Finding fault location for broken power cables in walls

[calzap]

We recently bought a 25-year-old house from the original owners. They were pretty backward technologically and not great on keeping records on the original construction and subsequent modifications.  As typical in California, it was built on a slab with copper pipes running through the concrete.  And as often happens, a few years ago water started coming out under the slab.

So, they had the house replumbed with PEX.  This process involved plumbers tearing out drywall and drilling holes in walls and framing to run the PEX.  Then carpenters covered things over with some extension framing and new drywall.  There were no pictures, drawings or notes made.

I tested outlets in the house.  One was dead because of crummy hardware, and one was mis-wired (yep, installed by the owner).  However, one in the garage and one on an outside wall opposite the one in garage were dead as was one on an upstairs, outside balcony.  I’m presuming the plumbers or carpenters cut the cables (type NM) to these outlets somewhere. But where?

I have experience tracing underground cables using the 60 Hz emanations or by injecting RF and tracing that.  And I have equipment for injecting RF into house wiring for tracing cables.  Problem is that in this house, a lot of the cables run in bundles and the RF jumps from the injected cable to others.  The two garage outlets, inside and outside, are connected to each other.  They should be connected to a GFCI (RCD) outlet in the garage.  Neither an RF signal or back-fed power put into the orphan outlets makes it to the GFCI outlet.  Through inspection of wiring in boxes and deduction, I’ve narrowed the fault location to a 4 m section of wall.  But injecting RF and trying to trace it is useless because other cables in the wall pickup the signal.  So, there’s no discontinuity when moving the RF receiver across the wall.  It looks like a rewire job from the attic is needed unless the fault can be located precisely.  Any ideas?

The upstairs outdoor outlet is less of a concern.  I can create a new outlet further along the wall opposite an outlet on the inside wall and bring wiring across.

Mike

[BILLPOD]

Good Morning Calzap,  Several of my DMMs have a feature of sensing AC voltage when the meter, or a specific portion of the meter, gets near the AC voltage.  Maybe your DMM(s) have this feature.   If so, it might help you narrow down the location of where the voltage stops.
There are also small detectors that detect only, such as this Fluke FLK2AC/90-1000V Voltage Detector.  Try whatever detector you have, on live conductors that aren't behind a wall, to get a feel for the distances involved.  That might get you 'in the ballpark'.

[CatalinaWOW]

Put a fast pulse into the wire and watch for the reflection from the break.  Since a nanosecond is about a foot the timing will give the distance to the break.  There are oscilloscopes specialized for this.  With a fast scope and a fast rise time pulse generator you can do it without the specialty equipment, but it is challenging.  Among the problems is that the actual propagation speed is dependent on the cable type, and the value for NM is probably going to be hard to find.

[calzap]

These are good suggestions to consider.  I have inductive 60 Hz emanation detectors (one handheld standalone and one in an Extech DMM).  Unfortunately, they cannot pick-up a signal from wiring through 50 mm of air and 15 mm of drywall.  But the much bigger detector I use for underground wiring might work.

I hadn’t thought of pulse reflectometry and should have.   I have the equipment (generator with fast rise/fall times and 200 MHz scope) and have done it on coax before.  Strangely, it never occurred to try it on power wiring.  I wonder what velocity factor so use ... same as coax maybe.

Mike

[coromonadalix]

best luck would be to find tdr cable tester equipment ... you'll find the distance ...  but getting them ...

[jpanhalt]

Have you considered using a borescope camera?  You can get access through the outlet box, then move along.
I got this: https://www.amazon.com/gp/product/B0CJY5C9QX/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1  in June and it's OK.  The articulated tip and adjustable focus help.  Used it to trace some cables in a daughter's home built in 1950 (3-layer old-fashioned plaster walls).  With drywall, a small hole will allow access and is easy to repair.  If I was in the profession,, I would get a much more capable unit that could be steered..   

A second possibility is that the CB at the panel was either removed or disconnected.  Are there any unconnected black wires there?  Perhaps not very likely, but very easy to verify and fix.

[CatalinaWOW]

Velocity factor is probably closer to the old 300 ohm TV antenna wire.

[Kleinstein]

For a crude guess of the lenght of connected wire one can also look at the capacitance. The difference between cable types is surprisingly small. Its usually some 100 pF/m or 30 pF per foot.

