Cabling test Failure

Hi,

We've been running our own Cat5 cables for about 5 years, for the first time today we had a failure that isn't a wire map problem. The test fail the attenuation test. Plenty of Googling has added to the confusion. We have run two cables, both taking the same path, and both fail.

Whats the best way of resolving this problem?

Would an incorrect NVP value cause this?

I've also read that undoing to many twists can also cause attenuation problems. Is this correct?

All advice would be most gratefully received.

Best Regards

Reply to
Ian Young
Loading thread data ...

Sounds like overlong, streched, kinked? or just cheap cable. Maybe the jacks aren't good. OTOH, running in iron pipe might hurt.

First, I'd repunch. Smartly.

I doubt it, but you can always fiddle. Weak batteries on a cheap sender might. They should shutdown when the voltage regulators fall out-of-range.

That should usually increase crosstalk more than attenuation.

-- Robert

Reply to
Robert Redelmeier

Undoing twists actually affects NEXT (and, consequently, ACR, FEXT and all other cross-talk related tests), not attenuation, so that possibility is out. Setting up a wrong NVP affects your length test as well as delay skew test, which is also different from what you've observed. However, if your NVP is WA-A-A-Y off, then the tool would get distance so wrong that by dividing good attenuation value by wrong length it could have arrived at wrong acceptable attenuation value for this particular link and fail your tests. If you keep your NVP between 0.67 and 0.69, you should be just fine no matter what is the actual exact NVP value for this particular cable.

Couple possibilities you might want to look into are:

#1: you could have pulled too hard on both cables when pulled thru a tight hole of sorts. If you have exceeded the standard max pulling tension of 25 lbs, the copper conductor could have elongated and therefore became thinner thus affecting the conductors' DC resistance and high frequency attenuation.

#2 the terminations were bad as in re-using patch panel ports for several times as well as simply not punching hard enough or straight enough at the jack or the patch panel end.

In other words, anything that affects resistance, such as bad contact, also affects attenuation tests.

#3 plugs on your test heads are worn off. They only last about 1000 connections or so unless you are using the latest and greatest heads from Fluke that last several thousands. Get new heads.

Once again, worn off plugs means bad contact and therefore high attenuation.

Good luck!

-------------------------------------

Reply to
Dmitri(Cabling-Design.com

How long is the cable run? Are you testing at both 100 Mbps and 10 Mbps?

Reply to
DecaturTxCowboy

And does it pass anything "electrically interesting"? 5' motor, power transformer, large magnets, etc...

Reply to
DLR

Hi all,

Thanks to all of you for your replies and my apologies for not getting back sooner. I'm going to reply to all your questions/comments in this one message.

Unfortunately these are the first two cables I have not pulled my self, although I did punch them in to the panel & sockets. Therefore I have no idea how hard the cable was pulled, some of those involved are well qualified electricians, others hang pictures with 6" nails:-). I had a look at the route taken and there is no problem with large electrical cable, motors etc. The cable is a little under 80m long.

I've tried repunching with no improvement. I also fitted a normal socket at the patch panel end to test whether the atmosphere* round the patch panel had deposited anything on the contacts. All the sockets etc we have used else where without problems. I retested with an NVP of .68 again no changes. I am testing at 100Mhz.

I ran a resistance test and cam back with the following results.

7,8 19.8 ohms 3,6 37.1 ohms 5,4 31.2 ohms 1,2 27.1 ohms limit = 20 ohms

Whether the limit (as listed in tester) is realistic I don't know, I couldn't find any suitable info on the net regarding the resistance characteristics of copper.

I downloaded a 650M file (multiple times) across the link and it was no slower than the rest of our network, although this test is not very scientific.

All suggestions gratefully received.

Thanks again

Ian

*We manufacture printed circuit boards and the cabinet is out on the shop floor where the atmosphere can have various vapours in it.
Reply to
Ian Young

Reminds me of when I had to test over forty cables that the IT guy said were pulled by an electrical contractor using a "come-along".

80 meters (about 260 feet) of a single 24 gauge copper wire is about 8.3 ohms. From your test results above, it looks like the tester is measuring 80 meters to the short at the far end on conductor 1 and 80 meters coming back on conductor 2, the you should have about 16-17 ohms.

Conductors 7 and 8 look all right, 3 and 6 are a tad high. Try using a standard digital volt-ohmmeter and see what you get.

Google for resistance of 24 gauge copper wire and that'll give you some good reference material.

Reply to
DecaturTxCowboy

attenuation

frequency

Another thing that might affect attenuation is the cable may have gotten wet from capillary action, and a lot of water is still inside. I had that problem with the old RG-62 cabling, 'cause it's hollow. In any case, just abandon or remove the bad stuff and pull a new one in.

Reply to
Watson A.Name - "Watt Sun, th

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.