I'll try going back to connecting a filter directly to each phone. On another note, since the extension cord no longer helps, I have to say goodbye to "strange wiring". I went and switched the "strange wiring" to the filtered side and the 100ft cat5e direct to the basement to the unfiltered, dsl side of my wiring.

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 19:40:25 EDT

Bicephale : Before i make any further comments please allow to bring to
your attention the fact that those snapshots show up times
which are insufficient, especially if one considers that you'd
need to let it settle for hours given the very low error rate...

It's even more delicate when the level changes are so small.

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 18:09:00 EDT

dslrocker3 : Here are 4 DMT screenshots. They show how using a centralized filter actually makes my signal to noise ratio go down! "Blue wire" refers to the wiring that I was using to feed my DSL jack previously and "white wire" refers to the one that I am now using to feed my modem. When the centralized filtering was removed, all house wiring was left connnected but phones were obviously disconnected. The results go completely against what I would expect. I do not believe that it is actually the centalized filter that is causing the drop in signal to noise ratio margin but I do suspect that the extra attenuation added to the loop when the filter is removed helps the connection. It's a shame that my strange extension cord is no longer able to boost my signal.

This information appears to be telling me to go back to attaching a filter to each phone device.

dmtscreensho···lter.zip 228,849 bytes

Re: Help/Advice with my Line Statistics please

Wed, 11 Nov 2009 18:03:44 EDT

dslrocker3 : If we remember, I have a line where plugging in an unfiltered extension cord boosts my signal to noise ratio margin a couple of dB and gives me a an attainable rate that is a few hundred Kbps higher. I have now removed that extension cord but I something odd has happened. Now, even plugging in that extension cord to exactly where it was before makes no difference. There is no longer a "signal boost" provided by that extension wire. When I removing the staples from the wall, I did notice that the staples were suspiciously close to the wire. Using a staple remover wasn't sufficicient because the gap between the wire and the staple was too small so I had to use my finger nail. ouch! I suspect that there might have been a short in that extension cord. but why in the world did it boost the signal with a short!!!!???? That is one theory but the other theory is that I damaged the signal boosting extension cord when removing the staples...

before w

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 23:31:25 EDT

Bicephale : Thanks for keeping us informed, good luck with your schedule!

:D

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 22:36:14 EDT

Mauricio :

said by Bicephale :

Mauricio?

»In-depth analysis of line stats

Yeah, had a problem with my flickr account. Will try to post them back eventually.

Concerning the RF chokes experiments, I had to call it off due to a busy schedule at school. I will get back to it eventually, unless my line gets upgraded... :D

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 22:23:07 EDT

Bicephale : Mauricio?

»In-depth analysis of line stats

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 22:19:39 EDT

Bicephale : No, i meant Mauricio as he now has the tools to measure the difference.

In your case the remote must have made your DSL signal stong enough
to consider using even more turns, especially if no effect is noticed... I
recall you wrote about noise on the "voice" part so here's at least one
section where i wouldn't hesitate to use a fair amount of RF Choking.

What's puzzling about your case isn't the presence of noise but the fact
that the UpStream band ("TxFecErr") gets hit by noise in such manner:

Just signed up for DSL. Hello fellow Teksavvy users!, Bicephale, 2009-Nov-3

Just signed up for DSL. Hello fellow Teksavvy users!, Bicephale, 2009-Nov-9

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 22:00:09 EDT

Oinktastic :

said by Bicephale :

Speaking of waiting how are your RF Choking experiments going?...

I didn't notice much of a difference between 0 and 5 turns. I'll see what happens if I leave 8 turns around the choke tonight.

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 21:51:19 EDT

Bicephale : Speaking of waiting how are your RF Choking experiments going?...

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 21:25:39 EDT

Mauricio :

said by Oinktastic :

Heck ya I was first in line.. no budding! :D

Boooooo :p

I guess I'll have to wait then.

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 21:08:46 EDT

Oinktastic : Heck ya I was first in line.. no budding! :D

Let's see a graph of those 5 errors. Should be quite riveting... :p

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 19:23:29 EDT

Bicephale : Hummm...

Quite frankly, i doubt even a five hours noise vacation
will make a significant difference in your specific case!

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 16:58:03 EDT

dslrocker3 :

said by Mauricio :

Hey dslrocker3, want to switch lines with me ? I am sure you would love to tweak my 1728/640 fastpath line... ;)

Sure, why not!? But I think that Oinktastic was in line first. Here I am worrying about 10 or so CRC's per hour but either of you will be glad to take that. LOL. Besides, you forget that my line was set to 1728/640 interleaved, 2496/800 fast path, 2496/800 interleaved, 4032/512 interleaved (never actually synced that high), and then 1728/640 fast path for a few days until they transfered my line and set me to 5056/512 interleaved. That 5056/512 profile last for about 5 minutes while I argued for an 800Kbps upload profile and now I'm on 6016/800 Kbps fast path. :)

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 16:48:49 EDT

Mauricio : Hey dslrocker3, want to switch lines with me ? I am sure you would love to tweak my 1728/640 fastpath line... ;)

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 16:34:25 EDT

dslrocker3 : bicephale, I think you know very well that I was hinting at the error rate.

I could just be a big coincidence but last night I took some of the phone wiring that comes into my room and bundled the excess into a bundle with an elastic. I know.... I know... it would probably be better to just cut it but have extra wire/slack gives me a little more flexibility where to place the phone jacks, phone devices, and modems. This isn't the jack that feeds my modem but it does run this wire does run alongside the wire that feeds the jack to my modem.

This morning I started noticing a crazily low error rate. I'm talking about 2 FEC's in a period of a 3 hours based on the fact that the DMT counter wasn't changing. So I started playing around a bit and thought that the phone wiring must have been transfering noise to my modem because I had just had the excess wire shoved underneath my desk beside some speaker and electrical wires.

A few hours after that, the error rates returned to normal so I started wondering that maybe I didn't find the source of noise after all. So now I'm really not sure what to think. Why would noise just suddenly decide to take a 5 hour vacation?

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 16:10:17 EDT

Bicephale : Give us a hint! What's the 1st letter of the word?

"V"??? As in "Victory"?... Or "Vacation"?!!

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 13:26:05 EDT

dslrocker3 : I think that I might be on to something....

Re: Help/Advice with my Line Statistics please

Mon, 09 Nov 2009 02:37:33 EDT

Bicephale : That looks like a 1st order low-pass filter but
the PC Board was designed to be populated
with optional components which would turn
it into a two poles filtering device... Perhaps
a hobbyist who owns two of those might be
tempted to combine them to get something
roughly equivalent to a POTS splitter but in
a wall-mount outlet, euh... Please don't ask
me if it's worth enough to bother though!...

Too bad these short cords aren't terminated
on both ends, in any case.

Re: Help/Advice with my Line Statistics please

Sun, 08 Nov 2009 23:23:00 EDT

Oinktastic : Here's the wall mount filter I was talking about.

Back of unit		Front
Inside
Filter Zoom (not much to see though)

Re: Help/Advice with my Line Statistics please

Sun, 08 Nov 2009 23:10:01 EDT

morphector : I know it's probably not what it's causing errors on your line.

You should remove the old protector with no fuse in it, it happened a few times to me that those old protector caused some noise, I even saw one that caused a drop of 3mbps even with no fuses and no grounding on it. I think that I still have that one in my truck as a keeper...

I don't know why but everytime I find something that has spectacular/funny troubles i'm keeping it and identify it... Those are like small trophey.

Re: Help/Advice with my Line Statistics please

Sat, 07 Nov 2009 13:21:31 EDT

Bicephale : If you provide clear close-ups of each side it
may tell us how many "poles" this filter has...

Re: Help/Advice with my Line Statistics please

Sat, 07 Nov 2009 12:48:40 EDT

Oinktastic : Hi eveyone,

Years ago when I was with AEI Internet, I was looking all over the place for those wall mount filters and I couldn't seem to find one. Finally I decided to try a Bell World store. Even though the only products they had visible were cellphones at the time, the guy I talked to went into the back and emerged with one of these about 15 seconds later. He handed it to me and said I could have it.

I use it for my old purely wired wall phone that works with or without electricity. I haven't opened it up to see if it's just a small microfilter with the plastic plate around it, or if it has its own electronics. I'll open it up when I get home and report back :)

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 19:37:20 EDT

wayjac : Its been years since I've seen a modem kit that included the the wall mount filter in my other post

Now days the wall mount filter is not included with the modem if one is needed it must be requested and it will be something like the one below

[att=1,l]

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 18:36:13 EDT

Bicephale : Hi WayJac,

What's nice with your post is that if someone asks for short
patch cords now we know one place where to look for them!

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 15:49:53 EDT

dslrocker3 :

said by Bicephale :

What do you plan to do with the "Z-Blocker" outlet???

was thinking of possibly using it as an indoor demarcation point but now that I think of it, it just looks like a splitter and filter built with a peice of plastic around it for wall mounting. I would have had to plug into an already existing jack which would have accomplished nothing. I think that I was possilby mistaken and was just getting my in/out directions mixed up and was thinking that one of the "outs" was an "in" which would have made it a good centralized test point.

I just look at my error rate for the past few hours and it averaged out to about the same as before. Oh well, at least the hiss is gone from the voice phone now.

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 15:26:12 EDT

wayjac : I have one that looks just like that, the front middle port is filtered, this filter is intended to be used with a wall mounted phone the odd thing about this wall phone filter is its just a dual port microfilter

[att=1,l]

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 14:50:21 EDT

Bicephale : What do you plan to do with the "Z-Blocker" outlet???

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 14:21:27 EDT

dslrocker3 : Even though it was only one particular phone that picked up "hiss" on the phone line, I can't help but think about if can hear "DSL hiss" on that line, I bet it's possible that any noise that is prsent along the way could travel back through the line that the single filter couldn't prevent from being fed back to the dsl modem. So far, only 2 upstream FEC's over 15 minutes which is definitely lower than my previous averages. Only time will tell if the modem cares about using 2 microfilters in the basement instead of 2.

I came across the the filter/splitter in my stash. I'm kind of curious if I should implement it's use. It looks like something that would probably be best served being used as a demarcation point. I guess I could term that middle jack as a "line out jack" of sorts which I could use to plug in that extension cord that helps my connection. Too bad I'm not sure if it maters where I plug that extension cord in. Also, I'm not entirely sure if that jack in the middle of this wall mountable block is filtered or not (I would want it to be NOT filtered for my unique purposes).

Re: Help/Advice with my Line Statistics please

Fri, 06 Nov 2009 13:58:40 EDT

dslrocker3 : Today's little small contrilbuation/discussion about my line...

On previous occassions I had tried double filtering and stated that I found that double filtering only seemed to make a louder noise on my voice line. What just struck me was that I believe that I had not yet removed the ground wire in the basement that had no carbon attached to the other end. So perhaps the louder humming sound on the voice line with 2 filters was only caused by the incorrect ground

For absolutely no reason today, I decided to play with the amount of filters int he basement that filter out the voice line from the modem's wiring. With 1 filter, I here a very small amount of hiss in one particular phone in my house. With 2 filters, the hiss is completely gone and only a very slight "electronic noise" sound in the background that you only hear if you are listenning very, very closely. So I went ahead and tried triple filtering.

Attaching 3 filters actually make the voice line more noisy (still no hiss but more of an electronic noise sound came through). Hopefully, people understand when I say an "electronic noise sound" (a very faint background noise that almost just sounds like electricity - if you could actually hear it which you can't really).

I then tried playing with swapping filters around just to make sure there wasn't something fishy going on with any one of them. It didn't matter which ones I used in the end and the results were the same.

one filter=slight hiss on one on my voice phones, slight electronic background nioise
two filter=zero hiss, pratically zero electronic background noise
3 filters=still zero hiss but electonic background noise amount returned to about the same level as only using one filter

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 23:47:32 EDT

Bicephale : We'll do our best to accomodate you, promissed!

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 22:55:05 EDT

Oinktastic :

said by dslrocker3 :

said by Bicephale :

Poor Oinktastic, i'm afraid he's already overwhelmed as it is!...

yup, my increase in noise margin from 25 to 26.5dB might just be enough to push him over the edge and cause him to go ballistic!!!! LOL

I won't go ballistic :p However, it does just look a little bit too tantalizing

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 17:51:20 EDT

wayjac :

said by dslrocker3 :

I'm never seen higher than 10938 Kbps downstream on this line. However, I'm ceratin that my line is capable of more than 20Mbps on ADSL2+.

If your current max sync rate is short of 12000kbps now
I don't think it would be higher than 20000kbps using adsl2+

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 17:39:25 EDT

dslrocker3 : I'm never seen higher than 10938 Kbps downstream on this line. However, I'm ceratin that my line is capable of more than 20Mbps on ADSL2+. We can't really go by the Kbps numbers that DMT 7.35 provides because the Speedtouch 516 doesn't actually record an attainable rate result. It only keeps track of RCO and DMT 7.35 then calculates an approximation meaing that there is rounding off involved. I know that the attainable rates are actually calculated by the computer software because DMT 7.35 and OrbMt both shows 1025Kbps for the upstream but STMT shows 1026Kbps.

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 16:54:19 EDT

wayjac : What's the top max sync rate so far

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 15:43:45 EDT

Bicephale : Euh... Oinktastic going ballistic?!...

My brain will need a fair amount of sugar just to picture that!!!

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 15:15:00 EDT

dslrocker3 :

said by Bicephale :

Poor Oinktastic, i'm afraid he's already overwhelmed as it is!...

yup, my increase in noise margin from 25 to 26.5dB might just be enough to push him over the edge and cause him to go ballistic!!!! LOL

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 14:48:59 EDT

Bicephale : Poor Oinktastic, i'm afraid he's already overwhelmed as it is!...

Re: Help/Advice with my Line Statistics please

Tue, 03 Nov 2009 14:39:25 EDT

dslrocker3 : an all time high signal to noise ratio margin at 6016Kbps for me. :) It went up by between 1 to 1.5dB and I didn't have to do anything for it!

Oinktastic, ready to throw something at me yet?

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 23:36:14 EDT

Mauricio :

said by Bicephale :

Hi Mauricio,

Would you mind if i asked you about documenting how you
made 'RouterStats' work with your 2Wire by showing us the
step-by-step procedure illustrated with a few images taken
by Windows (PrntScrn, Cut'n'Paste into 'MSPaint') or else?

I think this DSL customer here could use a hand, it seems:

»New teksavvy customer slow download speed???

This thread has already reached its 4th page and yet it's
begining to feel like a bad dream to me - in slow motion!...

Hahah ! I will try to put something together tomorrow. ;)

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 22:49:41 EDT

Bicephale : It seems we've been throught this before, DSLRocker:

Help/Advice with my Line Statistics please, Bicephale, 2009-Jul-8

Help/Advice with my Line Statistics please, Bicephale, 2009-Jul-7

Right here in the present thread... I guess tomorow
i won't even remember where i've put my shoes!...

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 22:28:23 EDT

Bicephale :
I think i found it but i can't confirm positively by myself.

Mauricio - 2Wire XSLT Page B04 3rd Column (2009-Oct-19)

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 21:56:47 EDT

Bicephale : This is what i found:

In-depth analysis of line stats, Mauricio, 2009-Oct-19

It should work but i'm still searching for clues...

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 21:37:24 EDT

dslrocker3 :

said by Bicephale :

Hi DSLRocker,

For me the 2Wire "J45" page is only a text file about 80 Kilo-Bytes
in size so i'm not sure how 'RouterStats' is supposed to parse that
as i only have samples: i don't own a 2Wire myself!

In your opinion, can noise curves be collected at all, nonetheless?

Yes, noise curves can be collected. I think there's 2 ways to do it. From what I remember:

you can use the "detailed statistics" page in the main interface. From there, you could just use the total amount in each of corrected and uncorrected errors and tell Routerstats to graph the change since the last interval.

I believe that J45 can be used by always using the last possible value in each category. Since the last value is the most recent, you shoudl be able to highlight that and tell Routersats to use that number. The 2wire modem moves the values in the J45 up one notch after each interval (5 minutes for some item, 15 minutes for others I think).

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 21:25:07 EDT

Bicephale : Hi DSLRocker,

For me the 2Wire "J45" page is only a text file about 80 Kilo-Bytes
in size so i'm not sure how 'RouterStats' is supposed to parse that
as i only have samples: i don't own a 2Wire myself!

In your opinion, can noise curves be collected at all, nonetheless?

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 21:10:47 EDT

dslrocker3 : The important thing that I ntoiced with the version 4.25 2wire firmware was to make sure that I used the numbers provided in the main interface. Any attempt to use any data from the advanced MDC pages resulted in failure (couldn't get past the log in). The numbers from the main menus and the j45 page should work with routerstats.

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 20:56:21 EDT

Bicephale : Hi Mauricio,

Would you mind if i asked you about documenting how you
made 'RouterStats' work with your 2Wire by showing us the
step-by-step procedure illustrated with a few images taken
by Windows (PrntScrn, Cut'n'Paste into 'MSPaint') or else?

I think this DSL customer here could use a hand, it seems:

»New teksavvy customer slow download speed???

This thread has already reached its 4th page and yet it's
begining to feel like a bad dream to me - in slow motion!...

Re: Help/Advice with my Line Statistics please

Sun, 01 Nov 2009 15:16:44 EDT

Mauricio :

said by wayjac :

I don't think the modem actually indicates a bridge tap

»SBC DSL FAQ »What do the hybrid numbers mean and what do they tell you about your line?

I did not know that. Kudos !

If only I knew where my SpeedStream 5242 is...

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 19:46:37 EDT

Bicephale : WayJac already provided related information in the present thread, actually:

»Re: Help/Advice with my Line Statistics please

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 18:00:53 EDT

STN : Sweet. Thanks wayjac

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 17:19:38 EDT

dslrocker3 : At first, I was saying for people to plug in random extension cords just as a joke and bicephale actually tries it! it must be hallowe'en or something...

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 17:12:22 EDT

wayjac : I don't think the modem actually indicates a bridge tap

»SBC DSL FAQ »What do the hybrid numbers mean and what do they tell you about your line?

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 17:03:59 EDT

STN : Ok, just so I'm clear. Is the SS actually telling you specifically that there is a BT. Or are you interpreting things in the log and coming to that conclusion.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 17:01:41 EDT

Bicephale : Inconclusive preliminaries: my SpeedStream 4200 still uses
"Hybrid 2" and the bins didn't behave the same as with my
SpeedTouch. The connection/disconnection of the "Bridge
Tap" failed to cause the SS to renegociate, actually...

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 16:44:28 EDT

Bicephale : You'll find some "Hybrid" examples here, STN:

»/nsearch?q=Spe···m+Hybrid

:)

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 16:36:28 EDT

Bicephale : Every day you are proving that it is possible to attenuate the
effects of an outdoors "Bridge Tap" causing echo cancellation
issues, DSLRocker!

Perhaps i should put my SS4200 on-line to see if the "Hybrid"
will change...

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 16:15:53 EDT

dslrocker3 : so what are we theorizing here? If you have a bridge tap, plug in cheap, long, untwisted, dollar store extension cords and your signal to noise ratio margin and attainable rates might go up? that's scary and maybe i'm not crazy after all for trying it to begin with!

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 16:15:01 EDT

STN :

said by Bicephale :

This tends to confirm what my SpeedStream told me long ago,
though: i do have a "Bridge Tap" outdoors, effectively.

How did your SS tell you that

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 16:13:37 EDT

Bicephale : Seriously, i get one or two more bins but that's not conclusive
for my taste. I'd need to repeat the experiment a dozen times
to be convinced it's not just a pure coincidence.

More bins is good news but two more bins worth only a couple
bits doesn't translate into a significant SNR Margin difference...

This tends to confirm what my SpeedStream told me long ago,
though: i do have a "Bridge Tap" outdoors, effectively.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 16:07:11 EDT

Mauricio :

said by dslrocker3 :

said by Bicephale :

Since we've entered the domain of crazyness/scaryness i
gathered all the extension cords i could use and deployed
that antenna across the whole appartment. If i connect/
disconnect that thing to/from the LINE outlet of my POTS
Splitter this forces my SpeedTouch to renegociate and it
seems to gain bins with the parasitic item on-line!...

