Understanding the Cable Test Better

[mhoppes]

I'm trying to better understand the cable tester. For example, in this image, the cable is only 100 feet at the most - how is it testing at 620feet?

[sirhc]

Well I would say one of the following:

1) You are using an Ethernet surge protector which will mess with the test.

2) You have damaged cable

3) Your device is not yet powered and most unpowered devices will affect the length calculations depending on their design.

4) You have a damaged port

The main thing for running a cable diagnostics on an unpowered device is to insure no "CROSS SHORTS" exist but if you have POE smart enabled and are running v1.3.8 or newer it does this for you and now ONLY checks for cross shorts as earlier versions of POE SMART tried to check other things and resulted in too many issues.

[mhoppes]

1) None
2) Possibly
3) Could be - interesting
4) Tried multiple ports, not a switch issue

[lligetfa]

Matt, Is the cable a single contiguous piece or do you have keystones/patch panels in series? The transition from one cable to another through a patch panel can mess with the TDR measurements. A kinked or stretched cable can do that too.

That said however, my money is on #3.

[sirhc]

I agree and also place my bet on #3

[sporkman]

Similar topic, I'm looking at a switch with an AirGrid, a NanoStation and a Rocket M5. All are functional and seem to be working fine. No errors seen on the switch or AP end (CRC, short/long frames, etc.).

All "fail" the test and there seems to be an odd 10' difference in length between the PoE-carrying pairs and the data pairs.

edit: FW 1.3.8

[sirhc]

We are a little "strict" at what we call PASS on cable diagnostics.

Yes, when using older 10/100 devices where the manufacturer did not run pairs 3 & 4 through a center tap Ethernet Transformer you will see a pair length difference between pairs 2&3 verses pairs 3&4 especially on shorter cable runs.

Right now to get a PASS we expect OK on all 4 pairs and all 4 pairs to be the same length, as I said we are strict on handing out PASS.

However mind you this PASS or FAIL has no consequences with POE Smart or any other switch function it is just there to alert you to something is not as it should be according to ethernet standards.

[mhoppes]

Chris, This is why I don't run cable tests before turning on my radios :) Because normally with a radio attached the cable test comes back bogus.

[sirhc]

Chris, This is why I don't run cable tests before turning on my radios :) Because normally with a radio attached the cable test comes back bogus.

REALLY? The whole idea of running the cable test BEFORE you apply power is to make sure there are no CROSS SHORTS between pairs which can fry the port and possibly the radio.

If you do not think taking an extra 30 seconds to run the cable test on new cables BEFORE and AFTER is worth catching a miss crimped end causing a cross short or any short well then I give up.

Running a cable test AFTER the device is turned up is for seeing if there is an issue which may cause packet loss. Things to look for here after power is applied are DATA pair lengths match. So if using 10/100 that is pair 1 and 2 and for 1G all 4 pair should match.

Every piece of equipment reports differently before and after power is applied which if you go back and read my earlier posts I tell people to get familiar with how each piece of equipment reports with and without power. Always test NEW equipment never used before first in the LAB to familiarize your self on what to look for.

Look, I provide the tools, rather you use them is at your own risk. Not everything is clean cut yes or no and if it was no one would need IT people as they could just do it themselves.

[mhoppes]

I'm not saying the cable test is a bad tool. Just that once it's plugged into an equipment piece the results get a little wonky.