Circuit tracing on old NM without ground

[wsalomon]

This concerns a call I received from a client and is on the "practical" side of what I do (EE).

The client has an office in his 200-year-old farmhouse. It was originally wired with BX probably in the '30's, which in the course of renovations 25 years ago, was 90% replaced with NM with ground along with some new work, all on Square D QO panels.

The garage area (originally carriage area and hay storage) had new work with after the original BX. This was done in ungrounded NM and in some locales, with steel boxes that were never grounded. When the renovations were done 25 years ago, most of the boxes (except one) had at least one grounded NM run to each box, grounding all boxes except one. This happened a new door was put, requiring the existing circuits be interrupted (with extensions around the new door) with two new junction boxes. Given that 3 circuits run through these in a 4" square boxes with a plaster rings to allow a single blank plate, access is "problematic". The client says the entire area was opened up (having been done in old varnished knotty pine) and then replaced.

The specific problem is this... the lights to garage office are on 3-way, 4-way, 3-way switch circuit with the last switch (at the end) with two traveler (red, white) wires, and a return (black) to 4-way's junction box to feed the lights. That switch has not been used in years (and happened to be the run done in "original" ungrounded NM). However, it was accidentally flipped one day, rendering the other 3-way and the 4-way useless. Thus led to a search as what caused the "outage" - a flipped switch in out-of-the-way garage corner (problem temporarily solved) but still in need of a "definitive" fix.

Measuring the voltage on the first 3-way and the 4-way was as expected (analog). However, the measurements at the remote 3-way showed the white traveler wire as expected (0 or 120 v when the other switches were flipped), but the red traveler wire was about 10 v (the ground reference was through the garage door frame, admittedly, not the best ground in the world, but the best available at the time). My thought about this was some inductive voltage from white traveler wire, or some kind of ground loop.

For simple tracing on non energized tracing, I have a Klein Tools Tone & Probe Test and Trace Kit (VDV500-705) and a Harbor Freight Cen-Tech Cable Tracker (94181).

I tried tracing the red traveler backwards, hoping to find where the interruption is, without going to the two 4" junction boxes, or either the near 3-way or the 4-way and unscrewing wirenuts and doing a simple de-energized continuity with a 40' piece of wire. Unfortunately, neither of these testers was terribly helpful, either using a single wire signal feed (red traveler), or a second feed (the white traveler or the black return).

Any thoughts on this. I'm not anxious to let my client know that the electrician who did this 25 years ago may have mis-wired this, but "it is what is".

Another question is - can one "date" when the garage area was wired with unground NM (roughly speaking)?

My thanks in advance.

 

[Seven-Delta-FortyOne]

On an ungrounded system, you really need to use a Lo-Z meter to get any kind of meaningful test results.

Have you tried pulling the 3-way and testing it to see if the problem is with the switch, not the wiring?

Thats where I would start.
Assuming I understood your situation properly.

 

[wsalomon]

Great thought about the switches - first thing I did was to replace them all (for the same reason you thought of).

Interesting thought about the Lo-Z meter. Given that most analog VOM's are relatively high resistance, this makes perfect sense about this being a "floating" wire (induction voltage).

Of course, the problem now is to find the other end.

 

[1.21gigawatts]

From your explanation I believe this to be a “ dead end” 3- way. Where the common wire at remote switch should have a constant 120v power regardless of the 4 way or main 3 ways toggle position. Based on your voltage measurement, I believe the problem is in the wiring of the terminals of the switches. Here is what I would advise.
confirm in the first 3 way box that there is: a 3 wire nm, a “power in“ 2 wire nm and a “switch leg“ 2 wire nm.

confirm in the 4 way there are (2) 3 wire nm cables.

confirm the remote or last 3 way “Dead end” has (1) 3 wire nm cable.

The connections would land as described below:

The incoming 120v power (white and black) black supplies power to common in 3 wire NM. The other two wires in the 3 wire NM terminate on the traveler screws. The Other 2 wire nm switch leg terminates to the common on the switch. Both whites (Neutral) of the 2 wire nm cables terminate to each other.

The 4 way switch has a common that passes thru the box ( does not terminate to the switch) and (2) sets of travelers that terminate to the 4 way device.

The remote or dead end 3 way switch has the common wire that was sent from the first switches Power source terminate on the common terminal. (Constant 120v power). The other two traveler wires from the middle 4 way land on the other two traveler screws on the switch.

on a side note, it was general practice in 3 wire nm cables to make the red and black wires the travelers and make the 3rd white wire the common.

 

[James L]

Probably won't be useful to just look for voltage. You need to make sure you have travelers through the 4-way and to the 3-way.

Then you need to make sure you have a feed at one end, and switch leg at the other.

Continuity testing is probably the easiest.

 

[ptonsparky]

Get a tone generator that is used to identify phone cables. Or others.

 

[James L]

on a side note, it was general practice in 3 wire nm cables to make the red and black wires the travelers and make the 3rd white wire the common.

