atcomsystems.ca/forum Forums Wiring - Telecom Carriers - Trunks & Services T1's, CSU's and DSU's 6 t-1's random troubles

I have a customer that has 6 t-1's, all smartjacks are located in the server room. The business is feed by 18pr cable (total cable lenght is 8900ft). The cable pairs have been changed and cable techs have been brought in to check the pairs and naturally they are clear of trouble and bridge taps. The t-1's have been alternating on which one goes into trouble but when they go into trouble the loop light starts blinking and sometimes it goes lights out for 5 or 10 seconds. The t-1 will run errors and if you reset the card it goes back to running clean. By the way the t-1's are single pair adc cards (and the cards have been changed). Any help will be greatly appreciated thank you.

The Transmits & Receives for the 6 T's are all on the same 18 pair cable?

Sam

Ya 'spose it's random lay, Sam? Isn't 18 pair what the 1st 1A2's used?! How old is that cable! John C. (Not Garand)

We have several pretty sharp CO guys that are familiar with this set up. I've never delt with the single pair T-1's.

Woops! I missed that - "single pair ADC's" - so forget the xmit and receive question.

My next question (hopefully better than my last question) is:

Have the circuits been stress tested end to end? Has someone gone out there with a T-Bird or Sunset and run to a hard loop? Or better yet gone head-to-head with a pair of testers, one at each end.

T-1 is just a transport. Let's make sure that the potholes are out of the road before we start checking the suspension.

Sam

hey sam, Yes the t-1's have been stressed with a ameritech and a looping plug in the smartjack. After doing some testing on monday I am beginning to wonder about ac voltage on the cable side, but as for crosses and grounds I can't find anything. Back to the stress test, On monday 1 t-1 crashed in the morning, When I got there the customer had reset the card and it was back to running great. We took the csu off put in a looping plug and it ran for 2 hours before it took 22 crc's. I checked cable pair and it was good then put the remote card back in and it runs great again. Man I am confused. Thank for the replies

Could the AC voltage be building up on the pair and after the remote card gets reset it releases the AC off of the pair. Just a thought, Thanks

I doubt it. AC induction builds on unterminated pairs.

What type of tests did you run? If you haven't done it yet, I would suggest a QRS test. It will pick up physical problems with the span. Another possibility- how about the electrical service feeding the smartjacks? Have you tried a UPS or a Dranetz meter on that?


Sam

How about an upstream mux? Are these 6 circuits connected to a common multiplexer in the central office? Maybe the common mux is causing the problems on all 6 circuits. When you reseat or change the card on a circuit you may just be resetting the error registers in the card and allowing the circuit to come up until the error threshold is reached again.

If the loop light on the smartjack is blinking, the smartjack is loosing sync with the T1 equipment at the office before it ever gets to a mux. From your description, I'd bet money you're looking at HDSL2 (Adtran H2TUR) or HDSL4 (Adtran H4TUR).

In most cases, resetting the smartjack "shocks" the cable pairs and returns them to service. In dealing with Southwestern Bell (now AT&T) their cable guys have reported this, in some instances, as a high resistance open.

The Ameritech techs should be able to login to the smartjack and pull the PM's, which will verify your trouble and give them an idea where to start.

Great input guys, Let first start with the type of test I am running is a Qrss, and the smartjack does not have any power except for the t-1's. No Mux these circuits go from main co transport to node to soneplex and out to customer. I have thought about high resistance open but I have changed cable pairs countless times in that 18 pairs, on different circuits. As of today we have been checking grounds on every span of cable and in the enclosures, everything seems to be good. Today I also made one circuit a four wire t-1. What I am finding out though adc is the 2wire has a thereshold of about 10v ac and right now I am getting anywhere from 7.5 to 8.5 vac, I could be a little to close for comfort. Right now the co is running Qrss on the 4wire and after 6hrs there have been zero errors. Thanks for the help, I will reply whats going on tomorrow.

Do the circuits have a repeater in the middle of them?

