dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
1471
share rss forum feed


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
Reviews:
·Verizon Broadban..
·AT&T U-Verse
·Comcast

[Signals] Modem signals better when split twice on highdB legs..

Click for full size
-3.5 dB Loss Leg
Click for full size
-7.5 dB Loss Leg
Anybody who can explain this?
I use the Arris TG862G for internet and phone. Now this is true for my other modem which is a UBEE-DDM3513

My confusion is that my modem(s) perform better when put on the higher dB (-7.5dB) loss legs than on the lower dB (-3.5) loss legs.
The top picture is of when my modem is connected to the low loss -3.5 dB leg. The lower picture is of the opposite, when connected to the -7.5 dB leg. Why are my levels better? Isn't this a contradiction stated in this very site's FAQ?
See FAQ: »Cable Modems and Wiring Issues »What kind of signal levels do I want on my cable modem?

I live in an apartment complex where each building gets it's own demarcation box where the signal is split and sent to each apartment. I have walked around my complex and 0f 8- 2 story buildings we have 4 nodes. That's a lot of power, and almost even overkill if you ask me.
But my real confusion is that my modem get's closer to the IDEAL signal levels when I split it.
I mean split it unnecessarily. I moved in, ran RG6Q from the demarc box (700 feet) to where my modem, TV, and other device exists. All wiring is RG6Q with correct connectors. I am not dumb, I used to run cabling like this for RCN cable in Chicago, so I have the tools and knowledge, except for this.

Why is my modem working better when I am connecting it to higher loss legs on a passive splitter than the lower loss legs.

If you can answer any of this I will be grateful b/c I don't understand a bit of this.
Thanks.
Bill
--
Bill - Hattiesburg, MS - »www.billhamel.net

Cable Employ

join:2012-07-23
Saint Paul, MN

Re: [Signals] Modem signals better when split twice on highdB le

All cable equipment works best at around 0dBmV coming into the device. By putting it on the -7.5 leg, you are loosing 7.5dBmV instead of the 3.5dBmV on the other leg, meaning, you are getting closer to the 0dBmV that is ideal. However, you should not, in reality, see any real difference in performance between the two. While you want an ideal level of 0dBmV, the modem is designed to work between about -10dBmv and +10dBmV.


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
Reviews:
·Verizon Broadban..
·AT&T U-Verse
·Comcast
Then it would theoretically, since it is ideal to be at 0dBmV, to add another splitter, literally splitting the connection twice.

In fact, after the apt complex upgraded my COAX to RG6 from RG59 I HAD to split (using two 2-way splitters) and putting the modem on the last splitter down the line just to keep the power levels from going over 15dBmV, and i'm for real.
I think the power levels are higher around this apt complex (and general neighborhood) b/c RG59 was standard when these places were built. Maybe i'm wrong about the increase in power and RG59 connection, but when I lived in Chicago I had to (since I worked for RCN Cable) run a dedicated line drop from the telephone pole to my duplex just to get enough power (and a reasonable SNR) to maintain what was my 10Mbps internet package.
There, I used active splitters. Had to. I was sharing a 5 bed duplex, each with it's own TV, and the internet as well.
I come to Mississippi and I have to find ways to lower my power levels otherwise my modem won't sync w/ the headend.
Odd infrastructure...2 cable companies so probably 2 different methodologies to their network infrastructure.
Who knows....
-Bill
--
Bill - Hattiesburg, MS - »www.billhamel.net


EG
The wings of love
Premium
join:2006-11-18
Union, NJ
kudos:10
reply to whamel
FWIW here, the upstream power level needs to be taken in to consideration as well because adding or extracting different amounts of overall line attenuation affects it as well in an equal but opposite way. One does not want to throw their upstream / return power out of spec while trying to make the downstream be in spec / look pretty or they may be recreating the very problem that they are trying to eliminate.


Streetlight

join:2005-11-07
Colorado Springs, CO
reply to whamel
whamel,

What seems strange about your results is that adding another 3.5 dBmV of attenuation did not raise the upstream power level to ~44.0 dBmV. I recently rewired my house with high quality RG6 cable and added an extra 3.5 dBmV attenuation from a 2-way splitter and the downstream power dropped ~3.5 dBmV and upstream power went up ~3.5 dBmV. I would have liked the upstream to stay the same but reduce the downstream power level.

Anyone have a comment?
--
There is nothing more deceptive than an obvious fact.

Sherlock Holmes in
The Boscombe Valley Mystery
A. C. Doyle
Strand Magazine, October 1891


EG
The wings of love
Premium
join:2006-11-18
Union, NJ
kudos:10

1 edit
said by Streetlight:

whamel,

I would have liked the upstream to stay the same but reduce the downstream power level.

Anyone have a comment?

Then you would need a "forward path only" attenuator such as one of these;

»www.mac4mate.com/index.php?route···tenuator

They are rare in American DOCSIS systems but are common in Euro-DOCSIS systems. That site is in the U.K. Be advised that when I ordered some from it, they took about a month to arrive.


Streetlight

join:2005-11-07
Colorado Springs, CO
Thanks, EG. I'll look around for a forward path attenuator. Just what I need.

What about whamel's situation: no rise in the upstream power level. I guess we need to know the model of splitters being used. In my case I used Extreme Broadband Engineering's splitters which are what Comcast uses in our area. Are there forward path only splitters? I've not heard of them.

