The builders cut an ethernet wire during construction. They put in the Foam Insulation and the dry wall up without testing the wires. Upon terminating and testing them, I found that the wire to my AV room has been cut. I've re-terminated and re-set the keystone twice. I can't run a new wire. I have to splice the old one.

I need to tone out the course of the wire (if a toner can even sense behind the drywall that far with UTP ethernet).

A TDR was suggested to determine the distance before the cut. I'd like to minimize the amount of drywall and foam that I cut into. Do you think that this will work? I can hire an electrician and his tool to do the work. I just don't want to waste their time and $$ only to find that a toner can't sense the wire that far or that a TDR isn't that accurate.


  • With so many answers I will say TDR is fine for cables you can access but in walls up across ceilings or under floors TDR is almost useless compared to a scanner that will detect where the cable is in the wall but an open or short. You can be off by several feet when you don’t have access. For a non pro I would suggest a basic cable tester that checks for opens , shorts crossed pairs, more than once I have found cross over cables in network connections and that was the reason the cable did not work. In your case it could be cross over or cut or open all are checked making the next step easier. – Ed Beal Aug 25 '20 at 22:55

I have a greenlee cs8000 that will work through just about any wall even through conduit, if your electrician has this or a similar tool they can pinpoint a cut or shorted wire. Time domain meters depend on the information being correct if you thought you had copper wire but it was copper clad Aluminum the distance would be off. Also you need to know the exact path for TDR or you will miss the mark.

Edit; I felt I should expand on my answer because there is some question in my mind if those commenting and or answering have ever used any of these tools I do not have have a tdr meter but used to use One to test coax on plasma etchers. And is an absolute waste in general construction because it provides a distance based on the wire type and size . I have the other 3 and cable tester.

A toner is not for tracing wires through a wall it is for identifying a conductor in a bundle for example a toner is used in a phone jack in an office building then you go to the switch and can find that wire quickly by running the toner down the punch down panel or in a loose bundle the range is less than 1/2” I have several of these progressive 77 , etch tek (cheap). Data cables Can be live not powered mains Depending on brand

A tracer like the cs8000 can identify a live or dead wire within 1/2” even through metal conduit drywalls and stucco. Any live or dead wire up to 600v direct connect and pickup scanner

tracers for buried wire if you know how to use them can get you within 6” . on wires buried 2-3’ I think mine is a 520 greenlee . Live or dead up to 600v suitcase or direct connect and pickup scanner

Electricians that have and use these tools with competence can identify broken / shorted conductors very close to the open or short.

I think some one mentioned a cable tester (transmitter receiver) you plug in that tests for the correct connections On finished cables they show opens crossed or incorrectly wired connectors I have a couple of these ideal brand is the one I use the most but I don’t remember the model This is the least expensive of the tools mentioned and if you have a mis wire or open at crimp this could be your problem. Anyone doing data should have one of these. Non powered single cable xmit rcv modules on each end I think the ideal cost ~50$

  • This is great. Thank you. I'll know exactly what to make sure that they have before they head over. – Daniel Woodward Aug 24 '20 at 23:35
  • 1
    Using a fancy cable tester/verifier at work I can tell you that for one cut in ethernet you'll get 4 different numbers because of the different twist (yes, the length of wire is accurate, assuming a good quality TDR. But the wire is not straight, and the twisting of each pair is deliberately different, so it's not uncommon for the length of a single, correctly connected cable to be several feet different on different pairs.) Also - one more reason to use conduit for networking. – Ecnerwal Aug 25 '20 at 11:40
  • Yes I agree time domain is great for verifying cables prior to install , I used them to get the high power coax cut to the correct length so the matching networks could get a vswr of 1.1:1 really important when pushing RF power. But in construction not useful because of routing , I did not think about the twist being different to limit crosstalk so that is worth a +. – Ed Beal Aug 25 '20 at 13:12

Warning: Splices are not what they used to be

  • Splicing a telephone wire - no problem at all
  • Splicing CAT 3 cable - probably work OK
  • Splicing CAT 5, 5e, 6 - Watch out

You might be able to use a splice junction box like this:

enter image description here

but (a) these are really designed to be used in an accessible location, not thrown inside the wall cavity and (b) they depend on having a few extra inches of cable to work with. If you have a cable that was strung relatively tight, there may not be any excess to use for a splice. Which might mean adding two splices and an additional section of cable. All of which can degrade signal quality.

