dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
1169
share rss forum feed

TheMG
Premium
join:2007-09-04
Canada
kudos:3
Reviews:
·NorthWest Tel

Copper pair loss.

Google doesn't like me tonight, I can't seem to find what I'm looking for.

Basically what I'm trying to find is how much loss can be expected over a typical 24 AWG copper pair, at 1000Hz, over a distance of approx 3.9km. Both ends terminated into 600 ohms.

Getting about 20dB loss at 1000Hz wondering is that's in the ballpark of what I should be getting. Seems kind of high which is why I'm checking.



leibold
Premium,MVM
join:2002-07-09
Sunnyvale, CA
kudos:10
Reviews:
·SONIC.NET

1 recommendation

I pulled out my old telecom databook and it has a dB/km @ 800Hz table for some common telco cables in europe (sorry no awg wires).

For 0.4 mm diameter wires (26awg) the attenuation is 1.31 - 1.80 dB/km.
For 0.6 mm diameter wires (22-23awg) the attenuation is 0.90 - 1.22 dB/km.

20dB for 3.9km @ 1000Hz seems to be high but not unreasonable.
--
Got some spare cpu cycles ? Join Team Helix or Team Starfire!



54067323

join:2012-09-25
Tuscaloosa, AL
reply to TheMG

said by TheMG:

Getting about 20dB loss at 1000Hz wondering is that's in the ballpark of what I should be getting. Seems kind of high which is why I'm checking.

DSL Assistant is showing 11 db at 1.2 Khz with 12,795kf of PIC cable at 70 degrees.




As such 20 seems a tad bit high.

TheMG
Premium
join:2007-09-04
Canada
kudos:3
reply to TheMG

Hmmm yeah seems a tad high.

Thanks.



54067323

join:2012-09-25
Tuscaloosa, AL

said by TheMG:

Hmmm yeah seems a tad high.

Thanks.

Any ideas what may have happened to cause that loop to measure long? I had to push it out to 32Kf to get to 19.7

TheMG
Premium
join:2007-09-04
Canada
kudos:3
Reviews:
·NorthWest Tel

said by 54067323:

Any ideas what may have happened to cause that loop to measure long? I had to push it out to 32Kf to get to 19.7

Bad (resistive) splice most likely. It wouldn't be the first time. Telco infrastructure takes one hell of a beating from the weather here.

Mainly wanted to check if there is indeed an "issue" before getting telco involved. I don't have reference reading of what is normal for that particular line. AGC is able to compensate, but I found the 20dB loss during checks for a different problem to be somewhat suspicious.


54067323

join:2012-09-25
Tuscaloosa, AL

said by TheMG:

I don't have reference reading of what is normal for that particular line. AGC is able to compensate, but I found the 20dB loss during checks for a different problem to be somewhat suspicious.

If it is a point to point copper loop you might consider once it's repaired in addition to a baseline loss measurement at 404, 1004, and 2804, also measuring it's DC, A to Z resistance to a short.

At least with a baseline you can try to nudge the Telco circuit tech in the correct direction, the reason being nowadays I find it all too common for Telco circuit techs to have zippo experience with anything “audio or analog” the prevailing attitude being “if it touches it talks” which when trying to transport audio or low rate digital data service doesn’t cut it.

Seems like if the circuit isn’t HDSLx, their eyes glaze over, they break out in a cold sweat and their brains go blank.

lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1
reply to TheMG

You could borrow a JDSU HST-3000 and do a basic TDR check for any high resistance or wet sections.

»www.jdsu.com/ProductLiterature/h···m-ae.pdf



54067323

join:2012-09-25
Tuscaloosa, AL

1 edit

said by lutful:

You could borrow a JDSU HST-3000 and do a basic TDR check for any high resistance or wet sections.

A TDR would never be my first choice for locating a high resistance splice (high open), the reason being the majority of high opens are single sided events (half opens), that is to say the fault will only be on one side of the pair, with workers that will usually be the tip though on a dry circuit it can be on the ring adjacent to the tip of a worker.