With only a short distance / small area to look at the borescope looks worth a try.

[calzap]

This garage is a very electrically rich environment.  Among other things there is the service entrance panel, two subpanels, an automatic generator transfer switch, 3 load-shed boxes, water heater, whole-house vacuum, air handler, smoke detector interconnect relay, fire sprinkler flow switch, 2 garage door openers, inside and outside lights, and inside and outside convenience outlets. Just outside are two large AC compressors and a 22-kW generator.  All the wiring in the panels and boxes looks fine.  All CBs have wires of appropriate gauge, and the lug screws are tight.  There are no orphan wires in the panels.

All of the convenience outlets are supposed to be fed from a GFCI outlet on the east wall which is fed directly from a nearby subpanel.  There are outlets on the E, S and W walls, and I’ve confirmed all are fed from the GFCI outlet.   A cable from the GFCI outlet goes to the attic, and a connected cable there drops down to a single box on S wall and one down to a single box on the W wall.  Within each wall all the outlets are connected horizontally.

The dead outlets are on the S wall.  I know which box on the S wall receives a cable from the attic and have identified the destination of all the departing cables going to outlets on the S wall ... with an exception which goes in the direction of the dead outlets.  I also know the feeder cable for the dead outlets is coming from the direction of the box that receives the feeder from the attic.

So, I’m reasonably sure the fault is in a 4-m stretch of the S wall.  If I can narrow that down a bit with TDR or detection of 60 Hz emanations from back-feeding, then I could use a flexible boroscope to locate it precisely.

Thanks for all the comments.

Mike

[antenna]

4 meter is a little short for a 200MHz scope.  The fault could be hidden in the rising edge.

I have a couple techniques, depending on the situation, when the TDR route is impractical. 

My most favorite gadget is a Carlson Superprobe (diy project from youtube).  Instead of using a piece of coax as the probe, I added an SMA connector so I can change from an E-field probe (for open circuits) to an H-field probe (for short circuits).  Audio is a bit tricky getting any distance with, but if you AM modulate a LF carrier, you can easily reach through a wall.  Just don't go too high in frequency or it can couple to other conductors.

Another one of my tricks for really getting through walls, and this is primarily for following good cables, not necessarily for finding breaks, is to inject an HF carrier at the farthest outlet between the ground and neutral using a simple crystal oscillator and decoupling cap connected to an extension cord end and plugged in, and using a handheld radio scanner as a receiver with an H-field probe in place of the antenna.  In this case, since the detection is with the H-field, current flow is required.  Being the ground and neutral are connected in the box, that requirement is met.  Assuming your break is in the hot wire, this may still work in your case for following the path of the wire.

Edit: I mention the latter HF + H-field trick because knowing where the cable is in the wall greatly aids in finding the weak e-field signals on the broken wire.  As has been said, getting a tone from a few inches away is sometimes tricky, but knowing right where to look may help you pick up the signal after losing it in a corner or something.

Edit2:  and don't be afraid of trying to get a tone on the outside of the wall.  In some pre-fab or lazy builds, instead of drilling a hold in the stud, a notch is cut into the face of it and covered with a metal plate.  Often times, these are on the outside face of the wall,so getting a tone on the inside may be hard.

[jh15]

Before changing anything, do you have to get a permit first? Then a town inspector, or can you do anything and have a town inspector certify your work. Or do you have to have a licensed sparky in the works. If one is fully capable of electrical work, say industrial electrician, not licensed for outside work, but doing one's own home.

[Poroit]

G'day Calzap,

Would you consider cutting 100 x 100mm inspection holes in that 4 metre stretch of the South Wall?

If it is drywall ,then that is fairly simple to patch.

Yo might be lucky and find it first try.

[bostonman]

Put a fast pulse into the wire and watch for the reflection from the break.  Since a nanosecond is about a foot the timing will give the distance to the break.  There are oscilloscopes specialized for this.

I haven't done this personally, but I was with someone at my old job who did this. We were measuring the length of coax cable running in the building.

In any case, as you can see, plenty of creative options exist and this pulse suggestion is a really cool way of doing it. If I can recommend one simple suggestion and maybe it's not applicable. but sometimes I dissect problems by thinking about something logically such as: which route would someone run cable or you find a starting point and try to imagine the directions it takes.

