Weird voltage drop in only PART of circuit

Reply

  #1  
Old 12-09-14, 01:52 PM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
Weird voltage drop in only PART of circuit

A few days ago our hall ceiling light stopped working.

There are two 3-way switches that control that light, so before replacing the two switches, I measured voltage across two of the three leads attached to each switch. On both switches the voltage measured approximately 90 volts.

Here's the first weird part. I removed both switches and measured the voltage across one of the switch's pair of leads. 68 volts. However, there was no voltage between any pair combination of the three wires on the other switch.

Then I installed two new 3-way swtiches. Again, 90 volts across a pair of wires on each switch, although the voltage is not between the same pair of wires on each switch (which I think is normal).

Not surprisingly, the incandescent ceiling light still did not illuminate with the two new switches installed.

There is an eletrical outlet near the two switches, so just for the heck of it I measured the voltage across the two inner contacts and, sure enough, only 90 volts. I plugged a light in there and nothing, just like the ceiling light above it.

Here's the second weird part. There are other items, such as lights and outlets on the same circuit (that turn off when the circuit breaker is opened) that operate normally. I measured voltage in one of the light-sockets whose bulb is operating normally and, sure enough, 120 V.

Of course all wiring between fixtures is covered by drywall, which doesn't help matters. I'm at a loss as to how to troubleshoot this strange voltage drop in only part of a single circuit, so I'd really appreciate some advice on how to proceed.

Thanks.
 
Sponsored Links
  #2  
Old 12-09-14, 02:33 PM
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 10,515
It is probably a loose connection or broken wire at the last working, or first non-working device on the circuit. It's a tedious problem to track down. The 90 volts is probably a phantom reading.
 
  #3  
Old 12-09-14, 02:38 PM
ray2047's Avatar
Group Moderator
Join Date: Mar 2006
Location: USA
Posts: 33,066
I measured voltage across two of the three leads attached to each switch. On both switches the voltage measured approximately 90 volts.
Generally no useful information can be obtained by doing that so the reading doesn't really mean anything.
Here's the first weird part. I removed both switches and measured the voltage across one of the switch's pair of leads. 68 volts. However, there was no voltage between any pair combination of the three wires on the other
. Again doesn't really mean any thing. Voltage is measured from hot to neutral or hot to ground.

You have a open or loose connection. It could be at the last working device or the first non working device. Move any back stabs to the screws and redo any wire nut connections. Have you read: http://www.doityourself.com/forum/el...ther-info.html
 

Last edited by pcboss; 12-09-14 at 03:03 PM.
  #4  
Old 12-09-14, 02:55 PM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
Thanks for the advice, guys.

First I'll read the troublshooting guide and then I'll redo all wire nuts in the swtich and light-fixture boxes.

I'll report results after that.

BTW, all wires to the switches are wound around screws that are then tightened. There are no "back stabs".
 

Last edited by Darwin's Child; 12-09-14 at 03:21 PM.
  #5  
Old 12-09-14, 04:31 PM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
The 90 volts is probably a phantom reading.
While my wife held the leads of the multimeter to two wires on one of the switches, as the multimeter was reading 90 V, I opened the breaker and the voltage dropped to zero. I don't know if that phenomenon was a phantom dying, but, regardless, I wasn't going to touch the two wires with my bare fingers until that 90 went to zero.

Now a bit more of what I think is important new information.

I mentioned previously that the reading between contacts within the outlet was 90 V. I just now measured the same thing, but after reading the above advice again, this time I also measured between the shorter-length-slot's internal hot contact and the ground of the plug-in and also the center screw of outlet cap. Both measured 120 V! No voltage between the round ground of the plug-in and the center screw. However, a bit more weirdness. From the longer slot to both grounds, I read approximately 27 Volts.

At this point I'm happy to be reading 120 V at least somewhere in a fixture that is not working properly (I tried the light in the outlet again and, again, no light.)

Tonight I don't have the time to do more troubleshooting or nut-tightening (although my wife's complaints about no hall light are doing some nut-tightening of another kind), but the fact that there is 120 V between the outlet hot and ground is certainly significant. I'll be taking the outlet and swtich covers off tomorrow and taking some more voltage readings. I'll report back what I find. (I would not be a bit surprised to see 120 V between the switches hots and grounds, just like the outlet's situation.)