Bicephale, are you joking or are you really being serious? More bins is a good thing!

If it is true, I am curious to see if it's across the whole spectrum.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 15:50:41 EDT

dslrocker3 :

Bicephale, are you joking or are you really being serious? More bins is a good thing!

And when you say "line", do you mean the side that is meant for the dsl modem (unfiltered) or the side that is meant for voice phones (filtered)?

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 15:29:14 EDT

Bicephale : Since we've entered the domain of crazyness/scaryness i
gathered all the extension cords i could use and deployed
that antenna across the whole appartment. If i connect/
disconnect that thing to/from the LINE outlet of my POTS
Splitter this forces my SpeedTouch to renegociate and it
seems to gain bins with the parasitic item on-line!...

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 15:29:04 EDT

Mauricio :

said by dslrocker3 :

Mauricio, keep in mind that my modem isn't actually plugged into the extension cord. it's just necessary for me to have the extension cord plugged in somewhere along the house's wiring.

Oh, I see. Should be fun to experiment nonetheless.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 15:26:29 EDT

anon :

said by Groaner :

that was a real groaner :p
-1 for u

This is what years of working for Bell does to you.
You lose your sense of humour and become a dullard. :(
Have you got something better?
I'm sure there will be a Bell reference in there somewhere. ;)
Better yet start a scary Bell thread for Halloween.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 15:17:04 EDT

dslrocker3 :

said by Bicephale :

On the horror scale, from least to most:

DSLRocker's
Mine
Oinktastic's

;)

...except that my setup makes no sense. I think we need to start suggesting to people with line problems to just try whatever comes to mind even it shouldn't help because my line shows that crazy/scary wiring can improve your line stats.

Mauricio, keep in mind that my modem isn't actually plugged into the extension cord. it's just necessary for me to have the extension cord plugged in somewhere along the house's wiring.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 15:10:17 EDT

Mauricio :

said by dslrocker3 :

just out of curiosity, I mesed around with that extension cord today.

signal to noise ratio with no extension cord = 21.5 dB
signal to noise ratio with the usual extenion cord = 25.5dB
signal to noise ratio with a different extension cord = 23.0dB

Now, I'm not going to suggest that people start plugging extension cords into phone jacks to see if that improves DSL signal but there has to be something to this. I'm sure that all of the textbooks would say that my wiring is just plain wrong but hey, it works for me.

I should try using a 50' flat cord in my setup instead of Cat5... :D ;)

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 14:55:17 EDT

Bicephale : On the horror scale, from least to most:

DSLRocker's
Mine
Oinktastic's

;)

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 14:54:03 EDT

anon :

said by xbell :

Being Halloween and all I guess the scariest thing in your house is your line. :)

that was a real groaner :p
-1 for u

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 14:36:55 EDT

anon :

said by dslrocker3 :

I beleive that the fact that cat5 cable is generally not meant for outdoor use could be a concern, unless I purchase some expensive outdoor graded wire.

Home depot sells outdoor rated cat. 5 by the meter and it is pretty cheap.
I have some super heavy duty cat. 5 on the truck that is designed for feeds and to be bricked in but that is overkill for your app.
Being Halloween and all I guess the scariest thing in your house is your line. :)

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 13:43:13 EDT

dslrocker3 : I kind of miss having an unpredictable line. In reality, no matter what else I do to my current line, there will not be any signficant improvement.

Xbell, I know you're right about the cat5 cable as it is twisted pair as that will curb excess noise but I beleive that the fact that cat5 cable is generally not meant for outdoor use could be a concern, unless I purchase some expensive outdoor graded wire. I don't have an excess noise problem so I fear doing that change will just make me lose an outdoor graded cable (even it is old).

just out of curiosity, I mesed around with that extension cord today.
signal to noise ratio with no extension cord = 21.5 dB
signal to noise ratio with the usual extenion cord = 25.5dB
signal to noise ratio with a different extension cord = 23.0dB
Now, I'm not going to suggest that people start plugging extension cords into phone jacks to see if that improves DSL signal but there has to be something to this. I'm sure that all of the textbooks would say that my wiring is just plain wrong but hey, it works for me.

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 13:37:44 EDT

Bicephale : Bored? You probably mean desperate for action!

:D

Re: Help/Advice with my Line Statistics please

Sat, 31 Oct 2009 11:29:38 EDT

anon :

said by dslrocker3 :

I highly doubt it's twisted pair. The wiring coming from my side of the NID doesn't even have 2 pairs in the insulation. It is just old drop wire that simply has 2 thick, non-color coded, solid peices of copper inside of it (the type that you actually need a tool like a pair of plyers to be able to even bend).

It's an old drop wire pre- 15 years ago.
Basically it has all the same characteristics that a line cord would from the parallel conductors except the wire gauge is heavier.
If you are bored and you have access to it pull it out of the hole into the basement and a cat. 5 will fit in the same hole.

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 20:41:40 EDT

Bicephale : M'yeah, such a display of misery breaks my heart!

:p

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 20:05:21 EDT

wayjac :

said by dslrocker3 :

I'll probably only get a 0.5dB reduction in attenuation. oh my, what BIG problems I have with my line now!

He's back in the saddle again ;)

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 19:11:49 EDT

dslrocker3 :

said by Bicephale :

But you can still try to optimize length
as a function of the SNR margin, if you
need something to keep you busy!...

I do know that the extension cord clearly helps my connection. I also know that once the extension cord is connected, any of the houses jacks benefit from the increased DSL line quality.

What I have not been able to determine is if that cable needs to be connected on that exact jack or not. I wouldn't mind figuring that out as that would give me more options if it could be plugged in anywhere. The ability to move that extension cord would also allow me to have my modem's phone wiring take a more direct path to the basement while maintaining a minimal amount of unfiltered wiring. I don't actually want to have my modem fed by wiring that is held together by electrical tape in the basement and then goes through 2 different jacks before getting to my modem. That was only done to minimize the the house's dsl loop length. GRRR!!! someone stapled that wire to the wall! someone doesn't want me to move it! maybe, it's a sign from bicephale's crystal ball to just leave it where it is! LOL.

Even if I found a shorter extension cord that had the same type of effect and reduced the amount of wiring in my house, I'll probably only get a 0.5dB reduction in attenuation. oh my, what BIG problems I have with my line now!

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 18:52:44 EDT

dslrocker3 :

said by xbell :

Sounds good. Is it twisted pair?

I highly doubt it's twisted pair. The wiring coming from my side of the NID doesn't even have 2 pairs in the insulation. It is just old drop wire that simply has 2 thick, non-color coded, solid peices of copper inside of it (the type that you actually need a tool like a pair of plyers to be able to even bend).

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 18:34:30 EDT

Bicephale : Exactly, i don't think noise is a concern.

But you can still try to optimize length
as a function of the SNR margin, if you
need something to keep you busy!...

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 18:28:59 EDT

anon : Sounds good. Is it twisted pair?

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 18:17:42 EDT

dslrocker3 : It looks like my scary looking diagrams scared any line noise away. Either that, or the errors decided to go on vacation this afternoon! I might just be wasting my time trying to further optimize things.

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 17:55:54 EDT

dslrocker3 :

said by xbell :

Keep it simple stupid. That saying is the best inference for this. Less is better. I try to keep no more than one connection (scotchlocked if possible) between fuses and modem.

You have multiple protectors labeled btw. Do you mean splice points.

The 5 feet of wire coming in from the NID has zero other wires coming with it, other than the grounding wire.

They are 2 old former protectors that were once hooked up to carbons and a different ground wire than is being used now. I just used them as a convenient way to split the signal 2 ways. »/speak/slidesh···0zMjA%3D

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 17:51:05 EDT

anon :

said by dslrocker3 :

I'm not too overly concerned about going directly to the NID as the wire that goes from the NID into the basement is only about 5 feet long.

Of any section of wire in the loop that is often the most corrupted due often to proximity to electrical service panels in most homes.
Some people have the old nid in a jack at the panel and we have tested there and at the protector and that 5 feet made a big difference although short as it is.

Would running the wiring that goes to the modem through another jack/another step pose a concern/one more thing that could go wrong along the way? just wondering....

Keep it simple stupid. That saying is the best inference for this. Less is better. I try to keep no more than one connection (scotchlocked if possible) between fuses and modem.

You have multiple protectors labeled btw. Do you mean splice points.

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 17:45:11 EDT

dslrocker3 :

said by wayjac :

Can you post again the effect on the line stats the extension cord has

The first one is with the inside bridge tap/extension cord connected. The second screenshot represents the connection when the extension cord was unplugged.

»Re: Help/Advice with my Line Statistics please

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 17:36:22 EDT

dslrocker3 :

said by xbell :

said by dslrocker3 :

I'm trying to find an opinion on which setup would be preferable out of the first 2 pictures.

Number 2 but I would go Nid, cat. 5 white/blue, modem. Nid, cat. 5 white/orange, low pass filter, phones.

I'm not too overly concerned about going directly to the NID as the wire that goes from the NID into the basement is only about 5 feet long. Can you elaborate on the technical reasons for changing the wiring to the setup in the second picture instead of keeping it the same as in the first picture? Would running the wiring that goes to the modem through another jack/another step pose a concern/one more thing that could go wrong along the way? just wondering....

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 17:08:58 EDT

wayjac :

said by dslrocker3 :

I'm trying to find an opinion on which setup would be preferable out of the first 2 pictures.

I think you should use the setup that has the most desirable line stats
Can you post again the effect on the line stats the extension cord has

said by xbell :

Of course your line is cursed so I don't know what to say anymore.

Its not cursed anymore......now its just special

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 16:23:06 EDT

anon :

said by dslrocker3 :

I'm trying to find an opinion on which setup would be preferable out of the first 2 pictures.

Number 2 but I would go Nid, cat. 5 white/blue, modem. Nid, cat. 5 white/orange, low pass filter, phones.
As I run wire all day that is the easiest for me. Of course your line is cursed so I don't know what to say anymore.

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 15:26:11 EDT

Bicephale : Hummm...

Perhaps this simplified version will help you to make a choice.

If i've understood the situation correctly your extension cord
which i tagged "Bridge Tap" in my drawing won't be displaced
significantly. As for N1/N2 where "N" stands for noise source,
your DSL signal is so strong this may not be a real issue most
of the time. As for the rest, the filtered "voice" section must
be invisible at DSL frequencies, essentially speaking, while we
can safely ignore the EtherNet cabling section completely!...

Considering your need for that unterminated extension cord i
suggest you focus on that part of the wiring instead... For all
we know the odd behaviour you observed previously gives us
a reason to think that your indoors "Bridge Tap" compensates
for the impedance mismatch due to another located outdoors.

If i were you i'd investigate in hope to optimize the length. It
seems to me there has to be a sweet spot so in order to find
it i'd use a very long extension cord which will be meant to be
sacrificed. My test would proceed as follows: i'd collect some
detailed statistics for a couple hours, reduce the length by a
fixed amount, collect more data again and then repeat until i
have nothing left to cut... Analysis of the collected statistics
may or may not reveal a sweet spot. Good luck and have fun!

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 13:18:54 EDT

dslrocker3 : I'm trying to find an opinion on which setup would be preferable out of the first 2 pictures.

The current setup allows for the dsl signal to have to pass through less things to get to the modem. However, the proposed revised set up would allow a better isolation and truer removal of all other wiring (by a difference of 1 wire) if I wanted to test my modem in the basement witout making any alterations other than unplugging an rj11 connection.

I also know that implementing the change and then moving the modem to the basement would be even a better idea. That would also allow the removal of 50 feet of wiring and a jack that would not longer be used. I would have to use a splitter on the DSL side of the basement jack to allow that extension cord to remain visible to the modem.

The first proposed change is very easy to implement in about 2 minutes. the second proposed change is more of pain in the ass than anything else with very little to gain considering my current line state. So let's assume for now that wiring change proposal #2 is not an option (even if it is probably the best).

current setup
proposed wiring change
propsed wiring change #2

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 12:17:03 EDT

Bicephale : Input overload, central cognitive processor failure!
Brain rebooting, absorption of nutriments required!!

Euh...

Can you put this in visual schematic form please?...

Re: Help/Advice with my Line Statistics please

Fri, 30 Oct 2009 10:32:25 EDT

dslrocker3 : ... just a random picture of Bell equipment in my area. I'm not sure what is really in it. I pretty sure it's not an RDSLAM because I walked by a Bell technician at that site before and asked him when remotes would be installed in my area and he pointed to the tall grey box about 50 feet away.

just a wiring question for my line...
I know that the absolute best way is to usually have the modem's wiring going to the Bell side of the NID and then a filter on the customer side but as I've said before, that doesn't provide the best connection in my case due to a probable impendence mismatch. I have the wire that comes inside going to a junction in a basement and my modem is connected through wiring that is connected to that same junction point. From that junction point I have a wire going to a double RJ11 jack that has a filter plugged into one of those jacks with a line cord going to a second junction point to feed the rest of the house/the voice line.

Is it better to leave it like that or is better that the wire coming into the house is connected to that double jack and have one side of that double jack feed the junction that the modem wiring is hooked up to and the other side feed the second junction for the voice?

Bell brown cabinet box

Re: Help/Advice with my Line Statistics please

Mon, 26 Oct 2009 10:41:42 EDT

Bicephale : If they could just stay away from electrical power conduits!...

Re: Help/Advice with my Line Statistics please

Mon, 26 Oct 2009 09:50:55 EDT

Oinktastic :

said by xbell :

It's called cutting ahead and Bell frowns on it for the techs...

That's a shame that they discourage it, since it would keep a lot of their customers happier with their connections. My line could definitely use some TLC, but Bell's so far refusing to even admit there's a problem.

My family has been using Bell home phone for decades and we will probably continue to do so as long as the price doesn't go up. After all, we're using only the basic service with no features and it's not even really worth what we're currently paying. It's been very stable and of decent quality for the entire time we've had it, so we're happy to keep using it for consistency and loyalty's sake.

However, their increasing unwillingness to help customers, coupled with their constantly rising prices for less and less value are forcing people to go all digital and ditch all their Bhell products. If they continue to ruin my DSL experience, I may have to follow the pack and migrate elsewhere.

Re: Help/Advice with my Line Statistics please

Mon, 26 Oct 2009 09:44:20 EDT

Bicephale : I'd say you're so close it renders the antenna effect unsignificant!

Re: Help/Advice with my Line Statistics please

Mon, 26 Oct 2009 01:34:14 EDT

anon :

said by vintagewino :

Seems the easiest way would be to cut the excess that's going past your place so that it no longer exists on your line.

It's called cutting ahead and Bell frowns on it for the techs. I have been known to do it though.

Re: Help/Advice with my Line Statistics please

Mon, 26 Oct 2009 00:24:53 EDT

dslrocker3 : I have completed my little experiment. I have come to the conclusion that the extension cord does not attract any extra noise. I can see no signficant difference in error rates between the extension cord being plugged in and not and that extension cord only appears to help my DSL connection. The DMT screenshot provides a quick snapshot of line conditions after I returned everything to my regular set up.

I know a change in attenuation of 0.5dB doesn't mean a lot but my upstream attenuation has dropped to a slightly lower number. 4.0dB? I must be pretty close to that Ikanos remote!

DMTdiag2009O···.csv.zip 10,791 bytes
experiment complete, extension cord reconnected

Re: Help/Advice with my Line Statistics please

Sun, 25 Oct 2009 08:52:43 EDT

Bicephale : I can't wait to see the numbers flow, DSLRocker!!!

Re: Help/Advice with my Line Statistics please

Sun, 25 Oct 2009 07:44:24 EDT

Bicephale : Lets put things side by side for clarity:

Less DS SNR Margin = more DS Power...

Re: Help/Advice with my Line Statistics please

Sun, 25 Oct 2009 00:32:01 EDT

vintagewino :

said by Oinktastic :

I wonder how well Bell can test for those gremlins.

Agreed, nice find!

Generally, the nasty part would be once the F2 wire that goes to your place is not terminated to your drop, but is allowed to go past your place to ...???

As the picture shows, this can create standing waves (reflections) that would be a miserable pig to try to find, much less try to "tune out".

Seems the easiest way would be to cut the excess that's going past your place so that it no longer exists on your line.

Hmmm ... this gives me an idea!!!

Re: Help/Advice with my Line Statistics please

Sat, 24 Oct 2009 23:50:58 EDT

dslrocker3 : If we all remember, my line is the one that gets better results when plugging in a speicific unfiltered extension cord anywhere on the house's loop, as long as there's no device attached to it and as long as the modem isn't attached to it.

I have now unplugged that extension cord. The signal to noise ratio is down by about 3dB and the attainable rate is down by 700 to 900Kbps (approx).

I want to see if that extension cord that helps boost the signal to noise ratio and downstream attainable rates is also a source of noise.

I have DMT 7.35 running in diagnosis mode and plan on leaving that running for at least 24 hours. I am not encountering any problems on my line and this is all just for curiosity. I don't like the way my bits per tone chart looks with that extension cord removed but we'll see what happens when it comes to CRC and FEC line errors.

10/24/2009 23:40EDT, extension cord removed

Re: Help/Advice with my Line Statistics please

Tue, 20 Oct 2009 23:36:48 EDT

Bicephale : Yeah, i concur entirely: the 'DMT'-hostile v7
FirmWares are certainly a bugger for curious
DSL customers, there's no doubt about that!

Re: Help/Advice with my Line Statistics please

Tue, 20 Oct 2009 23:32:52 EDT

dslrocker3 : you are abosolutely correct about standing-still-in-time snapshots. I didn't have anything running to compare stats from one moment to the next. What I can tell you that the average has not altered sigficantly each time I'm checked since I last reset the modem. But that's about all I can tell you about it obviously.

Since I have no real problems with my line, it's hard to get motivation to actually "mess around" with the wiring when there's very little to gain. Any experiments will be pretty much just for knowledge and curiosity.

When it comes to the firmware, I'm reluctant to try newer version for reasons that you are already aware of. I don't see myself doing much tweaking as my line doesn't need it but I also don't want to lose the RCO and attainable rate reporting that I believe is lost in version 7 firmwares.

Re: Help/Advice with my Line Statistics please

Tue, 20 Oct 2009 23:24:14 EDT

Bicephale : Hi DSLRocker,

Two comments come to my mind. 1st, i'm a bit
reluctant to depend on instant snapshots now
that i've discovered the real power of long-term
records when processed through a SpreadSheet
to create daily "noise" curves... 'DMT' captures
are precious at the preliminary stage but can be
misleading when doing serious trouble-shooting.

In the present case the detail which strikes me
is about FirmWare versions. I'm observing very
different behaviours possibly due to differences
in FirmWares... For example, v6.1.0.5 gives me
a sawtooth-shaped SNR Margin curve while i'm
having a relatively flat curve with v7.4.3.2 (but
it may also be a consequence of the phone-line
quality which varies - or a combination of these
factors). In any case, 'DMT' instant snapshots
fail to reveal such differences.

My second comment will be about A.M. stations
being picked up by the DSL wiring. If you think
this is what's happening then RF Choking makes
even more sense to me now!

Re: Help/Advice with my Line Statistics please

Tue, 20 Oct 2009 18:24:15 EDT

dslrocker3 : I've also notice a recent trend over the past few nights when the downstream signal to noise ratio has been consistently lower during the night. During the night, the downstream signal to noise ratio flucuates between 24.5 and 25.0dB. During the day, it rises to 25.5 to 26.0dB. Also, one or more the tone's signal to noise ratios has been dropping to 14/15dB at night.

Today, I'm noticing a lower than normal error rate. The upstream error rate is about 1/2 to 1/4 of what is normally is. What strikes me is the tone gaps that showed up the last time that I reset my modem. Coincdently, there are local Toronto AM radio stations at 1010 and 1050 KHz. I wonder if tones 234, 243, and 244 are noiser on my line that the rest and if the fact thjat the my modem has disabled them for now has resulted in the lower error rate. Unfortunately, I no longer have a modem that is capable of the manual disabling of speicific tones.

