interesting residential issue

[mjmike]

I was working on a friends electric this past weekend and found it very interesting. The issue was no lights or power at a shed. Ok, sounds simple. The shed is fed underground with a 20A-1P branch circuit using #10 underground wire. The outside temperature is about 5 degrees. The wire enters the shed and splits to 2 different wires. the first set goes to 2 receptacles and the other goes to a switch which breakes the hot and serves a switched receptacle for an overhead flourescent light. Sounds pretty simple. I get ther and put my meter on the receptcale and get 120V and plug in my circuit tester and get 2 orange lights (this means ggod). I get an incadescent drag light and plug it into the first receptacle where I just read the voltage. When turned on, nothing happes liker there is no power. I plug my circuit tester in the top receptacle of the duplex. Turning the drag light on and off acts just like a switch turning the circuit tester on and off. I remove the tester and put my meter on it (hot to ground). With the drag light off, I have 120V, but when I switch the light on, the voltage drops to just about zero. Well, I definitelly believe there is a problem.

This is where I am confused, what would caust this to happen? If it was a loose neutral, the light would't turn on nor would you see a drop in voltage. I checked the splice at the main JB and all was good. I took the N and hot apart and re-did the wire nuts and checked power at the JB with the drag light. All was good at the JB. I also re-did the N wirenut at the switch. Put it all back together and the problem was gone. This one has me baffled. Also, when plugging in a small aircompressor, it will run with the voltage reading 120V. every once in a while it will choke and I will see the voltage drop to like 100V then jumps back up to 120V. If I plug the compressor in at the house, there is no problem. I am thinking because the shed is further away and it is winter, the shed is more susseptable to snow/ise on the overhead transmission lines to the pole transformer. Any thoughts?

I just found this whole thing interesting.

 

[gar]

110124-0930 EST

Your description is confusing.

My interpretation of what you said is this:

1. In the shed at a non-switched duplex outlet there appears to be near full voltage with very light load (meter or a circuit tester).

2. At this same outlet plugging a light, maybe 100 W, into the outlet causes the voltage at said outlet to drop to near zero.

Obvious conclusion is that somewhere in the hot-neutral path from the main panel bus bars in the house to the shed and any of those paths within the shed there is one or more high resistance points, and one or more may be intermittent.

This conclusion is based on no apparent problem in the house. Also there is no problem from the pole transformer to the main panel because you indicate the air compressor causes no problem when run in the house.

What is your experience in circuit analysis, and troubleshooting electrical circuits?

There are very simple procedures that will allow you to isolate the problems to a more localized region. At least if the problem is repeatable with any frequency. Very low probability problems can be very difficult to find, but there are ways.

.

 

[electricguy61]

Bad neutral, perhaps in the feed to the out building.

 

[gar]

110124-1055 EST

electricguy61:

Why is the response of most people a bad neutral? I have asked this question before.

I believe it was hurk27, in response to my question, that said that he usually found the greater probability of a failure in a hot lead.

The real point is that one does not have, a priori, a good reason to attribute the problem to any one place, and therefore a few measurements should be made to point one in a likely direction.

.

 

[electricguy61]

When you have normal voltage present (120v) with the very small load a digital meter presents, then a dramatic voltage drop or rise with higher loads, it typically means your load is "back feeding" through the other hot leg.

There are many individuals better educated on this forum that can explain this, but 30+ years in the trade tells me "bright or dimming lights" almost always equals bad neutral.

 

[ronaldrc]

Since this is a two wire circuit it could be in the neutral or hot. Sounds like your have a corrided connection
between house and shed.

Also if your going to continue to trouble shoot you need to invest in a good wiggy or selenoid tester.

Ronald

 

[LarryFine]

With the drag light off, I have 120V, but when I switch the light on, the voltage drops to just about zero.

This is where I am confused, what would caust this to happen?

That simply points out why many of us prefer a solenoid-type tester over a volt-meter for most troubleshooting. The typical volt-meter is too small a load to ignore phantom voltages.

That you read voltage with no load suggests the problem you accidentally found was in the grounded conductor, but not necessarily. A high-impedance meter may read voltage across a resistive connection, too. Comparing L-N voltage to L-g voltage also helps narrow it down.

A great troubleshooting tip for finding, not just eliminating the problem is to keep a drop-light plugged in while you look for a poor connection. When you find it, the light will begin to work. If you can't see the light, have a helper watch the light and alert you when it flickers.

 

[electricguy61]

Since this is a two wire circuit it could be in the neutral or hot. Sounds like your have a corrided connection
between house and shed.

Also if your going to continue to trouble shoot you need to invest in a good wiggy or selenoid tester.

Ronald

I han't caught it was only a 2-wire circuit.

You and Ron are correct that it could be either hot or neutral in this case.

 

[broadgage]

In this case, the voltage is near normal on virtualy no load, but drops to almost zero when an incandescent lamp is used.
Clearly a very high resistance in the circuit somewhere, could be either the neutral or the hot.
I would suspect water damage in a buried joint, with the wire eaten away by corosion, enough current to operate a tester might still pass via damp earth or the like.
Could be a poor connection somewhere else, but buried cables are vulnerable.

 

[gar]

110124-1328 EST

electricguy61:

There was no mention of brightening lights.

I agree that if you see a voltage on some circuit increase when a resistive load is placed on some other circuit that this probably and almost certainly implies a high resistance neutral.

