Testing Dry DSL Lines

[big john]

We've got some telemetry that takes place over 2-pair dry DSL lines.

I've been working with the phone company to verify operation and ran into something I don't understand:

They can inject a 1004Hz signal using a tester made by Tempo that reads the loop loss in dBm. When I connect a DMM to their tester, I can read that it's putting out 1004Hz at about 1.5 volts.

When I go to the other end of the DSL line, I can hear this tone with a butt-set and see it on my DMM.

Now, I've tried to reproduce this test myself: I connect a function generator to the line at 1004Hz and go to the other end, I can hear the tone with the butt-set but there's not a meter in my arsenal capable of picking up the frequency, even if I inject it as high as 10 volts.

Why on earth would I not be able to reproduce this test?

-John

 

[Sierrasparky]

I don'y know much about this , I wonder if you are not sending enough current.

I never heard of a DRY DSL line , I have heard of leased line and dry pairs.

 

[big john]

Bear with me, I'm not a phone guru.

I've heard these darn lines called everything from dry-lines, to leased-lines, to dry-DSL. Even within the phone company they don't seem to have a standard name: They're simple copper pairs that go through the CO with no amplification and no battery voltage.

Maybe it is as simple as current, but even with the 75Ω load at the CO my voltage measured between the lines stays above 3 volts, which is still higher than what the TELCO injected.

Is anyone familiar with the test they are doing that might be able to explain it in more detail? My version must be missing something.

-John

 

[Sierrasparky]

Yea the lack of consistent terminology can really be a pain.

 

[hbiss]

The only dry DSL line that I am familiar with originates at the CO. It is simply a pair that has no battery voltage. In your case I believe you are using a DSL "transmitter" and "receiver" and those should operate on a dry pair which is just a pair originating at one location and running to the other. I suspect that this is not what you have. What is the distance involved? Is there a CO involved? Is there DC continuity on a pair- in other words if you short one end and check the other end with an ohm meter would you see the short?

-Hal

 

[big john]

The only dry DSL line that I am familiar with originates at the CO. It is simply a pair that has no battery voltage. In your case I believe you are using a DSL "transmitter" and "receiver" and those should operate on a dry pair which is just a pair originating at one location and running to the other....

That sounds right. We have one pair for transmitt, and one pair for recieve with no batter voltage on either.

...What is the distance involved...?

One way, anywhere from 3/4 mile to 2 miles depending on the circuit.

...Is there a CO involved?

Only to the extend that the wires pass through a test point and lightning arrestor there. They aren't amplifying anything.

...Is there DC continuity on a pair- in other words if you short one end and check the other end with an ohm meter would you see the short?

This has worked on some, but not others. Not sure if it's because I don't have a meter than can push continuity through a circuit that long, or if it's because the circuits are damaged (some of them have definitely been damaged, I'm trying to verify the repairs.) So, that's why I was hoping to actually get a frequency through at a known voltage, it'd be a surer test and I'd be able to see the line-loss.

-John

 

[hbiss]

This has worked on some, but not others. Not sure if it's because I don't have a meter than can push continuity through a circuit that long, or if it's because the circuits are damaged (some of them have definitely been damaged, I'm trying to verify the repairs.) So, that's why I was hoping to actually get a frequency through at a known voltage, it'd be a surer test and I'd be able to see the line-loss.

Forget tone, the only test you need to do is continuity. 2 miles is nothing and any meter should work. If these are really dry pairs, for a 2 mile loop of 22 ga wire you should read about 340 ohms. That should give you a good idea as to whether the pair is good or not and how long it is. (0.01614 ohm per foot. Remember loop length is twice the actual length.) You need to measure line loss with equipment calibrated in db or know how to convert voltage to db to be meaningfull. Besides, DSL doesn't care much about line loss since it's RF that just rides on the pair. DSL can work even if the pair is open. There is a limit though for line length and DSL but I think 2 miles will make it depending on your equipment.

That said, why are you even doing this? The lines are provided by the phone company, just connect your equipment and if it doesn't work tell them to fix the line.

-Hal

 

[big john]

Hal, the only other thing is that there's the 75Ω load at the CO from the lightning arrestors/whatever, and that seems to also be bogging things down when it comes to continuity.