A TDR really cannot detect a half open or for that matter a high open be one sided or affecting both conductors, it just looks right through them for example, this pair I have my JDSU HST3000 TDR on has a serious problem, that being I popped the tip open at an adjacent terminal 626 feet from where I am running the TDR.







Note with a “normal” gain setting of 4 the half open is not even indicated and when I crank up the gain to 64 which is hot for the short range setting of the JDSU, it barely indicates the presence of the half open.







However, at 64 it does show something that looks like a problem beginning at 432 feet, which is really not a problem at all, just a replacement section of cable.




And false positives, which are common with a metallic TDR can really have a tech running in circles.

The correct method to isolate a high open is to go to one end of the loop and have the CO open the circuit to the far end and then place a short across the pair, note the loop resistance, then have the CO ground the tip and ring and then measure the resistance of the tip and ring to ground and those readings should be 50% of the loop reading +/- 10%. Then go to the other end and repeat.

BTW this can be done with a JD$U or a $75 Triplett 310.

Once the high open is isolated then it is time to divide and conquer.

lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1

said by 54067323:

said by lutful:

You could borrow a JDSU HST-3000 and do a basic TDR check for any high resistance or wet sections.

A TDR would never be my first choice ...
A TDR really cannot detect ...
The correct method ...

You cannot deduce locations without a TDR. I suggested a widely available but basic TDR.

There are more advanced TDRs which "really can detect" almost any flaw imaginable in each conductor and are indeed the "first choice" of many professionals.


54067323

join:2012-09-25
Tuscaloosa, AL

1 edit

said by lutful:

You cannot deduce locations without a TDR. I suggested a widely available but basic TDR.

I most certainly can and techs have been doing it for years, long before the TDR was introduced.

Ever heard of a device called a “cable fault locator” professionals use them to locate “faults” in OSP cabling, be it buried or aerial. Fact is I can locate a high open with one down to within a few feet, which is real important when digging a pit is necessary.

There are more advanced TDRs which "really can detect" almost any flaw imaginable in each conductor and are indeed the "first choice" of many professionals.

In the lab maybe, but not in the metallic "real" world, there are just too many variables in a working cable for a TDR to do the job each and every time, a TDR is merely one of many tools a professional uses to troubleshoot.

Plus as I mentioned, too many false positives results in too much tail chasing, hence the reason using a device that can locate almost any flaw gets old real fast, as such a “professional” will not make a TDR the first choice when he knows from “experience” a VOM is the proper tool for the job, this is something greenhorns learn after they burn out a bit.

For example as I clearly demonstrated an advanced TDR such as the JDSU HST-3000 will not detect a half open, however, it will locate a bridged tap within a few feet and the same for a WAO be it wet or dry, but in a working cable there are simply too many transients and noise for a TDR to be gained up to the point it can see a “dry” resistance change in a splice.

Good instrument, incorrect application.

Now fiber is a different world, in fiber one doesn’t have cross talk, leakage, cross battery, 710 mods and slightly un-twisted pairs here and there, fiber is pristine in comparison to copper, as such a professional which I am, will get out only two tools when shooting a fiber trouble, my Anritsu light pen and my Anritsu OTDR and with those two devices there is no fiber problem I cannot solve and solve quickly.


54067323

join:2012-09-25
Tuscaloosa, AL

1 edit
reply to lutful

said by lutful:

You cannot deduce locations without a TDR. I suggested a widely available but basic TDR.

JDSU offers a 2 day class on the HST-3000 in which you will learn the in and outs of a metallic TDR and what you should and need not use it for.

I took it and it is well worth the money for what you will learn.


54067323

join:2012-09-25
Tuscaloosa, AL

3 edits
reply to lutful

BTW Lutful, if you ever wondered what the correct procedure for testing OSP copper is, all you need to do is find a JDSU HST-3000 and checkout the startup menu, the engineers at JDSU laid out the menu in the order that the tests should be run.

The folks at JDSU actually understand OSP copper and the quirks that is brings to the table, unlike those idiots folks who designed the CopperPro.