Re: Help/Advice with my Line Statistics please

Mon, 19 Oct 2009 15:49:43 EDT

Bicephale : Follow the "Tomi Engdahl's site" link, magic word:

"TDR" (Time Domain Reflectometer)

It's amazing what the graphic signatures can be
saying (see at the botomm of that web page!)...

Re: Help/Advice with my Line Statistics please

Mon, 19 Oct 2009 15:24:57 EDT

Oinktastic :

said by Bicephale :

I just found this page explaining "Bridge Taps", it's a must read:

»adslm.dohrenburg.net/troubleshoo···taps.php

;)

NIce find :-)

I wonder how well Bell can test for those gremlins.

Re: Help/Advice with my Line Statistics please

Mon, 19 Oct 2009 13:32:53 EDT

Bicephale : I just found this page explaining "Bridge Taps", it's a must read:

»adslm.dohrenburg.net/troubleshoo···taps.php

;)

Re: Help/Advice with my Line Statistics please

Sun, 18 Oct 2009 15:51:52 EDT

Oinktastic :

said by dslrocker3 :

decided to run to some dslreports tests....

»/tweakr/block:···terpppoe
»/linequality/nil/2576933

Very little jitter. That's nice to see.

What's with 10% re-transmitted packets though? Is that a problem?

Re: Help/Advice with my Line Statistics please

Sun, 18 Oct 2009 04:59:42 EDT

dslrocker3 : decided to run to some dslreports tests....

»/tweakr/block:···terpppoe
»/linequality/nil/2576933

Re: tweaking

Fri, 16 Oct 2009 10:42:33 EDT

vintagewino :

said by Oinktastic :

Lol thanks for the information. I guess I have another veteran to consult now :D

Plug your modem into one of these:
»www.canadiantire.ca/AST/browse/3···imer.jsp

6 programmable "outages" with a randomize feature.

They will check the # of outages over the last few days, so just make certain it doesn't go off at exactly the same time several days running! :D :) :D Just a little too suspicious!

Re: tweaking

Fri, 16 Oct 2009 00:18:44 EDT

Oinktastic :

said by dslrocker3 :

another trick: turn your modem off. this one is harder to accomplish because they often review your line conditions before phoning you so you don't always know when they are going to be sniffing around. when they ask you about it, just act as if the connection isn't working. Turn the modem back on after a technician arrives.

Lol thanks for the information. I guess I have another veteran to consult now :D

Re: tweaking

Fri, 16 Oct 2009 00:16:17 EDT

dslrocker3 : another trick: turn your modem off. this one is harder to accomplish because they often review your line conditions before phoning you so you don't always know when they are going to be sniffing around. when they ask you about it, just act as if the connection isn't working. Turn the modem back on after a technician arrives so that they can't claim that you have a bad modem (tech won't be able to prove anything about you turning your modem off because he will check at the demarc at a different time. just act as if it must have been an intermitent problem. remember to turn your modem back on before he starts playing around at your NID!).

Re: tweaking

Fri, 16 Oct 2009 00:13:20 EDT

Oinktastic : I'll keep that in my backup bag of dirty tricks :p

I had someone from Bell call me up one time and say that they had spotted a short on the line between Bell's equipment and my house. Unfortunately, nothing ever became of that. The tech who subsequently showed up just tested with his equipment and claimed everything was dandy and wouldn't listen to anything I had to say.

I should have asked him "if everything is so great, why am I on an interleaved profile when my line should easily be able to handle this sync profile"... I'll put that in my regular bag of dirty tricks :)

Edit: I don't want to be stealing your thread, so I'll try not to ramble too much. As you can tell by now, keeping it short isn't my strong point.

Re: tweaking

Fri, 16 Oct 2009 00:06:07 EDT

dslrocker3 : With all seriousness, if Bell is reluctant to send a technician, literally hook up an extra wire to your junction box that you don't care about damaging and wait until you are speaking with the Bell High Speed Centre as they will being looking at the computer screen that shows information about your line. If you take a pair of metal scissors and cut that wire in half, that will register as a short circuit and will likely get them to send someone. Remove the wiring that you deliberately damaged. As you have discarded that wire, they won't be able to blame you for it because the wiring that caused the short circuit won't be connected any more. Playing these games with Bell won't actually fix anything but at least you will have a fighting chance and get someone to come to your premises.

You would only of course do something like this if you are 100% certain your wiring isn't the issue but this should work to get a techncian to at least look at things. However, as always, more trouble tickets increases the chance that you will be 64/64ed, or 1184/640ed, or otherwise have your profile setting slowed down.

Re: tweaking

Thu, 15 Oct 2009 23:53:51 EDT

Oinktastic : I'm thinking I should probably take notes! ;)

Re: tweaking

Thu, 15 Oct 2009 23:47:08 EDT

dslrocker3 :

said by Oinktastic :

So dslrocker, what are you going to do with your connection now that you don't have to fight with it?

purposely sabotage it (plug modem into dsl microfilter) because I miss the Bell techs so much!!!! How much would someone have to piss Bell off before they would 64/64 someone? I also miss not having defective modems.

or maybe even open a trouble ticket up and use a screwdriver to tighten the screws while they are testing the connection... or connect a bunch of extra wires to my NID and get a pair of scissors to cut them one by one while I am on the phone with the Bell techicians at the High Speed Centre/XDSL. It's fun hearing the phone ring when you cut a wire like that!!!

Re: tweaking

Thu, 15 Oct 2009 23:44:47 EDT

Oinktastic :

said by Bicephale :

Oh, i sense we may have two loyal Bell customers here!

Must be by choice, most certainly...

Most certainly ;)

I'm really glad for dslrocker that he was finally able to get things straightened out. I'd like to think I'm sympathetic, but it also gives me hope for my own situation :P

So dslrocker, what are you going to do with your connection now that you don't have to fight with it?

Re: tweaking

Thu, 15 Oct 2009 23:39:57 EDT

Bicephale : Oh, i sense we may have two loyal Bell customers here!

Must be by choice, most certainly...

Re: tweaking

Thu, 15 Oct 2009 23:11:15 EDT

Oinktastic : Well, I certainly hope it doesn't take them 10 years to fix mine, but I'm hoping for the same final outcome at least lol.

Re: tweaking

Thu, 15 Oct 2009 23:06:41 EDT

dslrocker3 :

said by Oinktastic :

said by dslrocker3 :

I agree with you. This tweaking doesn't accomplish anything. I just did it because i was bored! :p

I'll make a deal with you. I'll bring my line over to your place, and we can swap. Howzabout that? :)

I think I'm allowed to be silly and brag about my line every now and then. I've been having 10+ years of line problems including: 911 being called because of static/shorts on the line, no dial tone issues at least 3 or 4 times requiring a technician to fix it, fluctuating dsl line conditions, bridge taps, and years of Bell not showing up for appointments or just claiming I have defective modems. My line competely missed the 3Mbps profile upgrades and the 5Mbps was delayed by more than 5+ years. Over those years, my line was at 1728/384 or 1728/640 for most of that time with atteunuation numbrers above 60dB. This latest round with Bell took me 6 months to resolve before they probably decided that transfering me to the rdslam was easier than any other repairs that they could do.

Re: tweaking

Thu, 15 Oct 2009 23:01:14 EDT

Oinktastic :

said by dslrocker3 :

I agree with you. This tweaking doesn't accomplish anything. I just did it because i was bored! :p

I'll make a deal with you. I'll bring my line over to your place, and we can swap. Howzabout that? :)

Re: tweaking

Thu, 15 Oct 2009 22:00:53 EDT

dslrocker3 :

said by Bicephale :

I don't know what you mean by "rejection rate" but i see
no benefit by tweaking so far: your RCO went up and a
few bins were lost, it's probable no tweak can help here!

:p

When I say "rejection rate", I am refering to the peices of signal that the modem did not use because they didn't meet a certain quality standard.

I agree with you. This tweaking doesn't accomplish anything. I just did it because i was bored! :p

Re: tweaking

Thu, 15 Oct 2009 21:48:22 EDT

Bicephale : I don't know what you mean by "rejection rate" but i see
no benefit by tweaking so far: your RCO went up and a
few bins were lost, it's probable no tweak can help here!

:p

tweaking

Thu, 15 Oct 2009 18:28:12 EDT

dslrocker3 : oh boy, playing around with these tweak settings is helping my line so much!!! never mind me, life is dull with a dsl line that has no major problems! lol.

The screen captures appear to be telling me that I have very few tones that fall below the default threshhold.

default dsnrm
lowest target dssnrm, lowest rejection rate
highest target dssnrm, highest rejection rate

Re: Help/Advice with my Line Statistics please

Thu, 15 Oct 2009 01:29:35 EDT

Bicephale : The 24 h record you provided shown a steady error rate on the UpStream side so,
yes, i suppose if it's stable enough then perhaps the outcome hapens to be a bit
predictable but that's because your phone-line is relatively close/"clean"...

Re: Help/Advice with my Line Statistics please

Thu, 15 Oct 2009 01:08:42 EDT

dslrocker3 : so it probably won't surprise you when I say that within 30 minutes, I know how many errors will average per hour when I look at the results 24 hours later. It seems like if it starts off at 20 errors/hour, it will stay like that until I reset it or until conditions get worse. But if it starts off at 40 errors/hour it will never be lower than that than until I restart the modem. I have no solid proof of this but that's what it seems like.

Re: Help/Advice with my Line Statistics please

Thu, 15 Oct 2009 01:00:04 EDT

Bicephale : What you appear to describe corresponds to what i tagged the "bleeding syndrome":

I've show the same two graphics to Oinktastic arguing that this is a "feature" of the
Thomson SpeedTouch FirmWare, apparently... At 1st the DS SNR Margin is high but
bins are lost to noise and this loss accumulates through time until the CPE DSL unit
renegotiates the connection with its peer on Bell's side. A reset cures the "bleeding",
it would seem, so the more frequent those resets are the more stable a connection
should prove to be in the meantime. It wouldn't be a surprize for me to find out that
Thomson meant this to happen by design: in the begining, one of the few purposes
of a 3rd-party tool such as 'DMT' was to monitor the DS SNR Margin and to reset a
customer's unit past some pre-defined limit... Maybe that's one of the reasons why
Thomson made their v6.2+ ST5x6/585 FirmWare 'DMT'-hostile, it's possible they had
plans of their own, go figure!

Monitor both the DS SNR margin and the number of bins through time, i believe you
might end up finding that those parameters are linked directly to each other, under
the proper conditions...

Re: Help/Advice with my Line Statistics please

Thu, 15 Oct 2009 00:04:54 EDT

dslrocker3 : I"m trying to say that the error rate seems "made up/make belief". how does the error rate stay at a constant pace for 24 hours straight but then reduce to half the amount for the next 24 hours simply because I reset the modem? The error rate over the next hour appeared to just instantly change. It's like the modem was anticipating a certain amount of errors and reports them just because it is expecting them.

or in the case of the speedsteram 4200 or 5200 e242, they will report a certain amount of errors every 15 minutes but when you reset it, all of the errors will sometimes dissapear immediately for the next little whie.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 23:57:23 EDT

Bicephale : I'm not sure i understand what you mean exactly but i do concur
with the opinion that there's something bot intuitive relatively to
the error rate statistics as i've often observed that such numbers
don't always scatter evenly. Sometimes it's like the error rate was
obeying to a discontinuous non-linear function where numbers are
falling into "pockets" with voids in between... Starting from there
i decided that it was OKay to "clip" my error rate curves since the
numbers become contaminated with some FirmWare glitch once it
has passed some level: be it thousands or billions the connection
is severed when it reaches that point anyway...

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 23:24:12 EDT

dslrocker3 : I've noticed that error numbers often seem almost fictional. I noticed the reduced rate of errors immediately upon resetting the connection. It's almost like the modem is creating fake numbers that can't be corrected until you reboot the modem.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 23:12:48 EDT

Bicephale : No idea, Alcatel ST Home that's too old even for me!

:p

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 22:22:15 EDT

dslrocker3 : why do I have a gut feeling that an Alcatel Speed Touch Home would be fine on this line now, even at 6016/800 Kbps and on IKNS? The only reason I got rid of that modem had to do with it's lack of ability to reach farther for signals. That modem was built to different standards compared to today's throw-away models.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 22:16:16 EDT

Bicephale : If you really tried being plugged trough a DSL filter
and there was no noticeable effect then my guess
is you can use whichever MoDem suits you better!

:p

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 22:10:50 EDT

dslrocker3 : the only thing that I can think of that could possibly explain how the 2wire model gets pretty much no errors even when synced at 768Kbps upstream and at an RCO of 100% would have to do with a higher rejection rate of certain tones that are producing the noise on my line.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 22:07:31 EDT

Bicephale : Ah, yes indeed! I don't know what i was thinking!

It's Oinktastic's fault, my crystal ball turned almost
opaque because of the fog he's causing me!...

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 20:29:03 EDT

dslrocker3 : will changing the target signal to noise ratio on the downstream side really affect the error rate on the upstream side? As you know, errors on the downstream side are non-existent.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 19:25:15 EDT

Bicephale : Hi,

Having v6.1.0.5 installed on your SpeedTouch you
can try to "tweak" the RCO which in turn will affect
the error rate... Go to the MoDemOptions tab then
move the silder down one step at a time while you
take note of the new readings. If it works this may
allow you to get the best of both worlds: 800 Kbps
with fewer errors. You could also start playing with
RF Choking, i suppose...

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 18:28:45 EDT

dslrocker3 :

said by Bicephale :

That's desperately dull, you should start searching
for another hobby, IMHO...

I can fix that!!! hahaha!!! (modem plugged into filter)

Interestingly, the errors have reduced in the past 17+ hours since I have reset my modem. The 50 or so errors on average per hour from yesterday have turned into about 25 total errors per hour today. This is about the error rate I was getting a few days ago and I thought that the spike to about double was because of the furnace being turned on for the first time this autumn. However, the drop in errors today would argue against that theory.

Regardless, I think that the errors are there but they aren't anything to be overly concerned about. Bichephale, I think you are right though about being unable to find anything noticeable unless I leave DMT running for a very long time. I doubt I will really leave my computer on for 30 days non-stop.

The graph from yesterday's error rates would be very flat. However, the graph from my 2wire modem would have nothing on it! I am now struggling on deciding if it is better to have a modest but constant amount of errors with my Speedtouch 516 synced at 6016/800 or to have virtually no errors (only 2 fec corrected errors over 16 hours) while synced at 6016/768 with my 2wire 2700HG-B.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 08:52:37 EDT

Bicephale : Hi again DSLRocker,

Good news, the unit's performance is so steadily
stable and eventless there's no need for graphs:
this is outrageously close to perfection and you'd
need to collect statistics for weeks if not months
in order to make anything emerge from such daily
records!!!

That's desperately dull, you should start searching
for another hobby, IMHO...

DSLRocker DM···.CSV.zip 19,585 bytes
SpreadSheet-ready .CSV version of data previously provided

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 03:33:14 EDT

Bicephale : Of course, there's still 'RouterStats' which is
a graphing application right from the start...

SpreadSheet work becomes essential only if
one gets involved about making comparisons,
etc. A capture of the 'RouterStats' graphics
would be good enough for people who simply
need to monitor their DSL unit, IMO.

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 03:24:54 EDT

Bicephale : Hi DSLRocker,

You could start reading here for the basics:

Kitz.Co.UK - Use DMT Tool to log your Line Stats

It may be possible to use a forgotten utility
from the author of 'DMT' but i'm not familiar
with it myself:

»dmt3d.mhilfe.de/

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 02:26:17 EDT

projectx : rename file to csv, and open in excel, simplest way i could think of, hope that helps

Re: Help/Advice with my Line Statistics please

Wed, 14 Oct 2009 00:52:35 EDT

dslrocker3 : I've attached a DMT Diagnosis log from the past 24 hours.

Is there an easy way for me to graph this? or do I need to enter each number into a spreadsheet manually?

DMTdiag13Oct2009.txt 102,055 bytes

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 23:55:13 EDT

Bicephale : That's good news! Now lets see if the numbers can accumulate over days!...

:p

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 22:17:42 EDT

dslrocker3 : I decided to check to see what was going on with the results and I compared the number of recorded errors in the modem interface to what was recorded by DMT's diagnose mode and it stopped recording resuts about 10 minutes later due to the "keep alive" failure. In Diagnose mode, DMT doen't even tell you of the error. It just stops working.

update: I left DMT open for the last 30 minutes, and it is not giving me the "keep alive" error message. It does appear that the "disablepingcheck" parameter was the key to getting DMT to work properly.

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 20:18:42 EDT

dslrocker3 : it just struck me that selecting even "SNRM" in dmt 7.35 might even be a waste because it might be recording the upstream which never changes because of IKNS. I'll restart and select "DS snm" instead.

I think that I might have kept the wrong modem. It really annoyed me that my d-link unit was set to bridge mode out of the box while the instructions say it was in routed mode and to just type in the 192.168.1.1 address and it further annoyed me that it doesn't work properly in routed ppp mode behind my linksys router. However, performance-wise, I think the D-Link DSL-2320B is better than the Speedtouch 516 running on 6.1.0.5. The d-link shows slightly lower attainable rates on the downstream side but better atainable rates on the upstream side. The d-link modem was consistently giving beter througput numbers on the upstream side. This is probably is fixable on the Speedtouch 516 with a new firmware but everything after 6.1.0.5 doesn't co-operate with DMT 7.35 properly. For me to replace this speedtouch 516 with something suitable for my line, I would probably have to grab another dsl-2320b (overpriced) or a TD-8841 (almost impossible to find these days).

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 19:37:38 EDT

Bicephale : As i recall the killer items were "Bitalloc", "SNR" and "ATT":

OrbMT on D-Link DSL 2320B, Bicephale, 2009-Jun-15

In this capture they're unselected for a very good reason...

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 16:32:31 EDT

dslrocker3 : With DMT 7.35, I'm not getting a "keep alive" error message when in Dianose mode. The error only seems to happen when I am connected in the main dmt window.

Diagnose Mode has been running for an hour now. I did uncheck things this time that can't possilby change. I know that the Speedoutch 516 will not update things such as max attainable, current sync rate, power output, attenuation, etc. unless the modem is told to retrain. The only things I left checked off were: SNRM, tx/rx crc, fec, and hec.

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 16:22:13 EDT

Bicephale : Hi DSLRocker,

The Siemens SpeedStream 4200 retains 96 samples (15 min. sampling
rate) internally until it resets but the last time you used 'DMT v9.01' i
saw extraneous data in it which inflated your log files for no practical
purpose and i ended up asking you for your configuration file in order
to reproduce your actual 'DMT' setup here but there was no follow up:

»Re: Help/Advice with my Line Statistics please

For this reason i can't suggest to use the Siemens SS4200 at this time.

Also, it may not the best choice with your new profile/peer...

I guess 'RouterStats' should do fine if you're having trouble wit 'DMT'
but if you must use the later then i'd take a look at those parameters:

'DMTv7.35.ini':

[TELNET]
SendKeepAlive = 1
KeepAliveIntervall = 90

[DEBUG]
disablePingCheck = 1

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 13:48:17 EDT

dslrocker3 : are you curious about noise margin, crc's and stuff like that? I'll have to actually keep my connection up long enough to do that. every time that I see a certain amount of errors on my line, it ends up annoying me so I end up resetting my modem (as if that makes the errors dissapear!!!!).

When I run DMT 7.35, it always complains about something about "failing a keep alive request" if I keep the program open more than a few minutes. Once that happens, the only way to get DMT to connect to the modem is to restart the modem or renew the modem's IP address with my router. Would you know what that is about?

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 13:14:06 EDT

Bicephale : Any plans to document the new link by collecting long-term statistics?...