Thanks very much for the advice so far. I already feel I'm making progress.
 
  #6  
Old 12-09-14, 05:12 PM
ray2047's Avatar
Group Moderator
Join Date: Mar 2006
Location: USA
Posts: 33,066
I read approximately 27 Volts.
Consider it ~0 for now if using a digital meter. It sounds like you have an open neutral. Your probably going to find the problem at a receptacle. Start at the first good receptacle nearest the first bad receptacle. If there are receptacles on both sides of the wall on the same breaker take that in to account when figuring which receptacle to start at. The path of daisy chained receptacles isn't always linear but it is a good rule of thumb to start with.
 
  #7  
Old 12-09-14, 05:36 PM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
Consider it ~0 for now if using a digital meter. It sounds like you have an open neutral. Your probably going to find the problem at a receptacle. Start at the first good receptacle nearest the first bad receptacle. If there are receptacles on both sides of the wall on the same breaker take that in to account when figuring which receptacle to start at. The path of daisy chained receptacles isn't always linear but it is a good rule of thumb to start with.
Will do. Thanks for the advice!
 
  #8  
Old 12-10-14, 07:41 AM
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 10,515
The last receptacle someone used a vacuum cleaner or a space heater in is a really good bet too.
 
  #9  
Old 12-10-14, 08:30 AM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
The last receptacle someone used a vacuum cleaner or a space heater in is a really good bet too.
How about a hair dryer?

You know, there is a bathroom within 6 feet of the hallway ceiling light, its two 3-way switches and the aforementioned outlet. Just inside the bathroom, on the right, is a pack of 2 switches (one for overhead bathroom lights and one for ceiling exhaust fan) and, just below it, a GFCI. These switches and GFCI are operating just fine, but, significantly, just happen to be on the same circuit as the items just outside the bathroom in the hallway that are not operating properly.

Perhaps even more significantly, my wife plugs her hair dryer in that GFCI just about every day and of course there is a certain amount of physical movement (along with heating and cooling) of the GFCI and perhaps the nest of wires and wire nuts inside its box and maybe even inside the switch box just above it, which really does have a rat's nest behind and adjacent to the switches.

Two probably bone-ignorant questions here.

First, I believe that any fixtures in the circuit that are operating properly must be before, or closer to, the breaker than those that are not. Is this true?

Second, with the main panel circuit breaker for this circuit open/tripped, should there be continuity between all the neutral wires of the entire circuit, regardless of switch positions and where the probes are placed? That is, if I put a probe in the longer (neutral) GCFI slot and then (for simple convenience's sake) insert the other probe into the hallway outlet's longer slot, there should be continuity, correct? (At this point I believe that the hallway outlet is the last/outermost fixture on the circuit.)

Thanks again.
 
  #10  
Old 12-10-14, 09:06 AM
ray2047's Avatar
Group Moderator
Join Date: Mar 2006
Location: USA
Posts: 33,066
believe that any fixtures in the circuit that are operating properly must be before, or closer to, the breaker than those that are not. Is this true?
Normally that is true but of course there is always wiring that made sense before the Sheetrock that doesn't seem to make sense after the Sheetrock.
should there be continuity between all the neutral wires of the entire circuit, regardless of switch positions and where the probes are placed?
Yes. Just remember though white wires of switch loops aren't neutrals.
 
  #11  
Old 12-10-14, 11:57 AM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
Well, to make a fairly short story shorter, the problem was sort of where I expected. It was not behind the GFCI, but above it in the two-switch box, inside a wire nut that hopefully holds together 4 white wires. This gaggle is pointed out by the red arrow in the photo below.
Name:  bathroom switches and GFCI.jpg
Views: 349
Size:  41.0 KB
Just for the heck of it, after pulling the switch cover off and dragging out the switches and wire gang, I closed the breaker, went upstairs and operated the first 3-way switch to the hall ceiling light. Sure enough, the light lit and the outlet had 120 V between the hot and neutral.

To be absoutely certain of what was causing what, I then grabbed the plastic wire nut and very slightly bent the wire pack behind it. The moment I started to move the nut, the light in the hallway went out.