Note you don’t skip to test 4 until the others have passed.

BTW if you don’t already have one these folks have good pricing and fast shipping on the HST-3000.

»www.solidsignal.com/pview.asp?p=···63768538

TheMG
Premium
join:2007-09-04
Canada
kudos:3
reply to TheMG

Well, had the telco switch it to a different pair. Avg loss through 300-3kHz is around 7.5dB which is right in the ballpark. S/N 60dB. All is good now.



54067323

join:2012-09-25
Tuscaloosa, AL

said by TheMG:

Well, had the telco switch it to a different pair. Avg loss through 300-3kHz is around 7.5dB which is right in the ballpark. S/N 60dB. All is good now.

Good deal, 300-3K if I might ask what are you running on that?

TheMG
Premium
join:2007-09-04
Canada
kudos:3

Voice. Goes to a transmitter located in a separate building.


lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1

1 edit
reply to 54067323

said by 54067323:

BTW Lutful, if you ever wondered what the correct procedure for testing OSP copper is ...

I may have wondered about that way back in the 1980s.

You may have come across my 1991 M.Sc. project in a recent thread in the wisp forum. I designed several cutting edge TDRs, both O and M type and even a hybrid O+M box since then.

Being familiar with all kinds of TDRs, I may have forgotten how we can deduce location of faults in the middle of a cable without using TDR. Maybe you can explain the theory in a few paragraphs?

said by 54067323:

... all you need to do is find a JDSU HST-3000 and checkout the startup menu

I mentioned JDSU HST-3000, which is a very basic MTDR, because it is widely available and inexpensive. There are much better MTDRs.

TheMG See Profile EXFO is another option which is available widely. »www.exfo.com/Documents/TechDocum···ngHR.pdf

This appnote shows actual traces for many types of faults including high impedance and wet sections.
»documents.exfo.com/appnotes/anote168-ang.pdf


54067323

join:2012-09-25
Tuscaloosa, AL

1 edit

said by lutful:

Being familiar with all kinds of TDRs, I may have forgotten how we can deduce location of faults in the middle of a cable without using TDR. Maybe you can explain the theory in a few paragraphs?

Since I have already demonstrated high opens cannot be located with an actual TDR that you recommended to use for that application, not a copy and paste from a web page, why bother, it seems all you want to do is argue, not learn?

lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1

said by 54067323:

said by lutful:

Being familiar with all kinds of TDRs, I may have forgotten how we can deduce location of faults in the middle of a cable without using TDR. Maybe you can explain the theory in a few paragraphs?

Since I have already demonstrated high opens cannot be located with an actual TDR ... why bother, it seems all you want to do is argue, not learn?

Did you actually take a look inside my M.Sc. thesis? I have designed advanced TDRs since 1989 and have provided engineering consultation for several commercial OTDR/MTDR products. You could even be using one ... think of that irony.

You can borrow an EXFO unit, which is used to pre-qualify loops for ADSL2+ speeds, and check any full or partial opens which your particular JDSU unit could not detect.

*** Are you actually able to explain how fault locations can be found by testing a cable only from the ends without measuring reflections? No screenshots please ... just a few paragraphs.


54067323

join:2012-09-25
Tuscaloosa, AL

2 edits

said by lutful:

Did you actually take a look inside my M.Sc. thesis?

Why bother I really do not care what makes a TDR work I am only concerned with interpretation the information it is giving me and I have gotten damm good at it too, but I have also the experience to know when a TDR is not the correct instrument for the job, something you fail to understand.

I have designed advanced TDRs since 1989 and have provided engineering consultation for several commercial OTDR/MTDR products.

Bravo for you, how many years of experience do you have locating and clearing cable faults though.

You can borrow an EXFO unit, which is used to pre-qualify loops for ADSL2+ speeds, and check any full or partial opens which your particular JDSU unit could not detect.