Re: Help/Advice with my Line Statistics please

Mon, 12 Oct 2009 00:08:58 EDT

dslrocker3 : hhmmmmm... the pppoe authentication mystery continues. Over the past week, logging in has taken as much as nearly 8 minutes after modem power up/reboot sequence completed. At first, I thought this had something to do with my switch to an RDSLAM but now, I'm not so sure that has anything to do with it. Last week, I did one authentication to test@test and everything was connected within 2.5 minutes of the modem's power up but to my velcom login, it took 7.5 minutes after the modem booted.

but low and behold, I did a test just now and I was able to boot my modem, sync, and pppoe authenticate to velcom in 2 minutes, 13 seconds. I'm not so sure if this is a pppoe authentication server problem or not but that would seem to be a pretty large difference in authentication time.

I think that my AC was changed to bas1.toronto-26 last week, whereas I'm pretty sure it was bas1.toronto-1 before.

maybe, I should open a ticket and see what Bell does!!! (it's a joke that I think that some here will understand)

Re: Help/Advice with my Line Statistics please

Sun, 11 Oct 2009 17:54:56 EDT

Bicephale : Of course it won't happen, it's a fantasy and i forgot about crosstalk...

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 22:32:22 EDT

dslrocker3 :

said by Bicephale :

Honestly, my position about profiles is that it should be possible for
the DSL customers to manage theirs themselves.

Bell won't do this for the same reason they mess with the modems they provide. They will tell you it is to make it simpler and easier to support because there is nothing that the customer can mess up to prevent it from working (such as vpi's and stuff like that) but why in the world do they need to hide things like the sync rates and noise margins unless they are purposely doing that in the hope that customers do not notice that they are not getting the advertised speed?

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 15:38:54 EDT

Bicephale : Honestly, my position about profiles is that it should be possible for
the DSL customers to manage theirs themselves. This would reduce
the work-load for Bell and nobody is in a better position to tell when
it works. I imagine an additional star service linked to the DSL card:
one dials *DSL and picks the desired profile from a voiced menu. No
more tickets caused by improper profiles, smiles on every face... A
simple call and say good bye to disconnection evenings, thanks to
Bell-supported tweaking!

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 15:09:32 EDT

dslrocker3 : If I opened a ticket right now, I wonder what Bell would reduce the profile to? We all know that Bell doesn't read tickets and just plays the profile game a good chunk of the time. maybe i'll get lucky and get 672/160 Kbps!!!!

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 15:02:35 EDT

Bicephale : Correction: it's impossible in this universe, Bell's universe...
But what about the rest of God's creation where the usual
laws of physics still apply?!...

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 14:40:02 EDT

dslrocker3 :

said by Bicephale :

Hi,

In my opinion this would tend to prove that moving your DSL service
to that IKNS peer didn't remove the impedance mismatch caused by
what seems like a "Bridge Tap" to me; it also suggests that it might
be located close to your end...

But Bichephale, what I am showing here is impossible... besides, an RDSLAM can't help me, according to Bell!!!! rotfl

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 14:30:52 EDT

Bicephale : Hi,

In my opinion this would tend to prove that moving your DSL service
to that IKNS peer didn't remove the impedance mismatch caused by
what seems like a "Bridge Tap" to me; it also suggests that it might
be located close to your end... Really, some of those days, it would
be nice to have more convenient methods to compensate for those
"Bridge Tap" related impedance mismatches than to simply use some
unfiltered section of flat unfiltered telephone extension cord!...

Re: Help/Advice with my Line Statistics please

Sat, 10 Oct 2009 12:06:18 EDT

dslrocker3 : Remember how long ago that I needed to have an extension cord plugged in on an unfiltered part of the house wiring's loop to get the best results? Well look at this... After being hooked up to the RDSLAM, this still remains true. Disconnecting that extension cord worsens the connection. for this extension cord to be helpful, it needs to be visible to the modem (not centralized filtered like the rest of the house wiring), no devices can be plugged into it, and the modem can't be plugged into it. It would appear as if the possible impedence mismatch remains and what a Bell technician called impossible to have better results inside than at NID is still the case. The screenshots were taking immediately after each other.

Of note when unplugging the extension cord are the following:
-lower dowstream attainable rate
-lower dowstream signal to noise ratio
-increase downstream power output
-missing tones

extension cord unplugged
extension cord on unused jack plugged in

Re: Help/Advice with my Line Statistics please

Fri, 09 Oct 2009 19:40:10 EDT

wayjac : A remote dslam needs power see any meters?
The green box in the second picture looks a box that houses connections for local customer wire pairs/loops

Re: Help/Advice with my Line Statistics please

Fri, 09 Oct 2009 19:06:51 EDT

dslrocker3 : Just out of curiousity, which one of these would be one of my IKNS "friends"?

I'm just trying to get a general idea of the path that my line is taking. My assumptions probably won't be accurate though as there is some type of RDSLAM every 1/2 Km to 1 Km on every major street that surrounds by residential block. Like I said, just curious as this doesn't really affect me since I am already connected.

Bell equipment, Islington&Evans (NE corner), Etobicoke, Ontario
Bell equipment (maybe?), Islington&Gardiner (SE side), Etobicoke, Ontario

Re: Help/Advice with my Line Statistics please

Fri, 09 Oct 2009 01:33:47 EDT

Bicephale : Exactly!

Re: Help/Advice with my Line Statistics please

Fri, 09 Oct 2009 00:35:28 EDT

dslrocker3 : LOL, I think that I'm just trying to keep myself entertained now that I have a 6Mbps sync rate! I just have to realize that any further improvements probably won't yield any real-world speed improvements. The FEC and CRC errors that only happen on the upstream is kind of interesting.

if I was you, I'm not sure I would try to get connected to any RDSLAM, at least not at this point. I'm sure you know about the dangers of Bell trouble tickets and they might just slow your profile down on you even if there is no problem. Besides, I'm not so sure that Bell would be willing to connnect you to a remote dslam so that you can go from a 5056Kbps sync rate to a new sync rate of a full 0Kbps faster to 5056Kbps!

I'm wondering if I'll wake up one morning to see that I was switched to an ADSL2 profile.

Re: Help/Advice with my Line Statistics please

Fri, 09 Oct 2009 00:27:30 EDT

Bicephale : Oh but it's not off-topic at all to suggest that at this point it would be a
good idea to use our site's tools to document this adventure! If we can't
trust such IKNS numbers then these tests only make sense, i believe. All
you need to do now is to conduct multiple tests from multiple sources in
an effort to reject any possible aberration(s).

You'll need to make your MoDem "pingable", eventually. The procedure to
do this on a SpeedTouch is well known, there's a hint here:

»Disconnection TroubleShooting using an ST5x6

Re: Help/Advice with my Line Statistics please

Fri, 09 Oct 2009 00:08:27 EDT

dslrocker3 :

said by Bicephale :

Furthermore, i must be honnest: maybe i wouldn't like being on an IKNS
peer if i can avoid that...

I know we're probably off topic now but that's okay for this thread as my line problems are pretty much solved (for now). Plus, I'm sure the mods would rather have me talk about random things in this thread rather than to create a separate thread for every random thought that I have. :)

When it comes to the IKNS remote, I would also rather not be on it if I could get full speed without the RDSLAM. One of my other friends (not the one on IKNS) has a very similar attenuation to mine in the 10dB range and a max attainable rate of over 10Mbps but he is connected to an Alcatel DSLAM, according to my 2wire modem. I have a funny feeling that his line would suffer less from strange behavior and I'm also pretty sure that his line doesn't care what modem is connected (unlike mine). heck, he's using a Speedstream 5200 (the older version) and he still gets full speeds of what you would expect from a 6016/800 sync.

Is it just me or does it seem like the Globespan DSLAM's have a longer reach than Alcatel? The past few days, I keep seeing threads about attenuation in the 40, 50, and even 60's where the user gets above 3Mbps sync rates.

I just momentarily thought about if my line could now sync if I plugged my modem into a microfilter. that would be interesting. I know the friend who is also connected to IKNS forgot to filter a phone and he still got a sync of 5056/736 with his Speedstream 4200. It wasn't stable for obvious reasons but it would go a few hours at a time before dropping the connection.

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 23:40:55 EDT

Bicephale : This fall Bell's four feet boxes have been growing like mushrooms around
here so maybe i should consider asking to have my phone-line moved to
one of those myself, eventually, but then this would change my present
testing conditions and hence no further comparisons would be possible;
not to mention that i still haven't even asked for the "Bridge Tap" to be
removed just yet, considering it's been manageable for me so far...

Well, certainly irritating at times but manageable.

Furthermore, i must be honnest: maybe i wouldn't like being on an IKNS
peer if i can avoid that... Because of the lack of trust we have in their
FirmWare it seems this leaves us only a few basic options for testing:

»/speedtest
»/linequality

...and perhaps:

»/tweaks

I feel this would be the next step if i were in your situation...

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 23:14:42 EDT

dslrocker3 :

said by Bicephale :

Don't look at me: you've just entered a new dimension of the twilight zone!...

Bicephale, after all the strange things you've seen with my line, would it really surprise you? especially if you factor in the way that IKNS acts on the upstream side....

maybe I should open some more tickets to complain about the uploads when connected to IKNS. It'll be fun to see how many times Bell will attempt to make my line more "stable" (you know what this means and you know I'm just joking!). I'll probably end up on 512/512kbps with a downstream occupancy of 5% (but I bet they would still leave me on the rdslam). hahahaha

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 23:08:45 EDT

Bicephale : Don't look at me: you've just entered a new dimension of the twilight zone!...

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 22:56:40 EDT

dslrocker3 : It just struck me that all errors are on the upstream side. I know there's no way I could prove this and it's easy to blame something on it, but why do I have a funny feeling that this is just another IKNS characteristic?

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Thu, 08 Oct 2009 21:19:17 EDT

wayjac : For me the 2wire logs show were the problem is
None of the other logs have enough detail to show that

Thanks for posting the logs they answered some of the questions about the ikns workings

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 19:06:38 EDT

Bicephale : Maybe it's about time to experiment with RF Choking then!

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 18:37:36 EDT

dslrocker3 :

said by dslrocker3 :

Using the everything else remaining equal principle/assumption, it would appear that the 2wire 2700HG-B's strongest point on my line is the error rate. Over this 16 hours, the only errors that have appeared are 2 lonely FEC's on the upstream side

This isn't a big sample but it looks as if the Speedtouch 516 (and perhaps other Broadcom-based units) does have a fatal flaw, even when connected to this IKNS remote dslam. I believe it is becoming clear that this modem is affected more easily by noise and is less able to prevent line errors. I doubt this small amount of errors will have any effect on my ability to use the connection but the they are errors, regardless.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Thu, 08 Oct 2009 18:32:45 EDT

dslrocker3 :

said by wayjac :

Can you post logs from any of the other three modems?

Over time, my syncs upon initital power up are getting faster. When I first got connected to the RDSLAM, it was taking longer but I didn't document it with pure numbers.

However, look at the amount of time it is taking for the PPPOE log in. The first one is to test@test (a generic non-functional login to test the connection) but in the second log, the ISP's log in is taking about 7 minutes to authenticate.

modemlogs.txt 3,543 bytes

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Thu, 08 Oct 2009 16:13:13 EDT

wayjac : Can you post logs from any of the other three modems?

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Thu, 08 Oct 2009 16:01:58 EDT

dslrocker3 : Yes, the 2wire 2700HG-B can autodetct VPI/VCI settings while the other modems I have tried recently cannot.

I was simply trying to point out all DSL modems now take longer to sync now that I am connected to an Ikanos RDSLAM. Everything just takes longer now. - from syncing to logging in. It even takes any modem longer to even detect the DSL signal whereas it was instant before when the wriing went straight to the central office.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Thu, 08 Oct 2009 15:47:28 EDT

wayjac : Yea I didn't understand your question the first time and I'm still not sure I understand it now

Anyway, all four modems require the vpi/vci be specified in the gui

The 2wire can do a atm pvc search using its pvc search list.
Log messages will indicate if its done and the correct vpi/vci

Some speedstreams can do a atm pvc search using its pvc search list.
Log messages will indicate if its done

I don't think the 4200 can do a pvc search
I don't know if the 2320b or the 516 can do this
I'm sure the problem is the dslam, not the modems

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 13:58:27 EDT

dslrocker3 : I got my D-Link 2320B replacement in the form of a Speedtouch 516. I might have actually bought the Speedtouch for nothing becaue it's been a week and the buyer on Ebay hasn't even paid! That's okay because my friend who is also hooked up to IKNS says that he wants to buy the D-Link modem off of me if the Ebay person doesn't submit the payment. The friend doesn't really like used electronics from Ebay so he doesn't mind taking the d-link off me even if it costs more because he knows that I bought it new from the store and knows that there's nothing wrong with it. He was getting really annoyed with his the 736Kbps uplaod sync rate he was getting on his speedstream 4200+IKNS.

Just as I had suspected, these modems with this series of Broadcom chipset seem to be the only ones that I know of that can cope properly with the IKNS upload madness.

On a side note, what's with 6.1.0.5 and IE8 not accepting the Speedtouch user passwords? The passwords are accepted fine in DMT 7.35 and in Firefox 1.5.... doesn't matter, i can get in fine if the leave the password blank...

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 13:52:54 EDT

dslrocker3 :

said by Bicephale :

Life 's going to seem dull!...

Yes with only 2 FEC errors in 16 hours, there's really nothing to test for. How am I supposed to test for noise when the modem isn't finding any!!????

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 13:48:39 EDT

Bicephale : Life 's going to seem dull!...

Re: Help/Advice with my Line Statistics please

Thu, 08 Oct 2009 12:55:19 EDT

dslrocker3 : Using the everything else remaining equal principle/assumption, it would appear that the 2wire 2700HG-B's strongest point on my line is the error rate. Over this 16 hours, the only errors that have appeared are 2 lonely FEC's on the upstream side. I'm not worried about the errored seconds because I have found that it is normal for this modem model to accumulate the vast majority of those immediately upon startup.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 22:43:27 EDT

Bicephale : Euh...

I mean what if the DSL device already knew
the VPI/VCI pair and didn't have to search?

...

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 22:30:57 EDT

wayjac :

said by Bicephale :

I'm going to try a wild question here.
What if the DSL device didn't have to
search for a VPI/VCI pair?...

That would be another technology

Every modem makes a atm connection to the other modem [dslam] atm uses pvc's

The atm pvc vpi/vci pair is a basic part of adsl

The question you should be asking is why is the ikns atm connection slow to respond its the "always up" part of adsl
Other dslams are waiting for the modems atm

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 22:16:38 EDT

Bicephale : I'm going to try a wild question here.

What if the DSL device didn't have to
search for a VPI/VCI pair?...

:o

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 22:16:08 EDT

dslrocker3 :

said by wayjac :

Ok, like I said if the 7616 in not in the log it never happened

Yeah, I know that 7616 never happened but where in the world is that number coming from? Like I said, I know that spot in the MDC page is dedicated to the profile setting speed but Bell Canada has no such profile at that speed. it must be a firmware bug.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 22:13:36 EDT

wayjac : Ok, like I said if the 7616 in not in the log it never happened

Looking at the posted log

lmd: device0: dsl0 has no current version set

I think that line denotes the end of boot [2min 8sec]

From that to dsl sync looks like 47sec
From there it took 3min 35sec to find the atm pvc notice the

round 33

After that 2sec until the ip is up

If you have some old logs compare them
For me it looks like the atm is the hold up

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:53:40 EDT

dslrocker3 : 6.5 minutes from boot until the internet connection was usable (log attached).

I don't remember log ins ever taking this long until after I was connected to an RDSLAM.

2wirelog.txt 5,971 bytes

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:42:31 EDT

dslrocker3 : for the sync rate thing on the 2wire modem, here it what it says in the detailed log:
INF P0000-00-00T00:02:52 lmd: dsl0: up G.DMT Annex A interleaved Rate:6016/768 Max:10036/768
INF P0000-00-00T00:02:52 lmd: dsl0: Margin:24.0/6.0 Atten:11.2/4.5 Power:9.8/12.6
INF P0000-00-00T00:02:52 lmd: dsl0: Country: {B5} Vendor: {IKNS} Specific: {1}
INF P0000-00-00T00:06:27 lmd: apvc0: found BRDG_LLC PPPoE on 0/35 round 33

but in MDC is says:
Downstream Rate Cap: 7616 kbps
Rate 6016
Max 1 8128
attn 11.2
Mgn1 25.1

what i'm trying to say is that the 7616Kbps profile isn't one of the profile options with Bell.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:38:23 EDT

wayjac : I wish I could offer some reasons for this but I can't

Can you post some logs

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:12:17 EDT

dslrocker3 : What I"m really asking is why it's taking so much longer to log in than before ever since getting connected to the remote dslam.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:11:06 EDT

wayjac :

said by dslrocker3 :

I still don't understand why the PPPoe log in now taking so long.

If I'm going to change modems/pppoe clients I make sure I end the pppoe session or use a different pppoe user id/password

My isp will give "limited" tcp/ip info if the session is not ended and I try to log in with the same pppoe user id/password

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:06:59 EDT

dslrocker3 :

said by wayjac :

With the 2wire every page that has the sync rates uses a different term/pharse and every modem vendor has it own set of terms/pharses
Then you have the telcos/isp's who almost never use the sync rate for marketing

except in Canada where the DSL ISP's always use the sync rate for marketing

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:03:51 EDT

wayjac : With the 2wire every page that has the sync rates uses a different term/pharse and every modem vendor has it own set of terms/pharses
Then you have the telcos/isp's who almost never use the sync rate for marketing

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 21:01:55 EDT

dslrocker3 : I could swear the my syncs are getting faster when I turn my modem on compared to when I first got hooked up to the IKNS RDSLAM. I still don't understand why the PPPoe log in now taking so long. The PPPoe logging in now gets on my nerves more than the sync time.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 20:59:07 EDT

Bicephale : Hi DSLRocker,

DR> Does the RDSLAM have a learning curve...

I sure asked that very same question to myself on some
occasions before! One day everything was just fine until
i switched MoDems after a long-term test, then the unit
acting as a replacement would cause trouble for the next
day or two while momentarily switching back to the unit i
was using previously worked fine again as if nothing had
happened. To me, it was like my MoDem's peer somehow
remembered the unit's signature and was slow/reluctant
to acquire a new one...

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 20:34:57 EDT

dslrocker3 :

said by wayjac :

"downstream rate cap", atu-c current tx rate,
"connected at", "max allowed speed", "Connection Rate" and profile setting

Each phrase boils down to sync rate

I realize the telco has set profiles but the sync rate (profile setting) is set by a human and we all know humans do make occasional mistakes

I completely agree with you but I just happen to know that 2wire went against the conventional terms when they really should have just said "profile speed". On the same MDC page for the 2wire, it does show a Rate of 6016Kbps but just lower down on the page.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 20:16:42 EDT

wayjac : "downstream rate cap", atu-c current tx rate,
"connected at", "max allowed speed", "Connection Rate" and profile setting

Each phrase boils down to sync rate

I realize the telco has set profiles but the sync rate (profile setting) is set by a human and we all know humans do make occasional mistakes

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 19:38:18 EDT

dslrocker3 :

said by wayjac :

The modems detailed log is the last word on sync rates

I've looked at modem logs that show a dsl retrain taking less than 10 seconds speedstreams and 2wires

"downstream rate cap" in the 2wire's mdc page really is refering to only the profile setting. At one point Bell and had me set on a profile of 4032/512 Kbps and yes, "downstream rate cap" said 4032. I was speaking with a technician on the phone when he made this profile change. However,my actual sync rate was below 2.0Mbps.

Also, when I got swtiched to the 6016/800 profile, i was speaking to the Bell High Speed Centre that sets all profiles and he also told me he set it to 6Mbps. If you look at my screenshots while that I took while using the other modems, it is obvious that Bell is limiting me to 6016Kbps with absolutely no possibility of syncing higher no matter what I do on my end. Trust me when I say my line can do over 10Mbps.

Wayjac, yes the 2wire modem can retrain but you are forgetting that if it did retrain, this would show in the MDC page. It would actually say "retrain higher". The general statistics page clearly shows a 6016Kbps sync rate and in MDC, it shows that no retrain was ever completed. The MDC page shows complete training history since the last reset. All screenshots for the 6016/768 sync on the 2wire were from the same connection (no retraiining, no resetting, now power cycling, etc.).