I just saw your post and agree with uour explanation.

And I was thinking of the ungrounded NM, most likey 50s-60s, and how they mostly used black/red for travellers and white for common.

It's possible (maybe probable) that the switch was simply installed wrong 25 years ago

Good stuff

 

[ptonsparky]

Black and white were travelers. Red common.

Five hours travel time can’t make things that different. Can it?

 

[James L]

Black and white were travelers. Red common.

Five hours travel time can’t make things that different. Can it?

I've seen that, too, but mostly earlier. Maybe a hangover from guys who were doing California 3-ways using knob & tube

 

[Fred B]

Don't be too quick to jump on the original electrician, I've seen too many time a "handiman" get into a switch box and totally mess up a 3 or 4way switching that had been working, so that it might work most times but fail in 1 out of 12 switch cycles. Or even they found a jbox in an attic mess it up in that location. And even worse it was so messed up that it would trip the breaker when the 1 switch was put in that one position.

 

[wsalomon]

Guys - Thanks so much for your suggestions... There's a blizzard due tomorrow, so I'll be looking into all your suggestions.

BTW... in some parts of the system someone nicely marked the traveling or return white with a black felt pen. In the past, I have made that a habit for the next person to come. In the future, for a traveling white, perhaps a red marking (as suggested by the traveling red, if that's the way the system was done). It's always interesting how different "conventions" hold in different parts of the country. In renovating old houses, the "prime directive" is "don't work against the house". One learns how it was built (for whatever reason that seemed right at the time) and just carries that on (within reason) with Code updates when needed.

I'm just glad this house never went through intermediate knob-and-tube stage (had one of those personally in Northern California wired in the 1920's - a nightmare). Nor was it done in the "aluminum age" with its own headaches (owned two of those, remediated device-by-device).

One thing that was not thought about when the steel BX was replaced was doing it in aluminum flex. No doubt the thought was, being non-commercial, it was cheaper. In retrospect, with gnawing red squirrels in the attic, and the stone foundation, this was probably not the best idea. One place in the attic was partially gnawed upon (but only down to one wire), but was close in enough to a junction box that a quick paint with insulating goo and shrinkable tubing did the trick.

I usually work with data (Cat 5/6) wiring and labeling each end is a "must" in courtesy to those who have to follow 10 years hence.

 

[SSDriver]

On older houses I find a lot of cables/conduits all run to the light or a j-box in the attic with about double the box fill that is allowed. Then they make a bunch of connections in that one box. Maybe something went wrong in a j-box like that.

 

[wsalomon]

Fred - Understand perfectly. Unfortunately, there had never been a "handyman" messing with it since the renovation.

Pton - I did try the tone generators (see above) with the red dead wire alone, and both the white/black (de-energized). With both units (see above), the tone was stronger using 2 wires, much weaker with just the dead one. It just didn't seem to work either way. Not sure which is the best way to do this (with data cables, I can use both since they will (hopefully) end in the same place). Admittedly, this was my first time working in this situation (where only one wire ended, not both. Perhaps you might have a suggestion on how to do this better.

 

[wsalomon]

SSdriver - I think the "overfill" is what happened here, in both of the new work 4" boxes (as in why were there two new boxes instead of one?) - and it could have been fine had not the plaster ring been put in before closing up so only a single plate would be needed. It's why I was trying the "non-invasive" way before "unpacking" the "contents".

 

[synchro]

If the 4-way and the last 3-way are no more than 30 feet apart, you could consider replacing the 4-way with a Lutron Caseta wireless switch and put a Pico remote switch at the last 3-way location. That way the wiring to the last 3-way would not be needed.

 

[Fred B]

Unfortunately the toners will have trouble in a multiple switch system simple because of the nature of the switching process. Also even the higher end tracers have a hard time tracing through BX or MC to locate behind walls, the tracer produces an RF signal and the casing of the BX seems to block the signal.
Not sure if you've already done this but, try separating each cable grouping in each switch box (of course the switch also and lighting fixture) then trace each wire grouping, that will isolate each cable run to between boxes, and determine which set goes to which box. Once you have that information you can ohm the conductors switch to switch to see the lost or broken set. I've had to do this process more often than I like when a GC runs the wiring and his guys can't figure out how to get the 4ways to work and they've run the wire sequence A-D-C-B.

 

[hillbilly1]

If it just quit working after the switch was flipped, sure sounds like a crossed traveler and common? But I’m sure you already checked that. I had a neighbor who had two sets of three ways that didn’t work after he changed the switches. Knew exactly what he had done. I didn’t get over there until a couple of years later (LOL!) and he had the original electrician which was his brother come out. He couldn’t fix it! His brother was the one that I kept going behind and changing out 30 amp breakers on. I was dreading what kind of mess I was going to run into. I was surprised to find he actually used 12-3, although ungrounded, for the three ways. After I figured out how he ran the wire, had the problem fixed in five minutes. LOL!