It would be absolutely fantastic if all I had to do was call the TELCO and I could expect effective repairs, but our provider is one of the most incompetent companies I've ever seen. I've been told about a half-dozen times now that all these lines are fixed, but every single time I call them back out they find significant defects. To say I don't trust their assessment puts it very mildly.

-John

 

[hbiss]

If there is 75 ohms across the pair it's probably the heat coil in the protector. For giggles if you know it's there with an ohm meter you should be able to figure the approximate length of each loop either side of the CO where the protector is. It still should respond to a short at one end also. If not then it's open or it's not just wire from point A to point B.

I'm not sure that a protector with a heat coil is appropriate for your DSL circuit. I know I have had to use high speed protectors without heat coils when I use ethernet > DSL > ethernet converters to extend a network out to another building over a pair on the phone cable. Our normal gas tube fused POTS protectors (and those don't even have a heat coils) won't allow those converters to commmunicate. But then again phone company DSL works just fine with those protectors.


-Hal

 

[big john]

...I'm not sure that a protector with a heat coil is appropriate for your DSL circuit....

Sorry to hear you say that, because one of the engineers is worried about it, too. I've been trying to down-play it because I don't think there's a darn thing I can do about it, I just hope it works. How would I get the CO to agree to remove that protection?

-John

 

[hbiss]

Wouldn't have to remove it just change it. There are lots of different 5 pin protectors, the ones I'm talking about are 5 pin. Temporarily to troubleshoot pull the protector and replace it with a jumper. If your connection comes up you've found your problem. Now select a solid state high speed protector without a heat coil.

-Hal

 

[steve066]

Are you sure they don't have an attenuator on the recieving end of the line? They might call it a signal conditioner, or any other misnomer.

God knows why they want to put those on leased lines. It takes a nice 1 volt signal and turns it into a millivolt signal. Its to "reduce noise" on phone tech told me.

I've been in the same boat with the local phone company here. We had several leased lines that we replaced with IDSN lines because they worked so poorly. They didn't maintain the IDSN lines any better. When one would go down, & we'd call to get it fixed, somehow they would switch them back to regular phone lines. That was really frustrating.

And just try ordering a leased line:

"I need to order a new leased telephone line."

"We don't have anything like that. Just residential or business lines."

"Yes,you do. I've ordered them before. Its not a regular phone line, its......."

"We don't have anything like that"

"Can you ask a supervisor about it??"

"Well I guess....."

:rant:

 

[big john]

Wouldn't have to remove it just change it. There are lots of different 5 pin protectors, the ones I'm talking about are 5 pin. Temporarily to troubleshoot pull the protector and replace it with a jumper. If your connection comes up you've found your problem. Now select a solid state high speed protector without a heat coil.

Oh, the Cook Blocks! Sorry, like I said, I only know a little of the lingo. We actually have those at the central location for theses circuits, in addition to whatever protection they have in the CO. Not sure how to spec. them or where to get them, but I'll look around for a solid-state version.

I found the manual for the network cards they've been installing, so I am sure they aren't being used as amplifiers, because I can verify the settings on the cards. And for what little it's worth, the techs have assured me these cards aren't actually changing the signals at all.

-John

 

[hbiss]

I found the manual for the network cards they've been installing

Who has been installing these? Are they the DSL cards that are installed at your equipment or are these by the phone company?

-Hal

 

[big john]

I found the manual for the network cards they've been installing

Who has been installing these? Are they the DSL cards that are installed at your equipment or are these by the phone company?l

These are called "Data Station Termination" cards, they've got 120V power supplied to them, and the TELCO has been putting them in at each end. They can send a tone and put the card into "loop-back test" which enables them to test the circuit from the CO. I think that's the main reason they're requiring them.

-John

 

[hbiss]

Ahhh you have data circuits! Forget everything I said above. It's 4 wire HDSL and there are likely repeaters along the line as well as the CO. You need a T Bird or other data test set. Typically you would put one end in loop-back mode (or use a loop-back plug) then send and receive HDSL test data from the other end via the test set.

My knowledge of data circuits is limited so you might want to go here, sign up and ask an expert who is a CO engineer http://www.myphonetechs.com/index.php?board=38.0

-Hal