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 19:33:04 EDT

wayjac : The modems detailed log is the last word on sync rates

I've looked at modem logs that show a dsl retrain taking less than 10 seconds speedstreams and 2wires

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 19:10:56 EDT

dslrocker3 : I can absolutely promise you that the sync rate is only 6016Kbps. Please see the further details in the MDC dianostic page and you will see that it says Rate 6016. This is the actual rate that the modem has negotiated with the DSLAM.

As for "downstream rate cap".... I have known for a while this only refers to the downstream profile, regardless of the actual sync rate. This is just the way that 2wire words it which is contrary to how other manufactuers use similar wording to describe the sync rate. If you look at previous screenshots in this thread, this becomes clear. However, like I was saying, my profile isn't 7616Kbps and neither is my sync. That is why I believe it is either an error because of either the 2wire or IKNS firmwares. All of the 2wire screenshots in the most recent post with pictures were from the same active connection. Also, Bell Canada does not have a 7616Kbps downstream profile for G.dmt.

My speedtest match up perfectly with a 6016Kbps sync rate. (usually between 4900 and 5100 Kbps depending on the test site).

Yes, my line supports 8Mbps very easily but 6016Kbps is Bell's highest sync rate for ADSL. Anything higher with Bell is ADSL2.

a theory: maybe my download profile really is set to 7616Kbps. like i said is this a firmware bug? or.... I wonder what would happen if I used a modem that I could force into ADSL2 or ADSL2+ mode..

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 19:04:11 EDT

wayjac :

said by dslrocker3 :

The sync rate was actually at 6016Kbps. For some reason, the modem is reporting a profile speed of 7616Kbps but that must be a firmware bug or another IKNS abnomoly because if the profile speed really was set that high, the sync would have been a bit higher (there's enough of a buffer to sync much higher).

I think the sync rate was set to 7616kbps I've never had a modem display the wrong sync rate.
I believe bell has sync rates like that
That rate would show 6000kbps on a speed test
Do you know if the max rate stays above 10000kbps, if it does your loop would support 8000kbps no problem

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 18:18:15 EDT

dslrocker3 : I'm not sure what happened with the 2wire. On the previous time that I tried, it wouldn't sync when I was speaking to the Bell tech on the phone.

I didn't actually look in the 2wire logs, but I did look at a clock briefly to get an idea of how long it took to sync this time. For the speedstream modems, I did check the logs and you had to wait 5 minutes before your connection would be usuable.

I'm really wondering why I could not sync with the 2wire one day but can do so fine now? Does the RDSLAM have a learning curve, perhaps?

The sync rate was actually at 6016Kbps. For some reason, the modem is reporting a profile speed of 7616Kbps but that must be a firmware bug or another IKNS abnomoly because if the profile speed really was set that high, the sync would have been a bit higher (there's enough of a buffer to sync much higher).

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 18:03:36 EDT

wayjac : I'm shocked to see the 2wire with sync, I read your explanation twice I still don't understand.............

One of the 2wire images show 7616kbps sync rate?

I think the upload is a non issue (80kbps or so difference)

said by dslrocker3 :

both the speedstream modems and my 2wire modem took about 2 minutes to sync so no big deal there. My d-link dsl-2320b takes less than a minute to sync but I didn't actually time it - it just seems faster at it. There is an annoyance with the Speedstream modems that I have in the sense that they take approximately 3 minutes after sycing to complete PPPoE log in.

Do the modems logs confirm the sync and pppoe times

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 17:41:17 EDT

Bicephale : Not to mention that there are cases as with Thomson's
ST5x6/585 series which may improve once the FirmWare
is replaced with another - and we got plenty of those!...

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 14:28:06 EDT

dslrocker3 : I really should learn from my past modem experiences. I bought a speedtouch 516 at full price and then sold it back in April because it performed poorer than my other modems at the time. I sold it at a loss only to buy another one a few months later to replace a d-link 2320b that I sold because I didn't like the way it peformed on my line. Once I sold the D-link modem, I regretted it because it now outperforms the other modems. So I think I might actually save money by just keeping all my modems so that I don't have to keep rebuying them each time line conditions change. LOL!

and yes, I've already had a 2wire power supply decide to "laugh in my face" immediately following a loud thunder sound (within a minute).

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 14:22:31 EDT

Bicephale : Don't forget lightning storms which can hit nearby in the
summer and destroy a MoDem instantly. If the only issue
is dust then place your unused units in a bag and stick a
tag on it which will be marked: DSL Insurance!...

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 14:13:07 EDT

dslrocker3 : I know I am going to keep a backup modem to go with one of the Broadcom-based units but now I have to decide whether to part with some of the extra modems. It's almost a waste to have the extra modems just collecting dust in my desk but at the same time, part of me says to keep one modem based on each chipset just in case line conditions change or if Bell makes minor (hopefully not large changes) to my line. Each little line factor seems to influence which modem works best for me.

Re: preliminary modem comparisons(short term) with 6016/800Kbps

Wed, 07 Oct 2009 14:02:00 EDT

Bicephale : It all looks nice overall, except the 768 Kbps US BitRate
and "DSL Unavailable Seconds" perhaps. I'd still suggest
you try mild RF Choking, eventually...

Anyway, lets hope it will hold!

preliminary modem comparisons(short term) with 6016/800Kbps IKNS

Wed, 07 Oct 2009 13:32:12 EDT

dslrocker3 : It's been 48 hours since the profile adjustment to 6016/800 and I decided it was worth the gamble to take a look at how these modem perform after changes were made to my line. I'm hoping Bell doesn't see the disconnections as a sign up instability and just lower the profile but I think 48 hours is the normal time-frame window to resolve tickets so I don't think they are actively monitoring my line any more. (Besides, disconnecting 3 times in the last 2 hours isn't quite the same as rewiring the house or trying 30 different modems or leaving the modem plugged in while doing rewiring and making their logs show thousands of line errors). :)

In all reality, I don't think it really matters which modem I use on my line, with the exception of the slight variance in upload syncs.

The problems that I had before with the 2wire unable to sync might have been just because I was trying to get it to sync while I was still talking to the Bell technician. That still doesn't explain why it was syncing on my friend's IKNS line but kept droping sync every 30 seconds (could be a line problem on his end as he has a very similar attenuation but a noticebly lower signal to noise ratio). I also switched the line selection from automatic to line 1 to prevent the modem from timing out and switching back and forth from inner to outer pair.

both the speedstream modems and my 2wire modem took about 2 minutes to sync so no big deal there. My d-link dsl-2320b takes less than a minute to sync but I didn't actually time it - it just seems faster at it. There is an annoyance with the Speedstream modems that I have in the sense that they take approximately 3 minutes after sycing to complete PPPoE log in. I'm not sure why they take that long but I didn't notice a delay that got on my nerves like that with the other modems.

The speedstream modems do appear to have an edge in the inititial negotiation with a starting signal to noise ratio synced at 27.0dB for the downstream. However, these 2 speedstream models were unable to maintain that and quickly dropped that higher SNR while consistently fluctuating between 22.5 and 25.5dB.

D-Link DSL-2320B
Speedstream 5200 E242
Speedstream 4200 A8M
2wire 2700HG-B

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 19:00:10 EDT

Bicephale : I understand you! Actually, i wouldn't want to get annoying to
Velcom myself or i would not have planned to pass by the mail
office tomorrow to send them my quarterly payment otherwise!

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 18:46:57 EDT

dslrocker3 :

said by Bicephale :

I never feared asking VelCom for a copy of Bell's numbers
but i didn't have opened tickets when i did, nor did i have
the pleasure to find my profile set to 6016 Kbps lately. It
can wait if you don,t feel like it. Enjoy!

:D

don't get me wrong, I am curious but with the amount of tickets that I've opened, I don't want to get annoying to Velcom. They've been good to me so they probably deserve a break from hearing from me! :) And plus, I don't want anything I say to be misunderstood as asking them to open a ticket with Bell, because we both know Bell will play the slowdown game.

The Broadcom inconsistencies (900Kbps downstream sync one minute, 2.2Mbps another minute and nonsense like that) have dissappeared since being put on this IKNS remote. the numbers are relatively fixed and constant based on the few times I've reset the connection. The noise margin also stays pretty much stuck at a constant number too while the connection is up.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 18:44:00 EDT

Bicephale : I never feared asking VelCom for a copy of Bell's numbers
but i didn't have opened tickets when i did, nor did i have
the pleasure to find my profile set to 6016 Kbps lately. It
can wait if you don't feel like it... Simply enjoy the ride!!!

:D

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 18:29:35 EDT

dslrocker3 : I was talking to velcom because they called to let me know the ticket was closed and he wanted to know if everything was fixed but I'm scared to ask questions because all tickets say "slow speed" regardless of the actual reason for the ticket and nothing else once Bell gets it and that's when they just go around and lower profiles. I already had to argue with Bell not to set me at 512kbps for the upload.

I'm just not sure that it's worth the risk, espcially when I'm not sure that I'm even supposed to be even getting 6Mbps sync. Is that just an unofficial profile for companies like velcom/teksavvy/acanac, etc. or is it a Bell Sympatico only profile?

besides, would you trust anything anything that an IKNS reports?

by the way, what's a safe amount of time to wait before messing around with the connection? ie. when would any auto-script stop monitoring the line?

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 18:07:18 EDT

wayjac :

said by dslrocker3 :

it looks like it might be fast path or at least an interleaved path with only a very short delay.

Sure looks like fastpath to me


Tracing route to velcom.ca [64.86.17.30]
over a maximum of 30 hops:
1   1 ms   1 ms   1 ms  home [192.168.2.1]
2   10 ms   9 ms   9 ms  adsl-76-252-39-254.dsl.ipltin.sbcglobal.net [76.252.39.254]
3   11 ms   9 ms   9 ms  dist2-vlan62.ipltin.sbcglobal.net [69.217.79.50]
4   10 ms   9 ms   9 ms  bb2-g8-0.ipltin.sbcglobal.net [151.164.241.198]
5   187 ms   25 ms   176 ms  69.220.8.51
6   24 ms   33 ms   35 ms  asn6543-teleglobe.eqchil.sbcglobal.net [151.164.250.18]
7   60 ms   55 ms   53 ms  Vlan4.icore1.TTT-Scarborough.as6453.net [206.82.141.2]
8   48 ms   56 ms   47 ms  Vlan2.msfc4.TTT-Scarborough.as6453.net [207.45.220.139]
9   50 ms   47 ms   47 ms  Vlan451.msfc4.TTT-Scarborough.as6453.net [209.58.7.10]
10  49 ms   48 ms   49 ms  core-tor.velcom.com [206.53.48.254]
11  48 ms   48 ms   47 ms  velcom.ca [64.86.17.30]

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 17:51:10 EDT

Bicephale : Now that i think of it, have you called VelCom to get the
numbers from Bell's side?...

;)

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 17:00:27 EDT

dslrocker3 : it looks like it might be fast path or at least an interleaved path with only a very short delay.

Yes, I am at 6016/800 but I didn't really want to question the techncian when he came back on the phone about if it was interleaved or fast path when I saw that modem had synced at 6016Kbps.

My point was the fact that he first tried 5056/512 before even talking to me and I had to read out my modem's RCO statistics before he hestitently agreed to increases the uplaod. Because of the way the Ikns RDLAMS act, I think that there will be a lot of people who will have reduced upload syncs at 512Kbps in attempt to improve stability or perhaps just because IKNS seems to give strange readings.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 15:51:03 EDT

Bicephale : But isn't your present profile 6016/800 Kbps FastPath?

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 14:33:41 EDT

dslrocker3 : with the "marvelous" IKNS brand remotes, expect to see a lot more 5056/512 Kbps profiles. I think IKNS is the real reason the technician didn't want to set my upload to 800. When I was asking him to to try 640 or 800Kbps, he told me that there was no 5056/640 profile.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 14:03:51 EDT

Bicephale : Oh, i see... Well you found it anyway and it looks good!!!

Maybe there's a TraceRoute item within the MoDem's GUI
itself but i don't recall.

:)

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 13:40:08 EDT

dslrocker3 : C:\Users\->tracert www.velcom.ca

Tracing route to velcom.ca [64.86.17.30]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.3.1
2 15 ms 19 ms 14 ms 64.213.140.2
3 15 ms 14 ms 15 ms 64.213.140.1
4 15 ms 16 ms 17 ms 204.14.16.237
5 15 ms 40 ms 17 ms ge-6-12.car1.Toronto2.Level3.net [4.59.180.21]
6 46 ms 32 ms 35 ms ae-8-8.ebr1.Chicago1.Level3.net [4.69.140.245]
7 35 ms 33 ms 30 ms ae-1-51.edge3.Chicago3.Level3.net [4.68.101.20]

8 * * * Request timed out.
9 45 ms 49 ms 53 ms Vlan4.icore1.TTT-Scarborough.as6453.net [206.82.
141.2]
10 224 ms 47 ms 49 ms Vlan2.msfc4.TTT-Scarborough.as6453.net [207.45.2
20.139]
11 48 ms 49 ms 47 ms Vlan451.msfc4.TTT-Scarborough.as6453.net [209.58
.7.10]
12 50 ms 49 ms 49 ms core-tor.velcom.com [206.53.48.254]
13 47 ms 49 ms 62 ms velcom.ca [64.86.17.30]

Trace complete.

»/linequality/nil/2572427

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 13:31:37 EDT

dslrocker3 :

said by Bicephale :

Very nice, but it would be even nicer if it didn't say "Interleaved"!

Perhaps now is the time for a TraceRoute and, yes, lets not make
them panick by switching MoDems back and forth!...

:p

Yes, I'm sure Bell XDSL would be very annoyed if I started disconnecting/reconnecting modems, rewiring jacks, testing at the NID, etc. now. They would either be annoyed or just slow me back down.

I know this sounds like a really stupid question but how do I do that? I'll search as well for an answer as I might be able to find the answer about how to traceroute more quickly than waiting for a reply.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 13:00:41 EDT

Bicephale : Very nice, but it would be even nicer if it didn't say "Interleaved"!

Perhaps now is the time for a TraceRoute and, yes, lets not make
them panick by switching MoDems back and forth!...

:p

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 12:55:21 EDT

dslrocker3 : doesn't look like errors are going to be a problem either (there were a few but a very small amount).

hopefully, my other modems like this profile. but i'll try to wait until Bell isn't sniffing around to get stats from those....

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 12:01:43 EDT

dslrocker3 : well, well, will you look at that!!! I'll probably wait a week or 2 before conducting experiments. Now that the Bell tech just talked to me, I'm certain that Bell is still monitoring the line so I think now would be a really bad time to start playing around.

They wanted to set me to 5056 download and 512Kbps on the upload but I argued with him because he was telling me numbers about the upstream capacity. I guess he didn't know about my friend called "DMT 8.07!"

Thanks for everyone's help who provided comments and suggestions (espcially bichephale but also all of the other broadbandreport users), Velcom tech support, and yes even thank you to Bell for finally completely the work.

I did a speedtest and got a result of 5066/664Kbps, and another that gave a download result of 5127Kbps.. sounds about right...

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 11:19:05 EDT

dslrocker3 : nothing new here but this screenshot is just to document an error free line (I think?) in action with this combination of modem and current speed profile.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 09:58:02 EDT

anon :

said by dslrocker3 :

He also gets 736Kbps upload with a speedstream 4200 on a 100% confirmed IKNS line which is the same number I get with a Speedstream 4200....

The way my DSL-2320B modem connects at 800Kbps on my line and his too further confirms a likely IKNS RDSLAM on my end too. Both of our lines also had problems with a 2wire 2700HG-B modem with one line refusing to sync and the other having a large amount of difficulties maintaining one (modem is known working fine on a non-IKNS line).

Yup, thats an IKNS.

There is a code modems show in their logs (most modems) that can also confirm this as well. But what you describe above is good enough to show this is an IKNS without having to 2nd guess it.

The 2wire. Which firmware are you running (not like I know them myself), but there is an AT&T (I think it's AT&T) firmware floating around in the TSI forum that people have used (You can use search to find it). Don't know if it interests you to try another firmware on the 2wire.

I think its an IKNS problem (defect) that causes problems with certain modems. So in my belief (w/o really knowing) I don't think a diff firmware on the 2wire would change the situation. But Don't know.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 09:24:52 EDT

dslrocker3 :

said by Robrr :

Would it be possible for you to post full line stat info for each modem on the line you were using. Would be interesting to see a comparison.

Seeing a modem sync at a full 800Kbps on an IKNS remote is rare.

Please see one page back for stats from my own line. I didn't capture stats from my friend's line but I can safely say that they will almost be identical with the exception that he is already on 5Mbps. His attenuation is also ~2dB higher on the downstream. He also gets 736Kbps upload with a speedstream 4200 on a 100% confirmed IKNS line which is the same number I get with a Speedstream 4200 so I'm about 95% certain that I am also hooked up like that as well. The way my DSL-2320B modem connects at 800Kbps on my line and his too further confirms a likely IKNS RDSLAM on my end too. Both of our lines also had problems with a 2wire 2700HG-B modem with one line refusing to sync and the other having a large amount of difficulties maintaining one (modem is known working fine on a non-IKNS line).

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 08:19:31 EDT

Bicephale : Hi DSLRocker,

DR> I'm going to get really bored no longer needing DR> to figure out what is wrong with my line.

You'll have plenty of time ahead of you to investigate later
but if the "Bridge Tap" is still there i suppose it could mean
the "echoes" are also stronger and hence cause the bizarre
discrepancies from one brand/model to the next which you
have been observing so far... My bet is that this is likely to
get worst at the "normal" BitRate of 5056/800 Kbps FP.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 06:42:45 EDT

anon :

said by time still :

said by Farchord :

You are aware that, past mid-november, Unlimited internet will become extinct in Canada right...?

Nexicom and Primus are in Peterborough offering unlimited now on ADSL2+.
Can't see them changing that offering as a result of the Gas UBB decision.

Don't forget videotron's unlimited out in Quebec. And whoever offers unlimited still from this group:
»Re: ISP's that don't throttle and have their own equipment. LIST

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 06:10:19 EDT

Robrr : Would it be possible for you to post full line stat info for each modem on the line you were using. Would be interesting to see a comparison.

Seeing a modem sync at a full 800Kbps on an IKNS remote is rare.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 02:52:02 EDT

dslrocker3 : Re: speedstream 4200 vs dlink dsl-2320b when connected to IKNS RDSLAM

ok, I tested both of these modems on another IKNS line in the same neighborhood

on my friend's line,
speedstream 4200: 5056/736Kbps sync,downstream snr 22.5dB
D-link dsl-2320b: 5056/800Kbps sync, downstream snr 25.0dB.

It would appear as if the Broadcom 638 chipset has a clear edge over the AR7 chipset when connected to an IKNS rdslam and the downstream attenuation is about 10 to 15dB.

The d-link modem also synced noticeably faster.

Re: Help/Advice with my Line Statistics please

Mon, 05 Oct 2009 01:07:34 EDT

anon :

said by Farchord :

You are aware that, past mid-november, Unlimited internet will become extinct in Canada right...?

Nexicom and Primus are in Peterborough offering unlimited now on ADSL2+.
Can't see them changing that offering as a result of the Gas UBB decision.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 23:02:29 EDT

dslrocker3 : I'm aware about all the nonsense Bell is ramming through. we'll just have to see what happens.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:58:45 EDT

Farchord :

said by dslrocker3 :

Nah, I don't like that Rogers cable internet isn't unlimited. 3web also stopped selling cable intenet service in Ontario, to my knowledge.

I could always still lurk around once in a while to share what I've learned. I'm no where close to being an expert but I'm sure there will be others who have a line similar to mine who might be able to use information that I share.

It's really funny how Broadcom appears to be great for my line in it's current state. But if I wasn't on an IKNS (almost certain that I am indeed on IKNS), I probably wouldn't bother with modems based on this chip.