Note that you can do this - there are no legal/safety concerns as there would be with 120V power. But "can" and "should" are often different things.

  • 1
    I do agree that splice blocks are the best way to go (direct punchdown) compared to a couple of RJ-45’s and a connector they end up being cheaper with a better connection and can be located in a wall without a box (once the cut is found if cut) testing the cable first may show a bad Rj45 crimp and a tester similar to an ideal link master or similar may identify a different problem than a cut cable. – Ed Beal Aug 25 '20 at 13:24
  • 1
    @EdBeal I've got a few testers of that sort. First was an STM-8 - 25 years ago? A couple hundred $ at the time, and worth every penny. Now I've got a Klein and a couple of cheapies. Of course, it is also possible for a builder to partially cut a cable - that can be really confusing to figure out. – manassehkatz-Moving 2 Codidact Aug 25 '20 at 13:46
  • 1
    Yes a partial cut is a possibility but I have had more connector failures than cuts (I thought those simple testers would be down to 25$ or less by now speery had one for 35$ ) my first one was black box brand back in the day five hundred or so might be a good thing to look for in eBay the only reason I got the ideal it is 1/4 the size and doesn’t require 8 batteries. The black box still works fine last time I used it has lasted forever late 80’s but slow it takes ~3 minutes to run a scan. – Ed Beal Aug 25 '20 at 14:09
  • 1
    @DanielWoodward You definitely would not want to use male-male here - those are great for getting extra distance out of factory-made patch cables. Punch down connections are the only thing I would consider relying on in a case like this. – manassehkatz-Moving 2 Codidact Aug 25 '20 at 17:42
  • 1
    I've only had one cut, the builder ran screws through my network cables, rather than crawl through the mud I used six splice blocks like pictured inside the wall. when it came time to terminate them I ended up with one cable not working, (possibly it was one where I didn't spot (or didn't rate) the damage on it and didn't splice, I had five other cables to that location, so it was not a big loss. – Jasen Aug 25 '20 at 22:23

TDR are very accurate depending, of course, on the calibration and quality of the unit. But they are easily accurate to within a cm or two.

The problem is determining where that lies in the building structure.

A "toner" as you call it, will NOT give you that kind of accuracy.

There are cable locator devices that may be useful here. These can send a radio signal down the line and can sometimes be useful in finding a fault or cut. In my experience with underground cable locators you might be +/- 1m or more.

  • I don't know the exact tool name, sorry. I've seen toners identify cables from within jumbles. If I can somehow use a tool to trace the wire's path behind the wall, I can use a tape measure to measure it's distance travelled. This tool would have to sense the signal at 3" or so. Then, I'd know where to start cutting. The wire originally travelled laterally in the basement about 20', up one floor, and to the structured data panel. Only the basement wall is foamed. – Daniel Woodward Aug 24 '20 at 18:13
  • What is the name of a tool that could sense a wire from behind the wall. I could make sure that the electrician has one before they come over. – Daniel Woodward Aug 24 '20 at 18:17
  • I'm not clear what you're asking about now. Your title says "TDR" which is an acronym for a TIME DOMAIN REFLECTOMETER. This is a device that sends a "ping" down a wire (i.e. a transmission line) and measures the time it takes to return. Using the time plus the wave propagation velocity, you can determine how far along the wire the fault lies. See: en.wikipedia.org/wiki/Time-domain_reflectometer – jwh20 Aug 24 '20 at 18:25
  • Thank you for the device name clarification. Here's my proposal: I'd like to have someone use a device to trace the cable's path behind the wall (up and over and up, etc). As I understand it, I could use a TDR to determine the distance to the fault. Then, I'd just measure along the marked path to estimate where I would cut into the drywall. Does this seem reasonable? – Daniel Woodward Aug 24 '20 at 18:47
  • That does seem reasonable, in principle. Reality often disagrees, especially when it consults with Murphy. – manassehkatz-Moving 2 Codidact Aug 24 '20 at 18:51