Most likely this isn't applicable to your situation, but sometimes I just think about a manufacturing company and how they'd build something for speed or whatever the case may be.

[calzap]

Problem with inspection holes is there are vertical studs at about 0.5 m intervals.  So, a bunch of inspection holes would be needed for a 4-m stretch.  In this particular case, I think things are even more complex.  City water comes into the house in the same wall as the dead outlets.  That’s why the sprinkler flow switch is in it.

Based on certain structural features, it appears a lot of the PEX and fittings were fastened in or on that wall when the house was replumbed.  Then some new studs with lots of notches for PEX were fastened broadside to the wall, and those were covered with drywall extending floor to ceiling.  It’s possible that all or most of the original drywall was removed.

And of course, that’s the 4-m section where the fault probably is.  No surprise.  That’s where the plumbers and carpenters did their sawing, drilling and fastening.  So, best guess is there is a layer of drywall, 50 mm gap filled with PEX and fittings.  Then maybe another layer of drywall.  Then more PEX and electrical cables.  There is no knowing at this point whether the new studs line up with the old.  That’s why I’d really like to locate the fault as precisely as possible before drilling or cutting holes.

If I can’t do that, simplest fix will be to go in the attic and find the cable from the GFCI outlet.  Install a junction box and drop a cable through the wall to the dead outlets.  Haven’t been in the attic yet, so it may be more complex than described above.

Mike

[antenna]

Could you disconnect the suspect 4m piece at both ends, then make a loop at one end, then grid-dip it?  Maybe even use a nanoVNA and port extend noting the distance that way?  Even if the RF couples to other lines, I sure would think the stub resonance would be apparent. 

Never tried in practice though.

I was off by a decimal place on my TDR time, sorry, that scope may be fast enough

[antenna]

too bad it would be a fire hazard, could short other end and inject HV from a fencer or modulated tesla coil and listen for the audio where it jumps lol

[watchmaker]

Problem with inspection holes is there are vertical studs at about 0.5 m intervals.  So, a bunch of inspection holes would be needed for a 4-m stretch.  In this particular case, I think things are even more complex.  City water comes into the house in the same wall as the dead outlets.  That’s why the sprinkler flow switch is in it.

Based on certain structural features, it appears a lot of the PEX and fittings were fastened in or on that wall when the house was replumbed.  Then some new studs with lots of notches for PEX were fastened broadside to the wall, and those were covered with drywall extending floor to ceiling.  It’s possible that all or most of the original drywall was removed.

And of course, that’s the 4-m section where the fault probably is.  No surprise.  That’s where the plumbers and carpenters did their sawing, drilling and fastening.  So, best guess is there is a layer of drywall, 50 mm gap filled with PEX and fittings.  Then maybe another layer of drywall.  Then more PEX and electrical cables.  There is no knowing at this point whether the new studs line up with the old.  That’s why I’d really like to locate the fault as precisely as possible before drilling or cutting holes.

If I can’t do that, simplest fix will be to go in the attic and find the cable from the GFCI outlet.  Install a junction box and drop a cable through the wall to the dead outlets.  Haven’t been in the attic yet, so it may be more complex than described above.

Mike

You have my sympathy.  We went thru this with a wall sconce controlled by three switches.  After checking all 3 switches I called sparky.  He did determine the break was between the primary switch and the sconce (forgot how).  Still had to open the dry wall to install a new section.

May be easiest to just bite the bullet.  I would cut out the 12' section of drywall at the cable height (12' by 2') and then it is easy to replace the drywall.  You would also have the old holes for passing the cable.

Kind of like first aid, first you gotta visualize the issue even when you know it will hurt.

[themadhippy]

could short other end and inject HV from a fencer or modulated tesla coil and listen for the audio where it jumps

was thinking similar.personally id run a new cable in,or maybe

[timeandfrequency]

Hello calzap,

I could successfully locate a broken cable behind a 5 x 2,5 m drywall with this stuff. Could narrow down the exact location to 10 cm.
You need two transmitters, set with two diffferent codes. Additionnal transmitter here.