I'm going to get really bored no longer needing to figure out what is wrong with my line.

I should add something like this as a signature to my posts: "'Bell quote: a remote won't help because you are on a bridge tap' hahahahaha!!!"

You are aware that, past mid-november, Unlimited internet will become extinct in Canada right...?

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:51:46 EDT

dslrocker3 : Nah, I don't like that Rogers cable internet isn't unlimited. 3web also stopped selling cable intenet service in Ontario, to my knowledge.

I could always still lurk around once in a while to share what I've learned. I'm no where close to being an expert but I'm sure there will be others who have a line similar to mine who might be able to use information that I share.

It's really funny how Broadcom appears to be great for my line in it's current state. But if I wasn't on an IKNS (almost certain that I am indeed on IKNS), I probably wouldn't bother with modems based on this chip.

I'm going to get really bored no longer needing to figure out what is wrong with my line.

I should add something like this as a signature to my posts: "'Bell quote: a remote won't help because you are on a bridge tap' hahahahaha!!!"

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:49:39 EDT

wayjac :

said by dslrocker3 :

I'm getting impatient now.

If it was a co connection ok....be impatient
A rdslam has a 3 or 4 day delay?

I had a look at a emulator for the dlink......I only see a option for the lan ip address and nothing else....
I must be sleepy.........yea that's it....................

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:47:25 EDT

Bicephale : Hummm... Now that it's about to be a solved case i guess
you should drop DSL InterNet completely and start playing
with cable InterNet instead: new challenges, new learning
curves waiting for you to climb them!...

:p

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:36:08 EDT

dslrocker3 : I'm getting impatient now. Bell hasn't set me to 5Mbps yet. I'm going to miss complaining about my connection and being so negative. :)

I wonder how my line stacks up in the "Largest one day drop in attenuation contest".

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:15:15 EDT

Bicephale : If it's only a problem with the CLI then maybe
you just haven't found the right D-Link model:

best modem ?, Bicephale, 2009-May-13

Actually, i wish there were more like this...

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:11:51 EDT

dslrocker3 :

said by wayjac :

I don't think I really understand what you're saying............
Would giving the router a private ip that is the dmz work?

Yes, you're gonna have a rough time getting used to stabilty

I have a linksys router. When I set up the linksys in automatic mode, all appears fine and the router assigns an address of 192.168.1.1 to the modem (i already changed my linksys router's ip to 192.168.3.1 because of the conflict).

But as soon as the dlink modem completes the PPPoe login, the linksys will then take the the IP address assigned to your internet connection by your ISP and incorrectly map that as belonging to your modem. This results in the Linksys router failing to be able to communicate with the D-Link modem. Same thing happens if I set a static IP address in the Linksys router for the d-link modem. This makes it mandatory to run the d-link modem in bridge mode with the router. that's what I call a pretty badly designed firmware.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 22:06:07 EDT

wayjac : I don't think I really understand what you're saying............
Would giving the router a private ip that is the dmz work?

Yes, you're gonna have a rough time getting used to stabilty

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:44:34 EDT

dslrocker3 : in a way, i'm hoping that the person on ebay doesn't pay for this d-link so I can keep it because the d-link gives nice stats without the need to use third party software. However, d-link's firmware drives me nuts (there is a bug forcing you to use bridge mode because using auto dhcp or a static IP in your router causes the router to incorrectly think that your ISP's assigned IP address is your modem's address, see the problem there?)

even if this ebay buyer doesn't pay, I'm sure that my friend who has also been "IKNSed" might be interested in a modem that gives him the full 800Kbps upload sync. His speedstream 4200 also reports a rate of 736Kbps.

what a pain to check stats on this dlink modem while behind a router... but I did a quick check and over the past 1 hour, 3 minutes, there have been zero errors of any kind so that's also good news.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:38:44 EDT

wayjac : Yea, you're right I forgot about the chip it is the replacement and it works with the stats utility
Thanks for the image!!
I'm thinking you are gonna be content with the short wire pair

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:25:24 EDT

dslrocker3 : I know that the my 2wire modem will tell me the dslam/rdslam vendor in the MDC, DSL diagnostics page. It might even tell me before sync is even established but I'll try to confirm that once everything has been settled down with my line. (profiles adjusted, when my line is no longer be monitored for errors, etc)

The Speedtouch 516 that I got for $20 coming in the mail soon should do nicely (same chipset) except without dlink's buggy firmware.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:21:12 EDT

wayjac :

said by dslrocker3 :

I don't have 100% confirmation of IKNS because only my 2wire reports that information but I'm almost certain it is IKNS because of the way this remote dslam behaves.

Have you looked at the atm and wan page?

said by dslrocker3 :

I'm kicking myself becaue I hated this d-link so much on my non-rdslam line and the way it performed that I sold it on ebay.

That's too bad it has complete stats and the dslam likes it too
You may end up looking for a replacement

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:10:19 EDT

dslrocker3 :

said by wayjac :

said by dslrocker3 :

I'm trying to point out how the broadcom chipset has suddenly decided to shine when I got hooked up to the IKNS remote.

Looks like the ikns likes the broadcom chipset the best
Did the broadcom sync up quicker the the others? If it did that's your modem

I don't have 100% confirmation of IKNS because only my 2wire reports that information but I'm almost certain it is IKNS because of the way this remote dslam behaves (strange upload stats).

Honestly, I wasn't really paying attention to how long it took to connect. Now that I have a 800Kbps upstream sync, I want to try to leave the connection up until they run the line test for profile adjustment.

I'm kicking myself becaue I hated this d-link so much on my non-rdslam line and the way it performed that I sold it on ebay. that's okay... i managed to snag a speedtouch 516 for about $20 and I would expect that it would behave similarly on this line.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:04:28 EDT

wayjac :

said by dslrocker3 :

I'm trying to point out how the broadcom chipset has suddenly decided to shine when I got hooked up to the IKNS remote.

Looks like the ikns likes the broadcom chipset the best
Did the broadcom sync up quicker the the others? If it did that's your modem

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 21:02:04 EDT

dslrocker3 : yes, one of my modems is a Bell Speedstream 5200 E242 (all black) one that has the good AR7 chipset.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 20:59:31 EDT

wayjac : Ah forget the 2wire i caught a glance of the other modem
damm ikns

All this time I'm thinking you had a 5200 you actually have a bell 5242 correct?

Can you post a image of its lable?

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 20:34:56 EDT

dslrocker3 : ok, i will look into the 2wire once everthing is all set and good after all speed profile are properly set.

but these pics should be interesting... I'm trying to point out how the broadcom chipset has suddenly decided to shine when I got hooked up to the IKNS remote.

I wasn't going to bother with all this disconnecting and everything because I didn't want Bell to use the disconnection as an excuse not to raise the profile. However, I felt that I had to try to get an 800Kbps upload sync for when Bell runs the line test.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 20:11:08 EDT

wayjac :

said by dslrocker3 :

i wonder what happened now... no dial tones on my phones... related to the dsl work?... (never mind got my dial back now... that was strange...)

Your phone and dsl share the same wire pair?

said by dslrocker3 :

I was curious if this modem would sync at the ful 800Kbps uplaod speeds profile. The 5200 E242 also synced at 736Kbps.

The upload thing is one quirk of the ikns :D

Do you have any logs from the 2wire when it would not sync?
I think that is very odd and would like to know more about this when the "honeymoon" is over

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 19:51:28 EDT

dslrocker3 : i wonder what happened now... no dial tones on my phones... related to the dsl work?... (never mind got my dial back now... that was strange...)

Yes 5056Kbps/800 is the top speed for my ISP. I hooked up a Speedstream 5200 E242 and it reported an attenuation of 14dB. I didn't really to reset the connection in case the phone company thought that there was a problem with the line and then refuse to increase the speed profile but at the same time, I was curious if this modem would sync at the ful 800Kbps uplaod speeds profile. The 5200 E242 also synced at 736Kbps.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 19:48:25 EDT

wayjac :

said by dslrocker3 :

I'm paying for speeds up to 5056/800 Kbps

So 5056/800 is the top speed for your isp?

said by dslrocker3 :

Approximate distance from exchange: 0.76 km. :)

Yes you went from one extreme to another

said by dslrocker3 :

I think that that the speedstream 4200's underestimate the downstream attenuation by about 3dB though, based on comparisons with other modems including a Speedstream 5200 E242 (same chipset).

I think each modem is a "individual" each has its own opinion of what the stats are

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 19:32:14 EDT

dslrocker3 :

What's the highest sync speed your isp allows

I'm paying for speeds up to 5056/800 Kbps although some people have been mistakenly put on 6016/800 because that is a speed profile that exists on Bell's internet offerings (Bell owns the copper wiring and RDSLAMs in this area).

Approximate distance from exchange: 0.76 km. :)
I think that that the speedstream 4200's underestimate the downstream attenuation by about 3dB though, based on comparisons with other modems including a Speedstream 5200 E242 (same chipset).

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 18:00:15 EDT

anon : IKNS tends to always report interleave, even when on fast path.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 17:37:40 EDT

wayjac :

said by dslrocker3 :

But I don't want to start messing around quite yet because Bell Canada might see every reset/reboot/loss of sync as a sign up instability,

I think that's reasonable

said by dslrocker3 :

especially when i currently have another ticket open to boost the profile speed.

You'll have no problems with sync speeds, too bad the 2wire won't sync I'm thinking your max rate is "pegged" at 10000kbps

What's the highest sync speed your isp allows

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 17:15:21 EDT

dslrocker3 : that will come some soon enough, for educational purposes and to provide information to the fellow users on this board.

But I don't want to start messing around quite yet because Bell Canada might see every reset/reboot/loss of sync as a sign up instability, especially when i currently have another ticket open to boost the profile speed.

Maybe in a week, I'll try to do some comparisons (I think that Bell keeps monitoring a line for 48-72 hours when there is an open ticket).

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 17:12:42 EDT

wayjac : Have you checked the line stats at the nid
Don't forget that you have abnormal inside wiring :D

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 17:10:37 EDT

wayjac : The telco can set the "Interleave Delay" to a value they prefer

Most users get upset when they see Interleave, never checking too see if there is any real delay

It seems like interleave (with no delay) is a default setting for adsl2/+

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 16:48:58 EDT

dslrocker3 : this is interesting....

My modem's log reports:
|Link up 1 US 736 DS 3008 (INTL:G.dmt)

but the Sentry Siemens Speedstream 4200 Utility reports:
Interleaved Delay 0ms

also command prompt reports:
Pinging gta.igs.net [216.58.97.17] with 32 bytes of data:
Reply from 216.58.97.17: bytes=32 time=18ms TTL=57
Reply from 216.58.97.17: bytes=32 time=14ms TTL=57
Reply from 216.58.97.17: bytes=32 time=15ms TTL=57
Reply from 216.58.97.17: bytes=32 time=14ms TTL=57

I don't know but that almost seems like fast path to me.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 16:18:26 EDT

dslrocker3 :

said by Bicephale :

Ah yes, this is better! You'll be upgraded to 5056/800 Kbps FastPath
soon enough and then this case will be solved to your satisfaction...

Comgratulations!!!

maybe 5056/640 fast path (is there such profile?). the funny thing is the 5.5 dB attenuation on the upstream but only a 736Kbps sync. IKNS strikes again!!! Did the Lucents also have this same problem on the upload side?

30 minutes for sync? that's okay, like I was saying before, my 2wire 2700HG-B refuses to sync at all on my line now. My speedstream 4200 takes about 2 minutes to sync but now also takes about 3 to 4 minutes for pppoe authentication.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 16:12:32 EDT

anon :

said by wayjac :

I was reading a thread about a modem taking up to 30 minutes to sync with a ikns

Makes me wonder if this is modem brand dependent. Even different brand modems reports stats differently with ikns. Some seem more affected than others.

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 15:57:31 EDT

wayjac : That is a very big change in attenuation being on a very short loop can have its drawbacks

I was reading a thread about a modem taking up to 30 minutes to sync with a ikns

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 15:41:20 EDT

Bicephale : Ah yes, this is better! You'll be upgraded to 5056/800 Kbps FastPath
soon enough and then this case will be solved to your satisfaction...

Comgratulations!!!

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 13:48:56 EDT

dslrocker3 : this tells the whole story!

upstream attuation is 5dB, snr 6.0dB, hybrid number is still 1.

the tech told me that his tests show that my line is now easily capable of over 11Mbps.

by the way, bichephale, I suspect that you are right. I think Bell just did this because it was easier than removing the bridge tap and otherwise fixing my wiring! they probably got sick of hearing from me too! hehehehe.

Unfortunately, my 2wire modem now refuses to sync now that the technician has completed his work and based on the upload sync, I think it's fair to blame both on the following and for me to yell out "damn IKNS!!!!"

lol

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 13:11:50 EDT

Bicephale : Perhaps you'll get hooked to that RDSLAM, eventually: otherwise Bell
would need to fix the wiring so they may be tempted by an easy way
to solve the case...

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 13:01:55 EDT

dslrocker3 : AARRRGGHHH!!

The technician just phoned me and even though Bell told Velcom that RDSLAM was availalble for me, the tech's ticket says nothing about it. He says that he can see it as available but can't hook it up unless he gets authorization. He says he is going to check for me but how did the paperwork he receive not say anything about the RDSLAM?

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 12:55:47 EDT

Bicephale : Compared to this i'd say you're on a diet, looking at the yellow middle graph:

Re: Help/Advice with my Line Statistics please

Sun, 04 Oct 2009 11:36:45 EDT

dslrocker3 : Interestingly, the gap at tone 109 has dissappeared overnight resulting in a 1.5-2.0dB rise in signal to noise ratio.

I'm still crossing my fingers for today's technician dispatch that was assigned supposedly to hook me up to a remote.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 20:28:34 EDT

dslrocker3 :

said by Bicephale :

Maybe someone wants you to remain quiet until Sunday...

:p

hahaha, you are trying to make me paranoid. why would they be trying to keep me quiet until sunday? to give away all of the remaining spots? with Bell, anything would be possible.

someone phoned me earlier and I think it might have been velcom but i couldn't hear the person so I'm not sure if it was them. oh well, i'll just have to wait until sunday.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 20:03:08 EDT

Bicephale : Maybe someone wants you to remain quiet until Sunday...

:p

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 19:53:38 EDT

dslrocker3 : it is great news if it gets done but I'll believe it when the work has been completed. Then again, why would Bell schedule another dispatch and say the rdslam is aviable. that wouldn't make sense and it would be just as easy for them to just say that ticket/issue was already resolved and just do nothing again.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 19:50:34 EDT

Bicephale : Great news! When this is all over i guess it should be
a somewhat less frustrating situation to deal with!...

:D

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 16:11:29 EDT

dslrocker3 : I just got a message from Valery over at Velcom and Bell is no longer denying the existance of the RDSLAM.

In fact Bell has phoned Velcom and say the remote is available for me and expect it to be done on Sunday afternoon. I can only hope for the best.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 13:51:27 EDT

dslrocker3 : what doesn't make sense is how ALL of the errors have dissapeared for nearly an hour.. You think i would still get some errors with ah noise margin of only 9dB.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 13:50:05 EDT

Bicephale : The regularily spaced notches are still there, no good!

:(

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 13:32:46 EDT

dslrocker3 : lol.... opened another ticket to insist that I get hooked up to the rdslam, and guess what they did? changed the profile from 4032/512 back to 1728/640 - wait now connect back again at 2176/640 (guessing they put it back to 2496/640?) edit - and now back to 1728/640. I guess I waited to long try to get those higher speeds while tweaking. If I didn't know better, I would say that you might have been implying that I should do a bunch of stuff to mess with Bell's readings just to confuse them and make their job harder. I know you really meant the opposite but for once, it would be great for the customer to be able to mess with Bell. LOL, 2496/512 with a downstream noise margin of 0.9dB would have been funny if they saw that.

lol... now on fast path.... any difference that anyone can see from the screen capture other than the the error correction being turned off? (compare to speedstream captures from earlier)

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 13:26:07 EDT

Bicephale : I concur, today is no good day to start tweaking! Imagine what
Bell's people could do when they'll read your logs where the SNR
Margin drops next to zero! The magic word here was "before"...

:p

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 13:03:55 EDT

dslrocker3 :

said by Bicephale :

But honestly, i think you should have explored some of your other
options before this, relatively to RF Choking and Spectral Shaping

Bicephale, I know you're right. It's just hard to know where to start because its something that I know nothing about.

the other factor at play here is the fact that I know Bell is sniffing around right now and forcing numbers that don't make sense to them might make them to slow me down further.

yes, i did ask the technician to remove the bridge tap but he said that had to be a different person to do it (could have been the truth, i don't really know, could have just been yet another excuse). It's too bad that I don't know where the bridge tap is. Some wire cutters and ladder would do the trick!

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 13:01:03 EDT

dslrocker3 : Ironically, as much as I dislike the dlink dsl-2320b, it actually outpeforms my 2wire when my line gets like this.

using dsl-2320b:
first image is while using normal setings

second image - tweaked settings, target noise margin turned all the way down.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:56:36 EDT

Bicephale : The D-Link will support Spectral Shaping if you dare play with it
or you might as well insist that Bell gets that "Bridge Tap" cut...

:p

But honestly, i think you should have explored some of your other
options before this, relatively to RF Choking and Spectral Shaping
i mean...

:huh:

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:50:49 EDT

dslrocker3 : unfortunately, I don't have a speedtouch any more. However, I do have this peice of crap d-link... who knows... maybe I should just see what happens... can't hurt right? besides, if bell looks at my line right now at 2208/512, they will lower it regardless so might as well play around.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:49:17 EDT

Bicephale : But you have a SpeedTouch 5x6, right!?! That means you can force
the MoDem to negociate a connection at a lower BitRate so why not
just ask for 3008/640 Kbps Interleaved and deal with it yourself, at
your own pace when ready?...

;)

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:44:08 EDT

anon : Don't feel bad dude. This Bell tech had the same problem trying to keep his profile up past the OPI into the F2.
»[Internet] Bell tech with DSL problems at home

If he can't do it then what chance do you or any of us have?

Oh I forgot, divine intervention or win the lottery and have fiber strung to your house.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:42:51 EDT

dslrocker3 : I just spoke with Velcom and the person was a DSLreports user! :)

I explained to him again how my neighbor is on the Remote and we both seem to agree that Bell is playing hardball and we both agreed that we just can't prove it.

He said he would excalate. Who knows if this will work this time but I suspect they will just lower my speed profile again when they look at the line.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:32:18 EDT

Bicephale : I'd say you've lost more chunks than meets the eye at 1st glance:

Before

2nd Try

Bell's worker should have put you on 2496/640 Kbps Interleaved but
he made things worst (trenches) or perhaps you need to test at the
demarcation box again...

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:20:47 EDT

dslrocker3 : The technician didn't say anything about the upstream other than it was necessary because I couldn't maintain the 640/800 Kbps upload (purely based on his 104% measurement of the upload rco@640 - lie)

Bichephale, if you notice before I opened the ticket to ask for the RDSLAM, the trenches weren't there. here's the thing with my line - it will be without those trenches for a few months but then the trenches will re-appear for about 2 weeks.

when there are Trenches - speedstram syncs at 2200/2300, 2wire at about 1600Kbps download

when there are no trenches- speedstream 2496@8dB downstream, 2wire 2496@10dB max attainable around 2900Kbps

I fully expect that the shaping will get better on it's own and I'll connect at about 2.8/2.9Mbps using my 2wire modem within the next few weeks (if they don't lower the profile).

also, this current sync rate of the 2208/2240 neighborhood are the exact same numbers I got on the dowstream last week on the 2496/800 profile when things went screwy. but now, i've lost a chunk of my upload speed.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:14:12 EDT

Bicephale : Oh, and lets put them together for convenience:

Before

1st Try

2nd Try

It strikes me now, i'd say your UpStream Attenuation is off-scale!

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 12:08:16 EDT

Bicephale :
Well, i do believe what Bell's worker told you when he spoke about
a "Bridge Tap" all right: your spectral response graphs show deep
trenches occuring at strangely regular intervals while the overall
shape is so atypical it only makes sense!...