The EXFO is old news, the JDSU can do everything an EXFO can and more depending on how much money one wants to spend on SIM’s. Lets see the basic HST-3000c unit will qualify POTS, DDS, Analog Data, ADSL, ADSL2, ADSL2+, HDSL, HDSL2, and HDSL4, with other SIM’s, DS1, DS3, Ethernet, Gigabit Ethernet, optical Ehternet and since it is a software based platform most anything else that comes down the pike. OH I almost forgot it also does sync.

Are you actually able to explain how fault locations can be found by testing a cable only from the ends without measuring reflections? No screenshots please just a few paragraphs

The less the process is so simple, I wonder why you never thought of it and I will gladly give you a clue, it involves a spare pair and a device that internally is not much more than a precision wheatstone bridge.

You connect it to the pair under test and the spare pair, spin a multi-turn pot and when the meter centers, the reading on the dial is the location of the fault, within a foot or two.

By the way the JDSU can be used the same way if one knows how to use it.

And I will put it up against a TDR any day of the week and unlike the guy with the TDR I will not be nick-named “the gopher” nor wear out a shovel hunting down false positives.

Also a cable fault locator being DC based can work through loads which stop a TDR dead in it's tracks.


54067323

join:2012-09-25
Tuscaloosa, AL
reply to lutful

said by lutful:

You can borrow an EXFO unit, which is used to pre-qualify loops for ADSL2+ speeds, and check any full or partial opens which your particular JDSU unit could not detect.

BTW managers are not likely be handing out EXFO's or JDSU's to $30 per hour field techs to be left in the rear of their u-body or on top of a crossbox in the rain, if a tech is lucky he might get a hand-me-down fault locator and a 965 which by the way are used thousands times per day in the US to quickly and sucessfully locate and clear faults.

lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1

1 edit
reply to 54067323

Click for full size
said by 54067323:

process is so simple ... I will gladly give you a clue ... You connect it to the pair under test and the spare pair, spin a multi-turn pot and when the meter centers, the reading on the dial is the location of the fault, within a foot or two.

Do you actually understand the technical "process" going on inside that tool? For example, what if the cable had 2 different locations with higher impedance?

BTW I made a strategic decision by mentioning JDSU before EXFO because I sort of predicted how you will react. That is also why I did not mention more complex/expensive TDRs.

I am only hoping that my rhetorical questions will help you and others to understand why TDRs (and TDTs) are necessary to qualify long copper loops. The actual brand/model does not matter that much.

*** Quote from a 965 fault locator document:
In most cases, a TDR (Time Domain Reflectometer) which uses RADAR technology is a better test set to use for locating any type of OPENS [Complete Open, Partial (high-resistance) Open or Dirty Open (a combination of resistance and capacitance faults)]


54067323

join:2012-09-25
Tuscaloosa, AL

3 edits

said by lutful:

Do you actually understand the technical "process" going on inside that tool?

If you are referring to a cable fault locator then the answer is yes, if you want to know how a 965 works internally you will have to ask 3M.

For example, what if the cable had 2 different locations with higher impedance?

Because impedance does not matter to a DC based test device, like a fault locator.

And since you brought up impedance and long loops how does a TDR measure beyond a load coil?

*** Quote from a 965 fault locator document:
In most cases, a TDR (Time Domain Reflectometer) which uses RADAR technology is a better test set to use for locating any type of OPENS [Complete Open, Partial (high-resistance) Open or Dirty Open (a combination of resistance and capacitance faults)]

A 965 is not what I consider a fault locator it is merely kickmeter on steroids, though it does make a nice seat to sit upon when operating a cable fault locator.

BTW I made a strategic decision by mentioning JDSU before EXFO because I sort of predicted how you will react. That is also why I did not mention more complex/expensive TDRs.

Yea right, like you had any idea I owned one...

And isn't that kinda like trolling??

lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1

said by 54067323:

if you want to know how a 965 works internally you will have to ask 3M ... 965 is not what I consider a fault locator

You mentioned the 965 earlier. What brand/model do you recommend? FYI all classic "cable fault testers" use Murray/Varley bridge circuits and may also have a tone generator/receiver for opens.

Many newer "cable fault testers" are just a simplified TDR with a numeric display to show distance to very first fault.

