Any tricks for finding a break in an underground wire?

[CaptDon]

First off, don't go 'thumping' the cable with high voltage capacitive energy. The cable is probably rated at 600vac and you don't want to do arc-over damage to the cable's insulation in unknown or undetected places as you will set up a new failure mode. The next thing is let's start simple, Totally disconnect both ends of the cable, have your friend ground one lead at a time. Start with red for instance. He grounds it at the far end with a piece of wire to any cooper water pipe or ground rod (B.T.W. there should be a ground rod at the sub-panel far end!!!) Then you measure for continuity at your end by placing one of your meter probes on your local ground and the other lead on the red conductor. You should see very low resistance if that conductor is intact. Do this also for the white, black and even the bare ground conductor. At this point you should have determined how many conductors are failed. If you are lucky enough to have at least one good intact conductor you could then measure the 'capacitance' of the bad conductor to a good conductor at each end and look at the ratio. Lets take a really simple example, you say your conductor is 350ft. As example if you measured 1000pf on one end and measured 2500pf on the other end your break would be 'approximately' 100 feet from the 1000pf end or 250 feet from the 2500pf end. If the cable has been ripped open and the conductors are exposed to ground/dirt/dampness/water it will be more difficult. B.T.W., the 74ACXXX series of parts are a pain in the ass and very touchy due to their speed. You can accomplish a makeshift TDR device with a simple function generator producing a squarewave, a bit of series resistance perhaps 100 ohms and a decent triggered sweep scope. The function generator only needs to produce a signal at around 10 to maybe 100hz. Trigger the scope on the rising edge coming out of the generator (use scope ch1) then look for the bump along the waveform on the end of the resistor hooked to the cable (scope ch2). Simple example, if the bump is 100ns from one end of the cable and 250ns from the other end then your break would be 100ft from the 100ns end or 250 feet from the 250ns end. You don't need to know velocity factor if you measure from each end. The break will simply be the ratio of the numbers. Hope this helps!!

[richnormand]

Similar issue about 20 years ago between the house and the garage at my parent's place.

Made my own TDR with TTL (as suggested previously) and it worked well. However do not forget to compensate for effective velocity in cables.
Other issue to be aware of is water ingress. At the break site humidity and water can penetrate a long way in the cable and cause problems later.

To get an idea on the path of the broken cable I used a transistor radio on the AM band and injected a weak signal modulated at 2kHz on an unused portion of the band.
Alternatively did that from each ends and each conductor. Fortunately the cable was not enclosed in a metallic conduit.

Good luck with it.

[SmallCog]

I have been using TDR to interrogate buried cables for over 40 years.   Solid aluminum coaxial cables are grouted into boreholes.  When the ground moves (landslides, subsidence. etc.) the cables are deformed and it is possible to measure the location and magnitude of cable deformation.

And I've grouted those sorts of cables in, in a past life with more calloused hands

I know this is the expensive option, but it sounds like this power cable is not only damaged but also a bit dodgy in its installation and will be prone to similar damage in the future too.

Dig a new trench (hire a chain trencher they're not that expensive) dig it deeper than required and either use sand or conduit to protect the new cable. Personally I'd use both and I'd oversize the conduit to make pulling easier.

Subject to local codes/standards/regs, whilst the trench is open think about what else you may need to run between those locations. At the minimum I'd be putting some irrigation pipe in there and capping the ends.

[jerryk]

I appreciate the additional new replies.  I built a TDR using @w2aew's video#88 as a guide.  It was a good learning exercise but could not get near the perfect peaks in the wave form for measuring the reflection timing that he demonstrated in his video.  It was good for an approximation but since we will be hand digging I'm hoping to get close to the break.  I also purchased the TDR mentioned by @sokoloff.  It seemed reasonably accurate on all of the test pieces that I measured.  The ratio of capacitance, as simple of an idea as it is, also seemed fairly accurate.

The sad part was by the time I received the TDR and got my stuff together winter came to Alaska with a vengeance.  We are still going to try to locate the break this spring when things thaw.  I will sure post what happens as I don't take the help I get here for granted.

Thanks - Jerry

[m3vuv]

What is uf cable,is it an american thing?

[jerryk]

It's a direct burial cable.  UF stands for Underground Feeder.  It's typically used to bring power indoors to outdoor locations like a remote light, electric gate, small out buildings, etc.  The one I'm trying to fix supplies electricity to a remote fueling tank for an airplane.

As for American thing, I would not know about that.

Jerry

[cdev]

I know you have a TDR, but not knowing how its display works, I thought I would toss this in. With the right transformer connecting it to the transmission line, A cheap nanovna is probably a better TDR than an official TDR unless that TDR is a really expensive device.

 Is there any reason a nanovna (1 or 2) could not be used? Sure, youre not measuring a coax, but I suspect that if a 1:1 balun was made and used the TDR function would work perfectly with any two conductor power cable.

A two ware transmission line made with speaker wire is around 100 ohms. You could probably estimate the impedance using a program called atlc, and build your transformer accordingly. (in terms of turns ratio) I would make a1:1 or 1:2  transformer and try that first. That becomes your calibration plane. Why not?

I've not personally used TDR techniques on UF or NM type wires, but here's a tool that you could try:

https://www.globaltestsupply.com/product/megger-1002-227-tdr500-3-handheld-time-domain-reflectometer

The only thing is that I don't know the velocity factor of UF cable, nor do we know what has happened at the point of damage, so some experimentation may be needed.

How deep is the wire buried?

[m3vuv]

sounds like  a shitty idea to bury a cable if it isnt armoured or in a conduit/ducting,you would get big trouble if you did that here in the uk!!.

[MarkF]

JerryK,
Here is another video by w2aew explaining the TDR measurement technique.
Near the end of the video he measures speaker wire (which would be similar to your case) with a different velocity factor and possibly a larger resistor.  It will make a big difference in the calculated length to the break.




You might want to test the calculations with some extra of the same cable type that you can measure the length to verify your calculations.

Also, you probably want to use two adjacent wires when testing since this technique requires two wire pairs or coax.

[GeoTDR]

Is it possible for a splice in a coaxial cable to modify the impedance of the cable downstream of the splice

[coppercone2]

this might be useful for those that want to make more advanced projects for cable locating (commercial interests)


§ 15.213 Cable locating equipment.

An intentional radiator used as cable locating equipment, as defined in § 15.3(d), may be operated on any frequency within the band 9-490 kHz, subject to the following limits: Within the frequency band 9 kHz, up to, but not including, 45 kHz, the peak output power from the cable locating equipment shall not exceed 10 watts; and, within the frequency band 45 kHz to 490 kHz, the peak output power from the cable locating equipment shall not exceed one watt. If provisions are made for connection of the cable locating equipment to the AC power lines, the conducted limits in § 15.207 also apply to this equipment.