Did he say why you needed a 17 dB UpStream SNR Margin, by the
way?! Other than to make room for your DownStream DSL signal, i
mean...

Also...

Would you please edit 'DMTv9.01.ini' to remove sensitive data in the
"Login" and "Passwort" fields under the "[LOGINDATA]" section (top)
and send it to me? There's something wrong with 'DMT v9.01' which
must be corrected once and for all!...

DSLRocker - SS4200 vs DMT (2009-Oct-2 10h55)
DSLRocker - SS4200 vs DMT (2009-Sep-27 20h19)

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 11:26:41 EDT

dslrocker3 : I guess I'll have to try something new like RF choking.

you didn't believe anything that tech told me either, do you?

Who wants to place wagers that Bell will just lower my profile again? started at 2496/800 to 1728/640 (even though that made the downstream noise margin go up to 17dB on this profile and it was more than fine at 2496). Then something bad happened last night while the ticket was opened and I could barely get 1728. but then they go ahead and change it to 4032/512 (that change is funny because my line can't come anywhere close to that) to what next? perhaps 1728/512? I think that the change to 4032/512 is more of indication than anything else of what Bell expects this line to be capable of but which is it clearly isn't when they take measurements at the NID.

I think that it is also fair to say that when you have bridge taps messing around with things, your choice of modem becomes so much more important and you have to get just the right one that is the best match for your line. The frustrating part about the bridge tap is the fact that the signal to pick up higher speeds is probably there - my modem proabably just can't see it!!!

it would be great to compare what this technician told me in person to what Velcom tells me. I fully expect to get an email or phone call from Velcom that probably says that I have a defective modem or that there was no problem found. anything claiming a defective modem or problems with my inside wiring are b.s. because he didn't check anything inside. and if velcom's repsonse to me is that Bell found no problem, well... there obviously is a problem still!

DMTdiag.zip 59,771 bytes

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 11:25:10 EDT

Bicephale : Euh...

When we look at the shape of your spectral response graphs
it doesn't feel like the ticket should be closed at all: there's
still a "Bridge Tap" messing up your DSL signal! No one here
would believe that getting moved to a remote/stinger can't
help with a stronger DSL signal neither.

The presence of a bridge tap being reinstated once again do
you plan to explore your options with RF Choking, just in case
it might affect impedance matching in a positive way?...

:p

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 11:02:04 EDT

dslrocker3 : I'm not liking this. Once again, I was told 100% that there was a bridge tap on the line but this time I was told that switching to an RDSLAM wouldn't help because of that. I'm not sure whether to believe that.

Either way, when you want to raise the profile and your line is capable of that, they won't do it. When it was fine at 2496/800, they lower it to 1728/640 but when they shouldn't raise your profile, they go ahead and change me to 4032/512? That doesn't make any sense! He seemed very polite and nice to talk to but I'm pretty sure I caught him in a lie. I questioned him about the 512Kbps upload profile and he said that he had to because the upload occupancy was at 104% before (look at the screenshots from before for the upstream).

I did ask the technician about whether I would have better luck if I signed up to Bell Internet when it came to getting full speed, and take one guess at what his answer was?

The technican concluded there was nothing he could do about it except put in a request that re-cabling be done which might not happen for months. He did say that he was able to get a full 5Mbps at the "box" (he probably is talking about at the OPI where it starts going underground) but fully admits that the line is only capable of being on a 1728Kbps profile when he tested at the demarcation point. What a surprise here - he did not believe me when I told him that test results are always worse at the NID.

so the end result, things are just worse now after the ticket has been completed

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 09:58:51 EDT

dslrocker3 : just spoke to a technician on the phone and my gut feeling was correct about the how the line changed characteristics, lower attainable rates but less errors, speedstream now works better than 2wire (oppposoite as before)... The technician admits that changes were made at the CO/their end, just as I had thought also partly based on how my 2wire reported a signal interruptiion when the change was made.

connection down for about 30 minutes after technician contacted me. back up at 2240/512. I'm pretty sure they set the profile to 2496/512. The tecnician hasn't phoned back so I hope they are still looking at the line.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 04:00:49 EDT

Bicephale : I'll bet on whichever unit provides the best DS SNR Margin 1st,
hoping that its noise performance won't ruin that advantage...

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 02:09:53 EDT

dslrocker3 : I have Dmt 9.01 running in diagnose mode right now. But based on the 2wire diganostic pages above where the connnection drops and reconnects at a lower rate, I'm certain that Bell changed something at the central office. The way it is now seems to yield far fewer errors, even when the occupancy is very high. However, the possible sync rates are lower. This reminds me of when I had problems about 5 years ago when I had a technician say that the line was very clean after he made a change but I complained the he actually made my sync rate worse.

For example, 0 errors on the line of any kind over the last 30 minutes since I rebooted the modem. This is even though the noise margin is only at 10dB (1728). But when I would have a 2496 downstream sync with a noise margin of 10dB, I would sometimes get a lot of errors.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 02:02:34 EDT

Bicephale : Good. Here's a closeup version for future reference...

DSLRocker - SS4200 vs DMT (2009-Oct-2 1h13)

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 01:44:33 EDT

dslrocker3 : I think this last page or two of this thread clearly how one modem can be better for one line but a different modem for a different line... and even how different modems can work different for the same line when conditions change!

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 01:40:05 EDT

Bicephale : Great, this utility will complete 'DMT v9.01' beatifully:

Trev's SNMP Utility v0.1.0.9 (2007-Apr-8) for the Siemens SpeedStream 4200

...not to mention the 'DMT' "Diagnosis" mode requires little configuration at all.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 01:34:20 EDT

dslrocker3 : ... further proof that the speedstream 4200 works better than the 2wire 2700HG-B when the line goes to crap.

so i guess I need to keep both modems in case the line changes to the other characteristics.

The Speedstream 4200 that I swapped with my friend is now "unBelled" so I am able again to run DMT again.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 00:27:53 EDT

Bicephale : This was specific to the 2Wire:

OrbMT on D-Link DSL 2320B, Bicephale, 2009-Feb-22

...and this was for another brand/model:

From the ground up!, Bicephale, 2009-Feb-22

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 00:12:33 EDT

dslrocker3 : that's right router stats.. I haven't used that in so long!!!! I'll have to figure out how that works again. I do remember that I had to use the basic statistics because the advanced MDC pages couldn't be read by routerstats.

Re: Help/Advice with my Line Statistics please

Fri, 02 Oct 2009 00:10:24 EDT

Bicephale : Are you running 'RouterStats' in hope to capture Bell in action?

Re: Help/Advice with my Line Statistics please

Thu, 01 Oct 2009 23:01:05 EDT

dslrocker3 : Gee, thanks Bell!!!
Notice the reason for loss of sync... signal interrupted????!!! If you know the 2wire modem well, you know when you lose sync because of problems, it usually just says stuff like LOS limit, or Error limit. but Signal interrupted? that's almost like someone unplugged something and that's the exact same message I got when i noticed a similar degrading of the line 2 weeks ago. I was watching my modem page and this happenned and it reported an "infinite attenuation" - that sounds like a great thing (being sarcastic).

hopefully, work is being done but I'm not very optimistic about a positive outcome.

Re: Help/Advice with my Line Statistics please

Thu, 01 Oct 2009 21:29:51 EDT

dslrocker3 : wish me luck.... I am hoping to get onto the same IKNS RDSLAM that one of my friends got connected to some time on Tuesday or Wednesday. I'm not very optimistic though because they have already started the usual profile "adjustments" (read as slowdowns).

I have notice a 1.5 dB drop in attenuation on the the Speedstream 5200 and about a 0.3dB drop in attenuation on the 2wire 2700 since they changed my profile from 2496/800 to 1728/640. I suspect that might be because of the the upload speed "shaping" the connection differently.

I would capture some screensheets with my Speedstream 4200 except I gave that one to that firend who is hooked up to the RDSLAM. He needed a modem to use and I wanted my 2wire back so I had to leave him something to use.

A technician is coming in the morning. All I can hope for is a good one this time. However, I think it is more likely that I will get the usual nonsense.

Re: Help/Advice with my Line Statistics please

Mon, 28 Sep 2009 11:10:23 EDT

dslrocker3 :

said by vintagewino :

You can get some clip-on ferrites (various sizes) at Sayal Electronics. Depending on the size ~$2-5 each. They have a reasonable selection. I like the snap-on type, as you open it up, lay or wrap the wire, and snap it shut. The toroids are more fun in threading the wire through the loop. More interesting if there's a plug on the end of the able.

»www.sayal.com/

Note certain of your location, so for Toronto, Gordon Baker & Victoria Park, Vaughan, Creditstone near Rutherford, Mississauga Matheson near Dixie.

thanks for giving me an option of where to buy. The Matheson location would easily be the closest for me.

I'm in the Islington&Gardiner area. It is a very good area for RDSLAMs and there one on and new ones popping up every few months and are now almost every street corner with a traffic light. Yet, I can't get connected to one not matter what.

Re: Help/Advice with my Line Statistics please

Mon, 28 Sep 2009 09:03:25 EDT

vintagewino : You can get some clip-on ferrites (various sizes) at Sayal Electronics. Depending on the size ~$2-5 each. They have a reasonable selection. I like the snap-on type, as you open it up, lay or wrap the wire, and snap it shut. The toroids are more fun in threading the wire through the loop. More interesting if there's a plug on the end of the able.

»www.sayal.com/

Note certain of your location, so for Toronto, Gordon Baker & Victoria Park, Vaughan, Creditstone near Rutherford, Mississauga Matheson near Dixie.

Re: Help/Advice with my Line Statistics please

Sun, 27 Sep 2009 23:57:43 EDT

dslrocker3 : I haven't done any experimenting with that. I'm hoping that some of my friends' computer power supplies will fail so that I can take the part from those. I'm not really hoping that they problems with their computers but I'm hoping to spend the least amount possible for my experiments (saving money is always a good thing). Ebay might be my friend when it comes to that.

In particular, that extension cord that gives a downstream improvement but moreso when the line is doing well is still a curious element. In these recent "down" times, unplugging that extension cord made less of a difference. Ironically, the 2wire 2700HG-B and the DLink 2320B numbers were almost the same when during this "down" spiral (haven't checked today because I can't get the d-link to work reliably!) but based on my experiences, my 2wire modem is going to be far better for me than anything with a broadcom chip when I an on the "up" time/better conditions. There might be something fishy going on with that extension cord. But even it's going to hard to find another cord that does makes the line improve like that (modem is not connected to this extension cord, nothing is). I think I should focus on this factor because even if I get another cord that improves the line, that still doesn't explain why it does.

Also, the phone cord factor (the black ones originally from the Speedtouch Homes) wasn't there during the recent worsening of conditions. If you remember, these black cords were quite a bit better for my line when connected to my modems during times of the higher download sync performances. That wasn't the case when I had these recent difficulties (made little or no difference which cable).

Side note: with any modem that I have tried with these Broadcom chipsets (including speedtouch 516, TD-8841, DSL2320B), they all gave rediculously low sync rates whenever the unit is first powered on. what's that about? I just tried the dlink modem after everthing on my line has normalized and this modem does not like the current characteristics of the line. It gets a 1440/800 sync rate while my 2wire unit has a maximum attainable downstream rate of 2812Kbps. Wow! I've been critical about these Broadcom based modems before, and while really liking the tweaking abilities and the feature set, I can only now charactertize these types of broadcom-type modems as peforming okay on some lines but awful on others.

Re: Help/Advice with my Line Statistics please

Sun, 27 Sep 2009 23:07:59 EDT

Bicephale : I'm thinking perhaps it's not as much a problem of the line being
a "good" or "crappy" one but of finding the right combination of
line vs DSL device while the impedance matching conditions may
happen to be varying because of external elements. Did you give
RF Choking a thought lately?... I suspect RF Chokes can affect
the impedance and hence contribute to impedance matching in
presence of a Bridge Tap, under certain conditions. I wonder...

Re: Help/Advice with my Line Statistics please

Sun, 27 Sep 2009 22:02:30 EDT

dslrocker3 :

said by dslrocker3 :

I haven't checked yet but I'm almost sure that if/when I rehook up my 2wire modem, it will return to the normal pattern of being beter than the speedstream. (2wire having a rx snr of a good 2 to 3 dB higher at 10 or 11dB).

Confirmed....

for whatever reason, my 2wire clearly does better on the downstream side on my crap line when the line is working properly

On this same crap line but when there are problems, the Speedstream handles things better.

So my conclusion, at least for my case, is this:
-crap line, 2wire 2700HG-B better;
-even crappier line, Speedsteram 4200 better

Re: Help/Advice with my Line Statistics please

Sun, 27 Sep 2009 20:30:15 EDT

dslrocker3 : It's a good thing that I didn't open a trouble ticket because I'm certain Bell would have fixed nothing and simply "1184/640ed" me.

Much better.... rx snr fluctuation between 7 and 8dB... and I don't think that it is very likely that it is anything on my end that was causing the problem (although it wouldn't be impossible). I wasn't even home when the line returned to normal. I was at the Jays game all day and got home just 30 minutes go to see that my modem decided to resync at a higher speed. Since this last round problems started 2 weeks ago, my 2 wire has disliked what was happening to the line more than my speedstream modem did. I haven't checked yet but I'm almost sure that if/when I rehook up my 2wire modem, it will return to the normal pattern of being beter than the speedstream. (2wire having a rx snr of a good 2 to 3 dB higher at 10 or 11dB). It really does seem like a line with a split personality, doesn't it?

I know that it might not really matter how my line is doing to some of the readers on the forum here, but I find it interesting to look back at the patterns that have occured in the past. This 2 week problem window is almost identical to the last time I open a trouble ticket (no ticket opened this time) which also last about 2 weeks. The numbers were very similar too if I remmber correctly. The really do believe that Bell has 2 different ways of connecting my line - one that gives one set of numbers (around 1800 to 2200 rx sync, attenuation of about 0.5 to 2dB lower) and another way that gives me the 2496/800, but with attenuation about 1 dB higher. These past 2 weeks show me that different modems act differently on different lines and that if you have a bad line, you need mulitple modems because when the line changes it's "personality", the other modem might act better.

D-link 2320=junk, garbage, waste of money
I'm almost ready to throw this modem in the garbage. I don't even dare sell it in fear of geting bad feedback on my Ebay account, unless I can find better firmware for it. With the original, out of the box firmware, the computer and router couldn't automatically assign an IP address to the modem. Then I found out that by default, the modem enables certain vpi/vci settings but when they are active, the setup wizard simply fails and refuses to continue. I flashed a newer firmware and now the autodetect wizard works fine but I sitll cannot get it to work behind a router. I will work for a few minutes at a time but then my router will change the MODEM's IP address to that of the ISP's address, resulting in the modem becoming non-operational. I was reading something briefly in this forum and someone said something about the D-link has a problem with DHCP and often fails after 30 seconds. (i noticed this in move auto dhcp mode and even when setting a static ip). Pure crap is what this modem is!

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 19:03:26 EDT

Bicephale : I wish you this new quest to bring some
more satisfaction than previously.

It will be a pleasure to follow your thread
and to comment about it when suitable...

:D

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 17:16:30 EDT

dslrocker3 : tweaking time!!!!!!

pccanada's closing down sale had the TD-8841 for $25 but they were sold out. :( They had the D-link 2320B for $50 but I went for it anyways because if Pccanada's store is very local to me (walking distance). If I was to pay the normal pricing of $35 from another Toronto-based store, I would have had to pay for bus fare, gas/parking, or shipping so I made the purchase anyway.

Unless you know how to manually set up a static IP address, I wouldn't recommend getting this modem (D-link 2320B)! I couldn't figure out why I could only get the modem to work in bridge mode even after a factory reset. For whatever reason, I had to manually configure my network card's TCPIP 4 or had to manually assign the IP address and default gateway in my router for the d-link modem.

Bichephale, you'll be happy to know that from looking quickly, it appears to have all the same features as the TD-8841 but there is only 1 ethernet port! I think that you might be curious about that from a noise standpoint.

Time to play with it... Initially, out of the box, the d-link is syncing at 1856/800 while my Speedstream 4200 is syncing at 2240/800. I know that the d-link sync doesn't sound great but it is at least almost identical to what my 2700HG-B is doing currently. At least I can "customize" my connection.

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 09:41:32 EDT

Bicephale : I guess what Velcom means is that they
just won't request a specific profile and
it's no surprize because i believe it's the
very same story with all 3rd-party ISPs:
otherwise you'd only need to switch!...

Did you try to ask for advice from Bell's
people, in their forum? They got control
over your phone-line, after all!... Maybe
i was luckier than the average customer
when i needed help before but why not
send a few private messages, in case it
might be your lucky day? DeadPool is a
friendly guy, if he doesn't know what to
do for you then i'll bet he can still point
you into the right direction...

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 08:28:39 EDT

dslrocker3 : Good morning Bichephale,

Velcom tells me that they can't request specific profiles. I know that's B.S. because they are the customer of Bell but that's the story that they keep to.

However, opening trouble tickets is very "dangerous" when it comes to Bell. I think that we both know that Bell usually ignores what it says on the trouble ticket and does the "nonsense" (no show techs, defective modem stories, telling you to format your hard drive, tech showing up but then leaving after a few minutes saying they will look at the C.O. for problems but probably do nothing, or reducing profiles even if your line is perfect) 90% of the time. So I really see that as a 90% chance that Bell will slow me down... I'm not so sure how lucky I'm feeling today. ;p I'm pretty sure that they only way that I'm going to be able to get Bell to let me "play" with different profile would be if I was able to speak to someone directly at the Test Centre but that can be difficult because even if they do call you, which they don't always do, they don't allow you to phone them back if you missed the call.

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 07:51:39 EDT

Bicephale : If at least you could be allowed to try
a more appropriate profile in order to
experiment with it. The ST5x6 can be
forced to connect at a lower BitRate
if necessary. It would make sense to
ask for 3008/512 Kbps Interleaved as
the additional stress put on your line
would make changes more obvious as
you roam around trying to fix things...

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 07:40:50 EDT

dslrocker3 : still not a good line for this day and age but not as tragic as yesterday!

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 00:26:00 EDT

Bicephale : Yet, it must feel particularily unfair
when you see such 'DMT' captures:

Is a higher profile worth the low SN ratio?, SammyJ, 2007-Aug-30

:uhh:

Re: Help/Advice with my Line Statistics please

Sat, 26 Sep 2009 00:01:41 EDT

dslrocker3 :

said by Bicephale :

Your patience/tolerance are exemplary!

Bell fixing it? hahaha!!! I even gave one of the techs blue jays tickets (not expensive ones and plus they were extras that I couldn't use that day) and still nothing got fixed.

Patience/tolerance? It's not like I have much choice. Here, my choices are pretty much Robbers or DSL and the first choice isn't unlimited. Besides, a friend of mine who lives one street over keeps complaining how Rogers goes down for days at a time. While DSL around here may be slow, at least it works.

Re: Help/Advice with my Line Statistics please

Fri, 25 Sep 2009 23:42:29 EDT

Bicephale : It's really a tough case, maybe there's
a Bridge Tap but it just happens to be
too close to be seen. Euh... I wonder.

...and even if there is a Bridge Tap, it's
not clear Bell would be quick to fix it...

:huh:

Your patience/tolerance are exemplary!

Re: Help/Advice with my Line Statistics please

Fri, 25 Sep 2009 22:58:18 EDT

dslrocker3 : The hybrid number is still at 1. It's too bad that DSL is the only way to get reasonably priced unlimited downloads.

Re: Help/Advice with my Line Statistics please

Fri, 25 Sep 2009 22:28:09 EDT

Bicephale : Oh, it's even worst than i imagined...

So, how 's the "Hybrid" reading now?

Re: Help/Advice with my Line Statistics please

Fri, 25 Sep 2009 22:03:17 EDT

dslrocker3 : a cruel joke, huh?