.

 

[K8MHZ]

I took the N and hot apart and re-did the wire nuts and checked power at the JB with the drag light. All was good at the JB. I also re-did the N wirenut at the switch. Put it all back together and the problem was gone.

That indicates a likelihood of a bad connection, which you managed to fix.

This one has me baffled. Also, when plugging in a small aircompressor, it will run with the voltage reading 120V. every once in a while it will choke and I will see the voltage drop to like 100V then jumps back up to 120V

You may not have a large enough conductor with the 10 AWG to compensate for voltage drop. How much current are you seeing with both voltages? If the current increases as the voltage drops, you are likely seeing the resistance of the 10 AWG conductor. If it decreases, you may have a connection problem or a broken or compromised underground conductor. With my years of experience with air compressors, I suspect the former, voltage drop due to an undersized conductor.

 

[mjmike]

Thanks for the info guys. I am pretty good at troubleshooting, but can always use advise and learn more. The underground circuit was put in very recently over the summer and all worked good until the cold weather hit. When I redid a couple wire nuts, I got it fixed, but it still eludes me. If the problem comes back, what simple tests can I perform, assuming the underground is intact and the house electric is fine (fact). However the house is a trailer with the main panel is outside and this is where the branch comes from. The outside panel busing does show corrosion.

 

[gar]

110124-1931 EST

mjmike:

You first select a known good voltage reference point. I would use the main panel neutral from the panel as this reference. It is possible to have a bad neutral from the transformer to this point. Assume that is not the case, but it will become obvious if it is.

For this connection to the main panel neutral you may be able to use the outside the ground rod, or a cover screw on the main panel, or a close by EGC point on an outlet. You can do tests to verify that whatever you choose is likely to be a good reference.

Connect a long enough wire to this reference that will serve as one input to your meter. One or more 3 conductor extension cords may serve this purpose and also provide a convenient connection to the reference by simply plugging into a suitable socket. For the tests use a meter that can resolve 0.1 V on a range that reads above 250 V.

A convenient test load is a small 1500 W heater. This is about 10 ohms and applies about a 12 A load on a 120 V supply.

Use your meter to test various circuit points with and without the load. Try to have the load at the most remote point of the circuit of concern.

For problems like you had here it will be very obvious where is the problem. Depending upon wire size and wire run length will determine what voltage changes you would expect at different points.

For example a 100 ft length of #12 Romex, about 1.6 ohms/1000 ft, would show a voltage change of about 1.9 V for a current change of 12 A along one conductor, and 3.8 V if measured across the load.

When I start my DeWalt radial arm saw on 120 V at the end of about 130 ft of #12 I get upwards of a 40 V change from 120 V to 80 V across the saw motor from no load to starting load.

.

 

[readydave8]

don't look for something mysterious, find the bad connection first and then try to understand what happened and why.

 

[stevebea]

That simply points out why many of us prefer a solenoid-type tester over a volt-meter for most troubleshooting. The typical volt-meter is too small a load to ignore phantom voltages.

That you read voltage with no load suggests the problem you accidentally found was in the grounded conductor, but not necessarily. A high-impedance meter may read voltage across a resistive connection, too. Comparing L-N voltage to L-g voltage also helps narrow it down.

A great troubleshooting tip for finding, not just eliminating the problem is to keep a drop-light plugged in while you look for a poor connection. When you find it, the light will begin to work. If you can't see the light, have a helper watch the light and alert you when it flickers.

Good point Larry. I am surprised at the number of guys that I work with that have a DMM but dont own a solenoid type voltage tester.

 

[mtfallsmikey]

I guess Wiggys are good for something then....
I use digital Flukes at work, all analogs for home use, my fav is still my old Sperry Snap 6.

 

[LarryFine]

I guess Wiggys are good for something then....

One or two things. A voltmeter is the tool for reading exact voltages, such as measuring voltage drops, but a real(ish) load determines the presence of 'real' power, and relative voltage, such as 120 vs 208.

Discerning between 3ph 208v and 3ph 240v is easy by comparing all three lines and finding or not finding a high leg. 1ph 208v vs 1ph 240v is doable, but not reliably; then a voltmeter is the better instrument.

 

[Little Bill]

That simply points out why many of us prefer a solenoid-type tester over a volt-meter for most troubleshooting. The typical volt-meter is too small a load to ignore phantom voltages.

That you read voltage with no load suggests the problem you accidentally found was in the grounded conductor, but not necessarily. A high-impedance meter may read voltage across a resistive connection, too. Comparing L-N voltage to L-g voltage also helps narrow it down.

A great troubleshooting tip for finding, not just eliminating the problem is to keep a drop-light plugged in while you look for a poor connection. When you find it, the light will begin to work. If you can't see the light, have a helper watch the light and alert you when it flickers.

Larry
When you say a solenoid type tester, are you talking about an analog meter?
Also, what is a "wiggy"?

 

[kbsparky]

... what is a "wiggy"?

Click here for the official Square D spec sheet

Although I prefer to use Ideal's Vol-Con tester. Wiggy on steroids!!

 

[Little Bill]

Click here for the official Square D spec sheet

Although I prefer to use Ideal's Vol-Con tester. Wiggy on steroids!!

Thanks, I've been closed up in a plant too long and have stayed with my Fluke digital and Simpson analog meters. How much (approx.) do these meters run?