Now, why that pack suddenly loosened up enough to open the circuit is beyond me, but, regardless, all I have to do is wrap that pack of wires more securely and that's the end of the mystery (although I'd like to know why digital multimeters can read 90 Volts when there is none).

I don't know how much money an electrician would have charged for doing what I did, but I'd bet it would have been more than $100.

Thanks very much for your advice on this phenomenon.
 
  #12  
Old 12-10-14, 12:28 PM
ray2047's Avatar
Group Moderator
Join Date: Mar 2006
Location: USA
Posts: 33,066
Those don't look like quality wire nuts. I would replace the wire nuts with quality wire nuts. Properly done wire nuts do not need tape.

A question usually avoided here because it is almost as volatile as religion or politics is to twist or not twist before putting on the wire nut. In this case because of the number wires I'd say twist first.

Wire Nut is actually a registered trade mark for a type of twist wire connector. If you go with actual wire nuts made by Wire Nut you should be good. One tip some cheap wire nuts don't have a wire spring inside. Those should not ever be used (IMHO).
 

Last edited by ray2047; 12-10-14 at 01:40 PM.
  #13  
Old 12-10-14, 01:05 PM
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 10,515
One of the wires in the bundle may be broken or burned off inside the nut. Best practice is to take the bundle apart and remake the connection after inspecting the wires for damage.
 
  #14  
Old 12-10-14, 01:05 PM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
Those Don't look like quality wire nuts. I would replace the wire nuts with quality wire nuts. Properly done wire nuts do not need tape.

A question usually avoided here because it is almost as volatile as religion or politics is to twist or not twist before putting on the wire nut. In this case because of the number wires I'd say twist first.

Wire Nut is actually a registered trade mark for a type of twist wire connector. If you go with actual wire nuts made by Wire Nut you should be good. One tip some cheap wire nuts don't have a wire spring inside. Those should not ever be used (IMHO).
I feel that I have to apologize for saying that I re-used the "wire nut" (which does indeed have the wire spring-type arrangement inside).

I removed the tape and nut and then unwound and straightened the four white wires. (Before straightening, I could see that the bunch of wires was not wound properly.) Then I cut back the insulation another 1/4 to 1/2" to have plenty of copper to twist.

Then I twisted not only the copper, but back another 1" or so on the insulated part of the wires. The twisted copper part is now wound rock-solid tight.

Then I twisted on the nut to super tight and, much as I hate to say it, to ensure that the nut would never loosen, I wound on some electrical tape starting on the nut and then down onto the wires some considerable distance and then back and forth several times. Now there is no way that the nut is ever going to come loose or the pack loosen enough to lose contact.

I put everything back in place, installed the cover, closed the breaker and now everything works perfectly. Not only that, but my wife was impressed, which is no small feat.

Thanks again, everyone.
 
  #15  
Old 12-10-14, 01:22 PM
Forum Topic Moderator
Join Date: Feb 2005
Location: Near Lansing, Michigan
Posts: 10,515
to ensure that the nut would never loosen, I wound on some electrical tape...Now there is no way that the nut is ever going to come loose
Let me stir the pot a little bit :-) It had tape on it before yet it eventually came loose, right?

Glad everything worked out and your hallway is now well lit.
 
  #16  
Old 12-10-14, 01:41 PM
Member
Thread Starter
Join Date: Dec 2013
Location: Canada
Posts: 109
Let me stir the pot a little bit :-) It had tape on it before yet it eventually came loose, right?
Yes, you are right and I get your point. But the nut did not come loose. It was the wound copper inside that was not of equal length that allowed one or more of the wires to pull out just a bit and lose contact with the others. (But, again, why this happened 36 years after that circuit was creaed is one mystery that I'm not going to attempt to solve.)

I realize that what I said had a sort of "famous last words" quality to it. I should have photographed the "after". Then, anyone with properly operating vision balls would agree that there is no way, short of a nuclear blast (which is nowadays an ever-increasing possibility), that that pack of copper is ever going to loosen.

Thanks again, all.
 
Reply

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Thread Tools
Search this Thread
Display Modes