Re: Help/Advice with my Line Statistics please

Fri, 25 Sep 2009 21:55:14 EDT

dslrocker3 : okay, here's what's going on with my line that has multiple personalities....

2 weeks ago, I noticed somthing crazy on my line when using my 2wire modem. It was always syncing at 2496/800 for at least the last couple of months consistently but out of nowhere, something happened.

it would drop down to 2300/800, then 2200/800, 1900/800 and various speeds in between. Here's what I found interesting: Usually, when you lose sync because of instability or even because you unplug the phone cable, the 2wire usually reports reasons such as "error limit exceeded" or "LOS limit". But when this all started, I started seeing an explaination that I have never seen before that said something similar to "signal interupted". It would later say "retrain higher" and connect at 2496/800 for a few hours but would then say "signal interrupt" again and connect at a slower speed. But whenever, the slower connection occured, the attenuation would also get lower! It's almost like someone is turning on the "bad switch" or turning off some type of amplification at the CO.

Ever since this started, my speedstream 4200 has been providing better line stats than the 2wire 2700 modem. (even though it was reversed before).

So now for the past week or so, 2wire is connecting was connecting at about 1900/800 and the speedstream about 2200/800.

now today, my 2wire is connecting at 1632/800 and my speedstream at 1760/800 (prfoile is still set at 2496 according to the 2wire unit). Strangely, the slower the sync rate, the lower my attenuation gets. Also, even though the conditions are worsenning based on the connection is getting slower, the errors (fec, crc, etc.) on the line are less frequent!

so yes, here's my situation in a nutshell... sync rates are getting lower and lower (profile is the same) but attenuation and errors are also getting lower... which is usually the reverse of what should happen!

I don't dare open a trouble ticket because I know Bell will never repair this line. They'll probably just stick me on 1184/640. I'm just hoping that things just magically improve like they always have in the past.

Re: Help/Advice with my Line Statistics please

Mon, 03 Aug 2009 04:46:50 EDT

Bicephale : I'm sure those few inches don't really matter, twisted pair or not!

:D

Re: Help/Advice with my Line Statistics please

Mon, 03 Aug 2009 01:07:48 EDT

dslrocker3 : LOL, if the Bell worker who installed the NID wanted to do me a favor, he wouldn't have simply cut the old corroded drop wire that has insulation that is flaking off and hooked that up to my side of the NID! It's not like he wasn't already in the basement anyway to install the grounding wire!

btw, what in the world was i thinking in that picture when I wrapped a wire around in a counter-clockwise direction? It must have been the lack of sleep from all of the DSL nightmares that I've had.

Re: Help/Advice with my Line Statistics please

Mon, 03 Aug 2009 00:58:13 EDT

Bicephale : I'm sorry but i must say no twice.

No, the indoors station protectors stopped acting as such
the minute their carbon elements were removed, as shown:

No, Bell did you no favour! If their worker had wanted to
do a clean work he would have used an IDC bloc as this:

Euh... Or something with fewer circuits, hopefully, as this:

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 22:52:26 EDT

dslrocker3 : It would have stopped being a protector once I removed the rusted ground wire that was causing noise in the phone line....

Wasn't it so very nice of Bell to provide me with a second "ex-protector" so that I could install the filter in the basement?

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 22:23:13 EDT

Bicephale :

Hi,

Great and thank you DSLRocker! Now lets put them together:

Ah, well what you called "protector" doesn't protect anything anymore:
these are just "junction boxes" and your station protector has moved
outside... I've included a revised version which i might have posted at
1st if i had known better... There still are details missing, of course.

DSLRocker's Wiring (2nd approximation)

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 21:14:43 EDT

dslrocker3 : bicephale, you're more of an artist than I am but mine is more probably a more accurate represenation of my inside wiring. You will see that the strange wiring goes to my modem. The extension cord is the key to everything. Without it plugged in at the first floor, the improvement doesn't happen!

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 16:09:56 EDT

wayjac :

said by dslrocker3 :

my modem thinking there was an unfiltered phone even though could swear that there was no phone attached....

If that's a referance to the modems warning to check filters and alarm system.........
The 2wire modem can give that warning at any time even when a whole house filter is installed

warning to check filters and alarm system

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 16:00:59 EDT

dslrocker3 : My bad... about my modem thinking there was an unfiltered phone even though could swear that there was no phone attached.... It appears that unplugging the power for a cordless phone isn't good enough. I don't know why I didn't think of it before but last night, I actually unplugged the phone cord from the cordless phone. I learned something. DUH!!!!! Yes, it's time to laugh at me for this one!

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 14:03:46 EDT

wayjac : Why not, I would remove all the gray text before submitting the art ;)

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 07:03:36 EDT

Bicephale : Hi DSLRocker,

Please tell me, be honnest: do i have a chance to make it as an artist?

DSLRocker's Wiring

Re: Help/Advice with my Line Statistics please

Sun, 02 Aug 2009 01:25:50 EDT

dslrocker3 : ok, the fact that my modem was complaining about unfiltered phones even though I had no phones connected was really bothering me. The problem would fix itself if I simply re-inserted the centralized filter back into the basement to filter the wiring itself from the modem.

Using a jack upstairs that was connected to the separation jack in the basement was providing syncs that were no good (yes, of course I remembered to remove the filter when I decided to connect my modem to it).

Today I went and simply removed the wires from the screws/nuts and reconnected them. Now all is back to normal. Now, I have 2 different jacks upstairs that will work properly with my modem (assuming that I leave it unfiltered of course). The only difference being a 0.6dB difference in attenuation because of the balance of the house wiring not being filtered when using one of them.

Tonight, things started improving with my DSL connection even before I started working on the wiring. However, I happy that a wire that I thought that had somehow become bad only a few days after being installed is fine after all.

I know this is nothing new or anything but I just thought I would check to make sure that the strange wiring is still causing my connection to get better:

strange wiring disconnected in basement:
sync 1 2368/800 rx snr 6.0dB
sync 2 2400/800 rx snr 6.0dB

strange wiring reconnected (didn't matter if I connected modem directly to strange wiring or if I used the 100ft cat5, a difference in attenuation was the only difference):
sync1 2496/800 rx snr 9.3dB
sync2 2496/800 rx snr 9.8dB
sync3 2496/800 rx snr 9.9dB

I think that leaves absolutely no doubt that my crazy wires improve things.

With Bell deciding to "flick the good switch" on tonight, I can sleep well without DSL nightmares. It still amazes me that my line has 2 different unique personalities that can change without me doing anything at all to the wiring.

Re: Do you want the phone cops after you Bicephale!!!

Sat, 01 Aug 2009 14:27:56 EDT

Bicephale : Well, one "Bridge Tap" is bad enough, it doesn't help
our sanity to have two and the possibility of dealing
with a defective test module puts ghosts everywhere.

So, i conclude it's a nice precaution to check it! Why
not?!...

:p

Re: Do you want the phone cops after you Bicephale!!!

Sat, 01 Aug 2009 14:09:59 EDT

dslrocker3 : ok, so we don't know for sure but we now want to explore the questinon that the NID test jack might be defective. This shouldn't be so hard for me to test considering that I already spliced a short phone cord with an rj11 on one side to make a connection in the basement. I could borrow that cable for a frew minutes.

Re: Do you want the phone cops after you Bicephale!!!

Sat, 01 Aug 2009 12:57:37 EDT

Bicephale : Hi Ex Bell,

You remember what happened to Johnny Fever at the WKRP transmitter don't you?

Euh... Doctor Fever had all his magic compounds confiscated?

...I have had those test jacks go bad.

So it would be a good idea to proceed as with a vintage station
protector at least once just to make sure the test module isn't
defective!

It would take two of those, actually, one with a filter and the
other not. Should the module happen to be defective this would
be an opportunity to correct the situation, i believe.

Re: error rate

Sat, 01 Aug 2009 12:54:12 EDT

Bicephale : Split cores would be good enough and certainly more convenient, for sure!

:D

Re: Do you want the phone cops after you Bicephale!!!

Sat, 01 Aug 2009 11:12:52 EDT

anon :

said by Bicephale :

I appologize for asking but have you tried this suggestion...which then would lead to this domestic wiring topology:

Remember kids if you wire into the super secret Bell Canada side only of the nid to remove it before a repair call.
You remember what happened to Johnny Fever at the WKRP transmitter don't you? :D
The tech will probably be so offended by your blatant illegal transgression that he will find any way he can to give you a diagnostic.

BTW Bicephale, with the setup in the pic., to test I would remove the leads going over to the test jack.
Also if the person wanted a phone near the modem just take the white orange over to the screws which will filter your phone.

Also I have had those test jacks go bad. I had one that had a high resistance open between the fuses and the port.
Put static on the phone and knocked down the dsl completely which would indicate tip side. Luckily I carry those on the truck.

Re: error rate

Sat, 01 Aug 2009 10:25:25 EDT

vintagewino :

said by Bicephale :

The PC-XT "compatible" power-supplies had a tendency to use
large 1½ inches dia. cores - which makes it easier to wind up to
fifteen turns of ordinary telephone-line cable around the core...

Those PSUs operated at relatively lower frequencies because the
switching circuits weren't as efficient back then, resulting in some
larger ferrite cores. That's the reason why legacy is better here.

:p

Hey Bicephale:
Since Dslrocker says he's around the Toronto area, he has access to Sayal Electronics (2 locations), ElectroSonic, and Active Electronics. I know for a fact Sayal carries split ferrite cores in 3 or 4 sizes. The split ferrite is in a plastic case that opens & snaps shut. The largest one (~ 3x3x3 cm) is something like $5 at Sayal. I believe Digikey carries similar lines.

Rather than treading the wire looping through the ring, open the case, wrap as many turns as you see fit & snap closed.

Re: Help/Advice with my Line Statistics please

Sat, 01 Aug 2009 05:44:02 EDT

Bicephale : Hi,

It would help if you could at least obtain 'DMT' captures using
an equivalent but temporary setup...

As for corroded contacts, i'm afraid you'll have to shop around
in specialized electronics stores - which i suppose shouldn't be
a real problem in a town like Toronto! It would be a good thing
that you do shop around anyway, with toroïds in mind, for
example... While you're there ask if they carry moisture sealant
for electrical contacts!... Better be careful than sorry.

Keep in mind that your unorthodox wiring happens to help since
by some miracle it partially cures an impedance mismatch issue
and then you'll have a different perspective on the situation...

You still do have a problem when testing at the NID with every
telephone wiring but that of the MoDem disconnected, after all!

Re: Help/Advice with my Line Statistics please

Sat, 01 Aug 2009 02:16:18 EDT

dslrocker3 : I have done that before but not recently.

Here's the problem with doing this:
If I did that, the "funny wiring" that enables the faster speed would be filtered, resulting in the slower connection.

and plus, i currently have no hole in the wall that the wire could be fed through

I ended up reconnecting that 100 foot cat5e for use with a voice phone. I guess it doesn't matter that much when it's filtered. Doing this does not appear to have an adverse effect on the dsl line.

By the way, the contacts on my side of the NID below the test jack don't seem very clean. They aren't shiny at all. Is there a good way to clean the contacts?

Re: Help/Advice with my Line Statistics please

Sat, 01 Aug 2009 01:53:59 EDT

Bicephale : Hi DSLRocker,

I appologize for asking but have you tried this suggestion:

Phone jack Wires Question, BryanViper, 2008-Jan-9

...which then would lead to this domestic wiring topology:

Phone jack Wires Question, BryanViper, 2008-May-24

Re: Help/Advice with my Line Statistics please

Sat, 01 Aug 2009 00:29:15 EDT

dslrocker3 : for heaven's sake... I'm going to have more nightmares tonight... ROTFL!

I figured out which wire in the basement that was causing the problem. It wasn't really a problem because it was filtered out/separated from the DSL line but when no phones in the house are connected and the DSL signal is on a completely different wire, the DSL shouldn't be affected so much so that the modem complains that you have an unfiltered phone!

Guess which wire it was!???? It was the 100ft twisted pair cable that I brought in to help! Currently, there are 3 wires going to the protector on the right side that is separated from the one on the left side that is for DSL. All 3 of the wires on the right side feed first floor jacks. That leaves one problem. Since the only wire left that feeds the second floor is the one that is connected to the modem, I would have to connect the phone up here with a microfilter in one of the jacks upstairs. That defeats the entire purpose of the basement filtering!!!!

So, my options are to hook up a filtered phone upstairs to the same jack as the modem OR I would have to reconnect that 100 foot cat5e back up, use that for the voice and just filter it in the basement ignoring the fact that it acts strangely if left unfiltered when there is no device attached. For some strange reason, this 100 foot cat5e cable does not like the basement centralized filtering jack that I installed. yet before the centralization in the basement was done, the 100 foot cable was fine even if I used it to connect the modem to... AARGGHH!

Why me!??? GRRR!!!

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 22:07:00 EDT

dslrocker3 : I think that the other more likely possibility is a problem with one of those wires feeding to the protector on the right (now feeds all wiring except the DSL line). I am speculating it didn't show it's ugly self and provide interference to the DSL signal before because there was a filter on the jack end of it before that countered the effect. Once I did my rewiring, I would have removed that filter on the jack end meaning that it was free to wreck havoc on my DSL line again - except that I had a plan for a it. I filtered it from the source (or at least close to it) in the basement!

I was reading today one person's opinion that the main purpose of a DSL filter is actually to balance the line's impedence. Maybe the fact that there are no filters on any of those jacks any more caused another imbalance in resistance from one of them that only affects DSL signals. Wonderful, even though it doesn't appear to be affecting the DSL signal now that all is filtered, I just know it's going to bother me and give me nightmares until I figure this one out!

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 20:08:21 EDT

Bicephale : Hi,

I'm afraid Bell's unproperly terminated transmission line business
takes you right into mirage land where anything seems possible.

It played tricks on us once, i'm confident it can do that again...

:p

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 17:28:21 EDT

dslrocker3 : On further thought, the fact that the house wiring acts like it is a device shouldn't surprise me since I have a phone cord plugged into a phone jack and the other end makes contact with the rest of the wiring.

If I got a low cost phone junction box that has an rj11 conection at one end, I have a funny feeling that this "unfiltered phone even if not phone is attached effect" might be eliminated because the junction box might terminate the other side of the rj11 connection differently.

I know this wouldn't yield any better of DSL connection results but at least this would provide me more test points (phone jacks) that I could use within my house to test the connection.

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 13:48:15 EDT

dslrocker3 : Now that I think of it, the way that my house wiring appears to have the same effect as a unfiltered phone must be the result of either the phone jack that feeds the rest of the house or the phone cord used to make the connection. This is because of the fact that all of the other wires were there before except for what I just mentioned. So worst case scenario, I could easily put the wiring the way it was before but I kind of like the idea that my SNR doesn't drastically change any more whenever the phone rings. I wonder if connecting the balance of the house wiring using a different cord or different jack will elminate the way the wiring itself provides the unfiltered phone effect even with no phones connected.

I had a theory that I should take another one of the same type of cable (the same as originally provided with the old Speed Touch Home) and connect it at that jack just to see if the modem once again thinks I have a unfiltered phone attached. That theory didn't work as it didn't produce the same result. Some of all of the house wiring is providing that effect but I'm not sure why because it's all the same wiring as before (just connected to a phone jack now first before being screwed into place on the screw/nut).

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 03:53:08 EDT

Bicephale : If you're patient i can answer your last questions after generating
long-term curves... I'm not leaving the country and you neither, i
presume!

The present situation may happen to be transitory, if you can wait
then give it a few days to settle while recording data... It might be
premature to draw conclusions already. I'd simply suggest that you
keep track of which is which by writing...

Being a citizen of the most dynamic town in Canada it would be
a good idea to use that time to chase toroïdal ferrite cores, just
in case it might provide new avenues to explore! You can start
searching here:

»Re: lousy upstream SN margin

I wish you'll enjoy your time while shopping around!!!

:D

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 03:40:21 EDT

dslrocker3 : do you think I should be conerned about the fact that the house wiring with no phones attached acts like it's an unfiltered device according to my modem? Ater this rewiring, I strongly suspect that I now only have one jack that works with DSL properly, even if I go downstairs and remove the filter.

Or should I not worry about it because I did in fact filter it?

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 03:29:05 EDT

Bicephale : You know where to find me if you wish to get averaged curves!

:D

Just make sure the samples exceed 24 h in both directions: it's
too much trouble to try to re-synchronize half-day records, etc.

Ha, and check the log-file while recording to verify that its length
is normal - if you see what i mean!...

:p

In the summer, with lightning storms and high winds generating
electro-magnetic noise, many days are required, a few samples
may even turn out to be useless when corrupted by an isolated
event, for example. Of course, an ST5x6 + 'DSL_StatScope' is
the simplest way to reveal daily patterns if any... Good luck!

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 03:23:47 EDT

dslrocker3 : I have to get a larger sample size, observe more days and try seeing if powering on certain electronics in the house cause the error rate to go up. Besides, I only finished my rewiring about 12 hours ago.

The only reason I was thinking about electrical interference has to do with the fact that there is an old florescent desk light beside the computer that people use and it always makes a "pop" sound that can be heard in nearby speakers whenever it is turned on and off.

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 03:20:40 EDT

Bicephale : Or stratospheric movements...

You'll have a much more suitable perspective in a few days.

Re: Help/Advice with my Line Statistics please

Fri, 31 Jul 2009 03:18:10 EDT

dslrocker3 : I went and got a quick snack to eat before going to sleep and accidently woke my dad up. Since he was awake, he decided to turn on the downstairs desktop computer to use. In about 1hour and 20 minutes, I only had 30 RX FEC errors. In last approximately 15 mintues since he turned the computer on, I've had about 60.

Electrical interference?

Re: error rate

Fri, 31 Jul 2009 03:02:45 EDT

Bicephale : The PC-XT "compatible" power-supplies had a tendency to use
large 1½ inches dia. cores - which makes it easier to wind up to
fifteen turns of ordinary telephone-line cable around the core...

Those PSUs operated at relatively lower frequencies because the
switching circuits weren't as efficient back then, resulting in some
larger ferrite cores. That's the reason why legacy is better here.

:p

Re: error rate

Fri, 31 Jul 2009 02:48:09 EDT

dslrocker3 : why in the world would I have a power supply from an 8086 based computer? Old computer parts that don't work or that aren't at least in the range of Pentium 4 went into city "environment days" for recycling long ago! Back in the day of the early low quality ATX power supplies that only lasted 3 months and automatically kill one of your hard drive's PCBs whenever a capcitor pops, I should have kept a bunch of those to take the circle thing that holds the cords, whatever it's called.

come to think about it, since I filtered most of the house's wiring in the basement, I have not once seen that troublesome Impules Noise Warning on my 2wire modem's diagnostic page.

Re: error rate

Fri, 31 Jul 2009 02:41:52 EDT

Bicephale : I was thinking more along the lines of a PC-XT 8088 4,77 MHz...

...with CGA amber display...

Re: error rate

Fri, 31 Jul 2009 02:27:41 EDT

dslrocker3 :

said by Bicephale :

Would you, by any chance, happen to have some old/defective/
decommisioned computer power-suppy and/or VGA monitor, etc.?

The type which has a few ferrite toroïdal cores inside?...

I do have some old ATX power supplies around but the problem is that they still work!!! LOL... I can look around the house and I'm sure I can find something.

Re: error rate

Fri, 31 Jul 2009 02:21:59 EDT

Bicephale : Would you, by any chance, happen to have some old/defective/
decommisioned computer power-suppy and/or VGA monitor, etc.?

The type which has a few ferrite toroïdal cores inside?...

error rate

Fri, 31 Jul 2009 02:01:56 EDT

dslrocker3 : hmmm.. in the last 17 minutes since I last reset my 2wire modem:

0 rx crc
1 rx fec

I'm not use to that! hopefully, separating the house wiring from the modem wiring is helping!

Re: comparision of line statistics with different modems (July 3

Fri, 31 Jul 2009 01:44:54 EDT

Bicephale : No disapointment, i'm just puzzled by that...

I guess -27.4 dB of uncancelled echo is a good thing, though.

:)