how to test a T1 ?

I have been extending a lot of t1's lately. Other than a loop back Is there a way to test them ? What should I be looking for if I threw a volt meter on the line and what would I be testing TR T/R or tip to tip and what voltage should I be looking for ?

Also, what type of tester should I be on the look out for on eBay ?

Thanks

Voltmeters are 132% useless for any sort of T1 testing.

You need a BERT (Bit Error Rate Test Set) designed specifically for T1's, something along the lines of a FireBerd from Phoenix Microsystems (taken over by Sun a while back). Such can often be found used on Ebay and from used test gear dealers.

Whoa. Something is wrong there! Whatever is at the demarc, has a serious problem if it drops from 2.5 V to 0.7 V (assuming your house wiring is not literally a mile long). The 2.5V is good. The 0.7 V would probably work if that was what was coming in on the line, but an interface plus house cable should be losing that much signal unless your cable is really poor or really long.

The problem with a typical "voltmeter", is that it will not tell you what the pulse looks like, and unless it was designed specifically to look at a DS1 interface it will read something like RMS volts rather than peak volts. The RMS value of a T1 pulse is a worthless measurement, while knowing the peak voltage is somewhat useful. Even more useful is a waveform display.

Here is a short description, with a graph, showing what a T1 pulse is supposed to look like.

Most T1 test sets will give you a peak voltage reading, but some will actually trace out a waveform on top of a template that shows whether it is in specs or not.

The voltage will not change the speed (other than if it is too low, the speed will be zero).

A short time with almost any test set, and the instruction manual that comes with them, is usually very productive. Go to google and see what you can find...

The reason for the voltmeter question is, I am running a new line because the tech that tested the line said that there was 2.5V at the demarc, and ..7V at the jack that was extended over existing house wiring. I am Just curious how he tested the voltage. Apparently the line passed a loopback but the low voltage wasn't providing any speed.

Anyth>

hey Floyd, how well did that PCMCIA reader hold up?

I never heard of anyone having any problems with it. We had several of them in the field for years, being shipped all over Alaska on small planes. About 2000 or so we replaced all the cards with an update. The card slot is, if I remember right (I haven't seen one in a couple years now), on the back, and well protected (much better than on any laptop computer I've ever seen, but it also was not designed to be removed frequently).

Those units are about half as big and twice as rugged as anything else I've seen that will do the same job.

That's more like it.

Twenty pounds is exactly why that little 2-3 pound Sunrise unit is nice! Some of the T-Berd units can display waveforms though, but I've never ever felt that I needed that functionality (but I was rarely ever testing at the customer end of a cable either).

Neither of them is worth a darned if you need to really analyze any kind of trunk signaling though. So a place that does a lot of different kinds of digital services will have several different kinds of T1 test sets.

That line should have read "shouldn't be losing", of course.

Well that's probably why they had me run a new line, strung in a new 200 ft cat5E line and when the cpe tech showed up and tested it was fine .04 V drop.

Saw the tester also, a monster T-Bird. had to have weighed 20 pounds. Thanks for the link and input.

-->How about these?

Specifically, the NP7050I ?

The Sunrise unit does look really slick though.....

do these run? ( Price-wise )

Scotty

That /looks/ really nice. Modern technology! It has the potential to be upgradable too. But... it's very hard to say how rugged something like that is without hands on experience, and the same is true for the functionality of the user interface. I'd want someone (with enough experience to be able to know the difference) to use one for a day or two before buying one.

I did notice one thing not being described in the specs that would be an absolute requirement in my book. There was no indication what type of testing can be done on clock timing. Slips, jitter, interval errors, etc. Lots of T1 test sets can't do half of what is needed to be aware of timing issues, and since that is *frequently* where a lot of setup time and effort is spent, a test set *must* have that functionality to be the one and only test set available, particularly if it is used in the field as opposed to in a CO setting.

Same basic price. They start at $5000 and go up as you add optional features and hardware. I wouldn't get very concerned about price tag differences simply because even a couple thousand dollars difference can be saved by having the *right* test set on just one job. Multiply that by 1 per year for a few years and not just the difference is negligable, so is the entire cost of the any one of these test sets.

Well, I'm not sure what kind of a loopback was done. I was assuming that the provider did a loop from their office to the demarc. There are a variety of possible devices possibly being used at the point where your house cabling connects, and virtually all of them allow the provider to disconnect you and send a signal that will be returned. There are several ways of doing that, including just having an RJ-48 jack that physically loops the circuit when your connector is removed. The fact that it looped good to the demarc would point right at the house cabling. And we can assume also that your equipment probably has a loopback in it too, and that that did not work.

Various loopback "codes" can loop any repeaters or csu/dsu devices between your location and the provider's office, so it can all be tested by one lone tech without "windshield wiper time". Moreover, some test sets can be remotely controlled, and tech might be 3000 miles distant from any part of the entire circuit!

The type of tests can vary, depending on such things as the hardware configuration and the protocols used. But basically a known bit pattern is sent and bit errors are counted on the return side. Some of these tests can show really odd things too. A random enough signal will, for example, cause every possible frequency to be generated... and things like bridge taps (long lengths of unterminated cable jumpered to your pairs) will resonate at some frequency, and throw errors. Hence that is the most commonly used test. Others show specific characteristics, so sometimes 2 or 3 different test patterns might be sent.

As for testing cables... the usual method is to put a circuit on the pair, and if it works, the pair is declared to be good! For low density operations (if you don't have at least hundreds of circuits), that is usually sufficient. But you can also buy cable testing equipment that begins with simple DC resistance measurements and goes all the way up to using time domain reflectometry. The first isn't totally useless, but doesn't tell much. The second isn't totally useless either, but it sure tells more than is usually useful. Probably you don't need either and can save money by just swapping cable pairs to see if things work better.

Question 2 Obviously the house pairs that the T had been extended on were crap, even though it passed the loopback test. ( Is the loopback, Just continuity ? ) To avoid this problem in the future, what would you use to test the house pairs ?

Thanks