The lenght of cable is 8900 ft I have had it to where a repeater is needed.

Also as suggested above have you pulled the PM's on the circuit to see exactly what type of errors are occuring when the circuit goes down.

As for testing Qrss is a good test for basic testing you sould also run All ones and All 0's if it is a B8ZS circuit. This will also stress the T1.

Also having a test set on each end of the circuit verse a loop back plug is prefered.

Well the 4 wire t-1's are not holding up. I was at the business yesterday and one of the 4 wire t-1's loop 2 went amber so I plugged into the face of the card and it was taken framing errors bad. Instead of resetting the card I sat there and watched the errors, the snr was 27 on loop1 and 10 on loop2. After about 15 mins loop 2 snr started going up and when it returned to 27 loop 2 went green again and everything was ok. While I sitting there wondering what just happened loop 1 went amber and it started taken framing errors also, and snr was 13 for loop1 and 27 for loop 2. After about 5 mins the snr balanced out and loop 1 went green again. Currently I have three 4 wire t-1s and three 2 wire t-1s at the customer. I have never seen a t-1 that will take that many errors and repair itself and start taking errors on the other loop and repair itself again. There are no repeaters on the circuits, I need to copy and paste the pm's because I keep forgeting what errors their taking but I do know it is mainly framing and severe framing errors. I will try running head to head and all 0's. Thanks everybody

When running all zero's just leave it on for a few minutes. The only reason to run all zero's to to verify it will pass B8ZS. All zero's is not part of a stress test. Tberd had a test that ran all stress patterns randomly and was a very good test to find intermitten problems. Only bad part is I can't remember what the test was called.

EDIT: to add this it's a pretty good explaination of what each pattern is good for.

Try head to head. 5 minutes of All 0s, 5 minutes of All 1s, 15 minutes of 3 in 24 and an hour of QRSS.

Sam

When the loop goes amber and drops out, you've got copper transport trouble. Have you had anyone with the LEC cable repair group take a look at it? If you can determine why the SNR is 10 points off between the two spans, I'll bet you find the root of the problem.

What's the weather been like down there? A bad splice that gets wet and is crappy till it dries out?

When you go head to head - the side that takes the errors is the side with the problems. You may have to cut the circuit in pieces to figure out which segment is bad (assuming it's not equipment).

Sam

We have had some wet days but mostly nice and cool. I have had the cable techs look at the cable countless times and I was even lucky enough to have them out there right when two of the t-1's went down. They checked the cable with ohm meters and tdr's, we even put dial tone on the pair to make sure it wasn't going open and everything was clean and perfectly balanced. The customer brought up a good point, looking back at his trouble ticket logs, all of his reports show the circuits going down in between 10am and 1pm. I will definately run head to head but how good of a stress test is 55-octet? When the loop goes amber it never goes red or completely out. I would like to know what has a direct effect on the snr? When it comes to cable trouble I just can't understand how the pair comes right back and runs perfect for almost 24hrs. I am going to post the pm's so maybe somebody can help me understand them more. Thanks guys

HLXR PERFORMANCE SUMMARY

Elapsed time since counter reset: 23:08:10
HLXC HLXR
====== ======
HDSL MONTYPES LOOP1 LOOP2 LOOP1 LOOP2
===== ===== ===== =====
FCP 0 0 0 0
CVP 0 20 1 876
ESP 0 15 1 375
SESP 0 0 0 0
SEFSP 0 0 0 0
PA-H +14 +14 +14 +14
SNR-L +23 +24 +10 +14

Inewso, are these 6 T-1'r running out of the same cabinet? Maybe there is a problem with the backboard of the T-1 card housing box. We have weird problems before that were affected by temperature changing hot/cold and connections opening and closing. One site had a 25pr amp coming out of the telco feed box that could make and break connections by putting your hand on the amp connector. Try to get the telco to replace the card cage housing or moving one that has problems to a single card housing for testing purposes.....Also try to eliminate as many terminations points from telco as possible, even if it is temp for testing. Hope this helps.