Connect the transmitters at both ends of the broken cable.
Follow the cable with the receiver, by beginning close to one transmitter. When the code displayed on the receiver changes, you're spot on.
Read carefully the manual. Best result are obtained when the mains is switched off, but it is not compulsory. The transmitters can be connected to live circuits.
Try with and without ground/earth connexions (the transmitters have two banana plugs, one of those is ground/earth).
Good luck.

Same gear from Fluke. Probably compatible with the one above (same features/ frequency). A bit more expensive^^.

[MarkL]

I use one of these telco line toners from Tempo Communications (formerly Progressive Electronics), model 200EP and 77HP:

  https://www.amazon.com/Tempo-Communications-701K-G-Generator-Probe/dp/B083Z7V7C2

The transmitter can be connected to a ground and the dead wire, and then traced by walking around with the receiver next to the wall.  The reception range is around a foot, but can be varied with the volume/sensitivity control.

This is meant to be used on ***UNENERGIZED*** wires.  If you leave the AC on, that will be the end of the transmitter.

Cost is about USD$80.  There are plenty of knockoffs for $35 or less if you search for "200EP", but I don't know how well they work.  I would stick with the original.

EDIT: It probably goes without saying, but where the tone stops or gets significantly weaker is where the break is in the wire.

[Gregg]

A neighbor completely rebuilt his garage and wired it with romex per code.  He called me in a panic because the electrical inspector was coming later and a breaker kept tripping for unknown reasons. The stucco installers had used long staples to fasten the chicken-wire for the exterior stucco reinforcing and had pierced the romex causing a short.  Of course drywall was installed on the inner walls.  I located the fault with an Amprobe CT-326 tracer meant for finding circuit breakers on live circuits. Unfortunately this model is no longer made and some of the replacements by Amprobe are not as good at pinpointing faults.
https://www.amprobe.com/product/ct-326-c/

[calzap]

Let me repeat that injecting RF won’t work for tracing because it can easily jump from a wire to an adjacent wire.  The broken cable is bundled with other cables along its entire length.  When RF is injected, I can trace the broken cable from one end to the other.  But if I back-feed power into one end, there’s none at the other end.

Here’s what’s going to happen.  I’ll back-feed into the ends one at a time.  I’ll use the 60 Hz detector I use for tracing underground cables (made by Radiodetection).  Work from each end and see if there is a common weak or vanishing point.  If no success with that, I’ll rewire from the attic.  Rewiring from the attic might be simpler and certainly cleaner than repairing the break in the wall … no drywall to cut and fix.  It will take two boxes to fix it either way.  In the attic, they’ll be unseen.  In the wall, there will have to be blank covers over the two boxes.

Although it might be more satisfying to locate and repair the break.  At some point, practicality must intervene.

Mike

[bostonman]

Haven’t been in the attic yet, so it may be more complex than described above.

Not sure where you're located or how sealed the house is (twenty-five years isn't that old), but, if you're in a warm climate, consider that you may have bees up there and need a quick escape before you climb through some small entrance.

Around here many houses have tiny doors in the ceiling of closets, so once you crawl through, you're kind of stuck up there. I along with many other houses installed a pull down attic door.

[calzap]

Haven’t been in the attic yet, so it may be more complex than described above.

Not sure where you're located or how sealed the house is (twenty-five years isn't that old), but, if you're in a warm climate, consider that you may have bees up there and need a quick escape before you climb through some small entrance.

Around here many houses have tiny doors in the ceiling of closets, so once you crawl through, you're kind of stuck up there. I along with many other houses installed a pull down attic door.

There's a metal hatch to the attic in the garage ceiling very close to wall with the dead outlets.  Main concern is I might not have  easy access to the top of the wall in the attic because it's a fire wall ... wall between the garage and the rest of the house.  May be OK because usual method is to double-thick the drywall on wall and ceiling and not have anything special above garage ceiling.  And that's probably why the attic hatch is metal.  Will probably have peek next week.

Mike

[Poroit]

G'day Mike,

If you find the location of the Break/Cut, I imagine you will still need to cut a hole at least 300 x 300mm to be large enough to access the cut ends.

There may not be enough slack in the cable for a simple butt joint. You may need to insert a short length between the two cuts ends.

Wago are good for that type of work.

https://www.wago.com/au/electrical-interconnections/discover-installation-terminal-blocks-and-connectors/221#durchgangsverbinder