Cable tracer on unterminated CAT6A cable - unreliable, or signal bleeding?

[victorhooi]

I was trying to use a cable tracer (Amprobe AT-6010) the other day to trace some unterminated CAT6A cable.

This was unterminated UTP CAT6A copper cable - the wire itself is being used for Ethernet, as well as other uses (e.g. LED strip signals etc.)

I was using an crocodile (alligator) clip to clamp on the stripped conductor from one of the 4 pairs.

For the ground, I was just using the earth pin on a nearby outlet.

However, I noticed that the trace signal wasn't very reliable - it also seemed like there was bleeding from one cable to another, or sometimes the signal would randomly disappear every few seconds.

It's possible the crocodile clip was too large for the Ethernet wire, and not getting a good connection:



I did wrap it a few times around though - however, I've ordered some spring-loaded banana plugs, and also some screw terminal ones, to rule that out.

Is there something else I'm missing here?

I know there would be a cancellation effect if I was using another conductor from the same Ethernet cable - hence using the separate GPO ground. The twist in the Ethernet cable - I assume that wouldn't cause an issue? And this was UTP cable - so there shouldn't be any shielding to cause an issue.

Any other thoughts or suggestions?

[digger]

sorry for being off topic, but i'm shocked that UTP (unshielded) CAT6A is even a thing. sounds like an oxymoron. i thought the whole point of 6A was the shielding (and tighter twists).

did you read the manual for your tester? i haven't, but from my limited experience with cable tracing, i'm surprised you are not using a cable pair (instead using gnd from a nearby electric receptacle).

bleed is a thing, especially on long runs.

that's why the real way is to do wireless tracing for a rough idea but then do actual contact testing on the candidates distally. my cable tester changes the injected signal to a different audible pattern when the pair is momentarily shorted which is a useful confirmation. so i short the pair i inject into on the other end of the cable to be sure.

metal conduit also presents difficulties, but it sounds like you don't have to contend with that.

[SteveThackery]

Back in the day I used a "jointers oscillator" for cable tracing. You connect it across a pair, not with one leg to earth.  When you are tracing the cable, you can short out the pair on the probe tip - if it goes silent you've got the right pair.  The signal did leak into adjacent pairs, but shorting them didn't silence the signal.

Now, telephone cables employ a slow twist, so are much more leaky than ethernet cable. Nevertheless,  I would try using it across a pair first.  If you want more leakage, try one wire from one pair and another wire from a different pair.

[victorhooi]

The AT-6010 manual is available online here - https://www.itm.com/pdfs/cache/www.itm.com/at-6010/manual/at-6010-manual.pdf

On page 10, there's a mention of using a separate ground:



My understanding is that if I used a cable pair (or even just adjacent wires running together), the signals would cancel out - hence their suggestion of using a ground on a separate electrical circuit.

However, if I've misunderstood something here - please let me know?

With regarding the signal bleeding - this was only after 40-50 meters - would it be normal to have it happen after such a short distance? (The cables were very tightly bound together - cables ties 😒...I didn't run the cables).

And yes - I guess shorting the cables in a pair is more certain - is there a specific tester with that function you'd suggest here? Or would you just short one pair, then use a DMM or continuity tester at the other end?

I did try Googling for "jointer's oscillator" - I couldn't seem to find anything related?

[SteveThackery]

The "jointer's oscillator" is (still is, I think) a tool used by BT's technicians to trace pairs in cables. In theory you are right - the signal from adjacent wires would cancel out, but only if it is a perfectly balanced pair. If it isn't perfectly balanced - which no phone cable is - then it will leak to some extent,  and that is what the receiver unit picks up.

The pair under test has a differential mode signal on it, so shorting the two wires with the receiver probe kills the tone.  Adjacent pairs have the signal induced on them as common mode, so they radiate the oscillator signal but shorting them doesn't kill the signal.

I'm afraid I don't have any experience using other tracing techniques.  I wonder if you might need to take another approach, such as physical tracing or using the colour codes on the pair. You could use a multimeter in continuity mode with an assistant,  maybe.

Sorry I can't be of much more help.

[Gregg]

Wire tracing via tone signals is as much an art as it is science; practice and patience are required.  I have the Amprobe AT-6030 set but have not used it for data cable. The Amprobe works well for finding AC circuit breakers etc.  I have a Fluke MicroScanner2 that has never failed me for locating and sorting data cables. 
You might want to try connecting the transmitter to two separate pairs in the Cat6 cable which will tend to not cancel the signal as much as if you use a single pair. In other words strip the ends of one pair, twist them together for one clip lead and repeat for another pair.  This way you'll have the signal all along the length of the cable.  One big tip is to isolate the cable you are testing with the receiver from the rest of the bundle with your hand if you are trying to locate a single wire in a bundle. If the Cat6 were shielded, it might work better to connect one lead to the shield, but still connect the other to both conductors of a twisted pair.
As others have stated, the ultimate check for the exact cable is to short the wires at the opposite end from the transmitter and look for a loss of signal.