My levels after rewiring were, on 4 down channels, at about 9.5 dBmv and 3 up channels at about 42 dBmV. After adding an extra 2-way 3.4 dBmV splitter, I got an OK compromise of down of ~5 to 6 dBmV and up ~45.5 dBmV. My experience here is that on my Zoom 5341J modem if up goes close to 50 dBmV, problems occur and with our high summer temperatures, I didn't want to push my luck with another 3.5 dBmV of attenuation and have enough headroom for these signal level changes due to temperature and other causes . Also, it may be that this modem is very sensitive to upstream high power. So far, I've had absolutely no Internet problems with my current setup.
--
There is nothing more deceptive than an obvious fact.

Sherlock Holmes in
The Boscombe Valley Mystery
A. C. Doyle
Strand Magazine, October 1891


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
reply to EG
I like where you are headed with this...very true.
--
Bill - Hattiesburg, MS - »www.billhamel.net


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
Reviews:
·Verizon Broadban..
·AT&T U-Verse
·Comcast
reply to Streetlight
If you notice, the second (bottom) picture where I used the -3.5dB leg, the DOCSIS 1.X freq 29.5MHz increased 3dBmV. I know for a fact I use that frequency. I have another modem UBEE DDM3513 that is provisioned on the same line in my apt (I have 2 internet connections) and it it says I config. page says I use that upstream frequency. Weirdly enough, after re-wiring and installing a new passive 3-way splitter, the downstream frequency switched from channel 1 to channel 4 which means I go from 669MHz to 663MHz.
Can't seem to figure it out because I know certain areas are provisioned on a certain frequency to avoid "crowding" and loss of speed and signal quality. In Hattiesburg, MS we have 4 downstream channels available to us and only 2 upstream.
I have just noticed that while I was in Chicago, (my mothers home in the burbs) her modem had 8 downstream channels and 4 upstream. But that forward-path attenuator sounds like a good option.
Thanks for the suggestion.
-Bill
--
Bill - Hattiesburg, MS - »www.billhamel.net


LMoreland

join:2010-07-28
Littleton, CO
reply to EG
I ordered some forward path attenuators from mac4mate.com about a month ago and they arrived in about 8 days.


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
Reviews:
·Verizon Broadban..
·AT&T U-Verse
·Comcast
reply to EG
said by EG:

said by Streetlight:

whamel,

I would have liked the upstream to stay the same but reduce the downstream power level.

Anyone have a comment?

Then you would need a "forward path only" attenuator such as one of these;

»www.mac4mate.com/index.php?route···tenuator

They are rare in American DOCSIS systems but are common in Euro-DOCSIS systems. That site is in the U.K. Be advised that when I ordered some from it, they took about a month to arrive.

LOOKIE here: 3dB attenuator----->
»www.amazon.com/Parts-Express-In-···asin_lnk
6dBattenuator---->
»www.amazon.com/Parts-Express-In-···_sim_e_4
--
Bill - Hattiesburg, MS - »www.billhamel.net


EG
The wings of love
Premium
join:2006-11-18
Union, NJ
kudos:10
Those are not "one way", they are *overall*.


LMoreland

join:2010-07-28
Littleton, CO
reply to whamel
In order to attenuate only the down stream channels, you want attenuators like the ones in the link that EG posted that attenuate only 65 to 860 MHz and do not affect the upstream channels which are below 65 MHz.


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
Reviews:
·Verizon Broadban..
·AT&T U-Verse
·Comcast
reply to EG
said by EG:

Those are not "one way", they are *overall*.

MY bad i was all excited...
--
Bill - Hattiesburg, MS - »www.billhamel.net


Streetlight

join:2005-11-07
Colorado Springs, CO
reply to whamel
whamel, and EG

Oops, I didn't notice the 2nd QAM signal had gone up in your changed setup. Not sure, then, why the higher upstream frequency was attenuated and not the lower.

I found the manufacturer's spec sheet for these things and it seems that there is some small attenuation back for frequencies below 65 MHz of 1 to 1.5 or 2.0 dBmV, but at higher frequencies they have the labeled attenuation. That is, a 6 dBmV one will attenuate 6 to 7 dBmV at 88 MHz and higher. Generally, that would be OK for internet frequencies - upstreams are 25-36 MHz, downstreams are 580-606 MHz on my modem - but maybe not the higher frequencies used for cable TV since there's no data for >860 MHz. This suggests that these things should be plugged in to the cable input of the modem and definitely not at the input to a splitter that my connect eventually to a cable box.

I found a UK blog that shows these attenuators attenuate both up and down streams:

»www.cableforum.co.uk/board/12/16···int.html

These results leave me with some questions about how well these things work.
--
There is nothing more deceptive than an obvious fact.

Sherlock Holmes in
The Boscombe Valley Mystery
A. C. Doyle
Strand Magazine, October 1891


EG
The wings of love
Premium
join:2006-11-18
Union, NJ
kudos:10
Too much is being read in to this IMO... The forward path only attenuators that I have perform fine and DO NOT attenuate the return path.. Perhaps they introduce 0.1 dB of attenuation in to the return path which is insignificant.


whamel
billhamel .net
Premium
join:2002-05-09
Hinsdale, IL
kudos:10
Reviews:
·Verizon Broadban..
·AT&T U-Verse
·Comcast
reply to Streetlight
said by Streetlight:

whamel, and EG

Oops, I didn't notice the 2nd QAM signal had gone up in your changed setup. Not sure, then, why the higher upstream frequency was attenuated and not the lower....
I found a UK blog that shows these attenuators attenuate both up and down streams:

»www.cableforum.co.uk/board/12/16···int.html

These results leave me with some questions about how well these things work.

I found attenuator "conditioners" that worked on both up and down stream. Was very confused about those. I will have to find the website again...
--
Bill - Hattiesburg, MS - »www.billhamel.net