Anyway, I am curious what (non-TDR) tool you will personally consider for testing the loop when you have access to ring/tip only from TheMG See Profile residence? That is a very relevant question.


54067323

join:2012-09-25
Tuscaloosa, AL

said by lutful:

You mentioned the 965 earlier. What brand/model do you recommend?

For a POTS or circuits tech shooting the MG's trouble, the 965 DSP is really all that is needed it’s simple and intuitive and will serve a field tech very well for 90% of what they come across.

The 965DSP is literally testing for cavemen.

If the MG's trouble becomes a chronic, then you bring in say a HST-3000 and using that device as a DVOM you can “stress” the pair, the JDSU does this by pulsing the pair with 200+ VDC and measuring the resistance under the higher voltage and the lower voltage, a large difference between the two fails the test, this test also because of cable capacitance has the tendency to fuse open “high opens” making menu selection 2 a lot easier to use..

Anyway, I am curious what (non-TDR) tool you will personally consider for testing the loop when you have access to ring/tip only from TheMG See Profile residence? That is a very relevant question.

That depends, if a tech has CO coverage then the best way is head to head testing, lacking CO coverage, I personally would recommend placing the JDSU FED at the customers prem and going to the CO or the RT’s protector field and doing a full wideband test.

Now that may seem like overkill for a “audio” circuit and I may burn up a few pairs in the process, but I will deliver to the end user a loop that will last.
Expand your moderator at work


Jason
Stowage Class Traveler
Premium,Mod
join:2001-01-24
38.2967 Lat
kudos:3

1 edit
reply to TheMG

Re: Copper pair loss.

hold on..

Experiencing technical difficulties.. (i.e. Jason pushed the wrong buttons again.)
If you're experiencing missing posts, its just me gone and done something stupid.. I'll see if I can get them back.. I apologize for any inconvenience


Erm.. Nevermind.. My public humiliation is complete

Nothing to see here, please move along.. LOL


--
When cryptography is outlawed, bayl bhgynjf jvyy unir cevinpl.



54067323

join:2012-09-25
Tuscaloosa, AL
reply to lutful

said by lutful:

Many newer "cable fault testers" are just a simplified TDR with a numeric display to show distance to very first fault.

No one in the telecom world has a need for a cable fault tester, because the technician in-the-field has already determined there is a fault in the cable.

What is needed is a cable fault locator, so the location of the “fault” can be determined and cleared…

lutful
... of ideas
Premium
join:2005-06-16
Ottawa, ON
kudos:1

1 edit
reply to 54067323

said by 54067323:

If the MG's trouble becomes a chronic, then you bring in say a HST-3000 and using that device as a DVOM ...

I am looking back at my original suggestion to TheMG See Profile and your very first response.

said by 54067323:

said by lutful:

You could borrow a JDSU HST-3000 and do a basic TDR check for any high resistance or wet sections.

A TDR would never be my first choice for locating a high resistance splice (high open), the reason being the majority of high opens are single sided events (half opens), that is to say the fault will only be on one side of the pair

So it seems (now) that you did NOT have a problem with me suggesting JDSU HST-3000. You only wanted to make sure that he does not use it's TDR functionality to "check for any high resistance or wet sections" because of your personal experience.

But that opinion is actually contrary to user manuals and support engineers of any non-TDR "fault locator" ... they will happily explain about deficiencies when testing for full/partial opens. I already quoted from 965DSP earlier, you can look up a few others.

You also check EXFO and other "true" TDR user manuals for exact waveforms from all kinds of open faults and decide if you can finally change your personal opinion about TDRs or at least refrain from making comments that are sure to start an argument.

TheMG
Premium
join:2007-09-04
Canada
kudos:3
Reviews:
·NorthWest Tel

1 recommendation

reply to TheMG

Wow, what the heck happened to this thread? Never expected this to turn into a huge debate about TDRs and fault locators!

Signal did not get from demarc-A to demarc-B the way it was supposed to, telco was contacted, telco switched the pair, problem solved. Case closed.