A continuity test of each conductor in the cable is not a bad idea: it sounds like you've assumed the cable was cut because Ethernet doesn't link, but there can be other causes of failure to link. Use a cable tester, or check end-to-end resistance of each conductor in the cable by using a multimeter and some variety of extension: a long patch cable, another piece of low voltage cable, etc.

Supposing you do conclude that all the conductors are severed, an underground cable locator might be helpful in charting the path and finding the break. It is much like a cable toner but costs US$4000 instead of US$200. The operation is similar: attach a transmitter at one end of the conductor; follow along the path of the conductor while watching the signal strength meter to determine where the conductor travels. When one reaches the end of the conductor the fields radiating from the conductor spread out and the peaks and nulls on the signal strength meter aren't so sharp. That's how one knows that he has passed the end of the conductor.

You might be able to find an excavator or private utility locating service who has a utility locator and would give it a try. I've never actually tried mine in a building before.. it might be interesting to see what happens!

That said.. a locating service might not be inexpensive. Compare it with its chance of failure or unfavorable information against the cost and guaranteed success of opening and repairing drywall to fully replace the cable.

Unfortunately a toner has to get very close to the conductor to pick up the signal. If you know the path of the cable and can touch it somewhere along the path a toner could be useful. Attach the transmitter box to one end of the cable, probe for tone at that accessible place, attach the transmitter to the other end of the cable, probe for tone again. This will tell you that you've got the transmitter connected to the right cable and whether a break in the cable is upstream or downstream of that accessible point.

  • Greg if you are paying that much for your scanners you need to check out greenlee eBay new cs8000. 550 and up I did not see any used ones that were for sale but I have seen them used in the past around 200. In ground scanner greenlee 521a multiple for 450+ on used ones , new 500- 550$ This guy goes deep 18” I can find the break within a shovel width 3’ sometimes I have to dig a few inches to the side but as you dig down and “dance the jig” To pin point it but within 6” if you get one you will understand dancing the jig moving and angling to pin point and even get a guess at the depth. – Ed Beal Aug 25 '20 at 23:26
  • @EdBeal I use the Vivax Metrotech vLoc3-Pro for locating buried utilities and for tracking a sonde (beacon) when directional boring. Deepest I've had to do with it so far is to confirm the blue stakes marks on an electric line (48" underground) and to follow my beacon when it dove to about 7 ft under. Thanks for the tip on the cs8000 though; it looks great for tracking wires in a building and I may add one to the toolkit. – Greg Hill Aug 26 '20 at 5:42

first you need to determine what sort of damage was done to the cable. is it shorted or open?

Then measure the capacitance (open) or resistance (short) from each end of the cable

the ratio between the two results will tell you how far along the cable the fault is.

  • I'm not sure why this was downvoted. It is correct, practical, and can be done with a $30 DMM. I can't speak for the accuracy, but again -as I mentioned in my answer- you can experiment on a sample piece and get a feeling for the accuracy. So +1. BTW, step 2 (measure) covers step 1 (determine), just in case a reader thinks this cannot be done without knowing what the type of damage is. – P2000 Aug 25 '20 at 20:01
  • It's not a literal answer to the question, it only answers the problem. that probably explains the down-vote. accuracy will be about as accurate as the meter. it can be done with a $10 LCR-T4. accuracy for shorts is dependent on good contact but the short resistance can be determined by probing with the far end both open and shorted. – Jasen Aug 25 '20 at 22:12

TDR is the right tool to determine the position of a cut, squeeze or short.

A TDR will not tell you where the wire is. So use a tracer to determine the route of the wire (if you don't have pictures), and use the TDR to know where the damage is along the route.

Issues that may murk the waters:

  • The TDR needs to know the velocity factor of the wire. If you have a sample, you can confirm it with the TDR
  • Hooks (short radius turns) near the point of damage may throw the measurement off, and the accuracy then depends on the sophistication of the TDR (pulse based, or correlation based).

Despite this, the method is very accurate and commonly used in telecom.

Velocity Factor

The TDR relies on a correct input of the Velocity Factor "VF" of the wire. For data wires it's in the range of 80% to 60% of the speed if light. (Maybe this is obvious, but for the general reader I'll add that this is not to be confused with the "data speed", as the two -data rate and propagation speed- are generally not related). A "faster cable" has to do with bandwidth and losses/attenuation, not the velocity of electromagnetic waves (although, the two are often related).

I'm not sure how familiar you are with how a TDR works, or whether the TDR you have in mind uses pulses or correlation sequences, but it all comes down to measuring the time delay of an electromagnetic reflection that bounces of a damaged, open or shorted section of wire.

The distance is calculated from the reflection, based on the electromagnetic propagation speed.

And the speed depends on the type/size/materials of the wire and to a small degree how it is deployed, and this is where a bit of uncertainty comes into play.

TDR with a Pulse or a Sequence

If the TDR uses pulses, then the shape (roundness) of the received pulse -due to the wire- will add some uncertainty, and a correlation based TDR would perform more accurately. A correlation based TDR sends a pulse train, and matches that against the received pulse train. It's very accurate even if the received pulses are rounded and otherwise reshaped. With a single pulse it's far more difficult to determine the delay if the pulse is rounded or reshaped.

Alternative to TDR

If you don't have a TDR (they are expensive), maybe it is more likely that you can borrow, from a friend, a scope and signal generator to perform a frequency sweep and obtain the reflection points based on the spectrum. You'd have to borrow the knowledge as well, to perform the right measurement and translate the result to a distance.

Practical Wiring Consideration

Remember, if you are cutting a hole in the drywall to confirm & fix the damage, you'll likely have 1ft of length exposed anyway, so depending on where the detected damage falls relative to any framing & passages, the accuracy can be moot (if you can cut a 12in x 12in hole) or critical (whether it's upstairs or downstairs).


You can calibrate the distance -reported vs actual- by letting the TDR measure the length of a sample of an open ended wire. Compare it to the actual length of the sample, and the relative difference (e.g. in %) is your adjustment factor. This works for well for minor adjustments well under 10%.

Use a sample of exactly the same wire (brand, type). Unroll it, and use a length that's close to the length of your install, down to maybe 1/8 of it if you don't have enough or if it's cumbersome, but do not use just a tiny left over cutting.

The calibration corrects for any discrepancies between the reported and the actual VF of the wire, so that you can be more confident about where the damage is, relative to your walls, floors etc..

VF vs Z0

As a side note, the Velocity Factor and the Characteristic Impedance Z0 are not the same thing but they are closely related to the physical properties of the wire (dimensions, conductivity, dielectrics). Just to note: "Z0" is not the resistance of the wire; so don't measure it with an Ohm meter and somehow try to calculate or correct the VF from it.


As with anything behind finished walls, luck is a factor, but remember the words of Edna Mode: "Luck Favors the Prepared".

Good luck!

  • More importantly it calibrates for the cable's "velocity factor", two cables can have exactly the same impedance but nevertheless have different propagation velocities if they use different insulation materials. – Peter Green Aug 25 '20 at 4:42
  • @PeterGreen good point. I edited my answer to relate the calibration to the VF not Z0. Thanks. – P2000 Aug 25 '20 at 17:10
  • measure from both ends, but this only works if there's a single fault. – Jasen Aug 25 '20 at 22:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.