Voltage increase when lights are switched off

[storeytime]

Need help. I?m doing a job in a metal warehouse. 12 highbay fluorescents(4 light 54W T5) on two circuits/switches. The 2nd set of 6 lights is fed through ?" conduit. The lights are on one circuit and there is a circuit for plugs. They share a neutral. (Single phase)

With the lights on, the voltage reading at the plug is 118 volts. When the lights are switched off the voltage jumps up to about 125 volts. We checked every J box and every fixture. There are no mixed up neutrals or switch legs. The circuits are fed from an existing panel. Two lights on the
1st set of six, which are on a completely different circuit and neutral don?t come on. (Actually one flashes on briefly, then turns off. These six lights aren?t on the same circuit and aren?t experiencing the same voltage fluctuation. Is it possible that bad ballasts right out of the box can cause this problem, affecting voltage on the other circuit? Sounds crazy, but I?ve checked everything I can think of to check.

 

[LEO2854]

Need help. I?m doing a job in a metal warehouse. 12 highbay fluorescents(4 light 54W T5) on two circuits/switches. The 2nd set of 6 lights is fed through ?" conduit. The lights are on one circuit and there is a circuit for plugs. They share a neutral. (Single phase)

With the lights on, the voltage reading at the plug is 118 volts. When the lights are switched off the voltage jumps up to about 125 volts. We checked every J box and every fixture. There are no mixed up neutrals or switch legs. The circuits are fed from an existing panel. Two lights on the
1st set of six, which are on a completely different circuit and neutral don?t come on. (Actually one flashes on briefly, then turns off. These six lights aren?t on the same circuit and aren?t experiencing the same voltage fluctuation. Is it possible that bad ballasts right out of the box can cause this problem, affecting voltage on the other circuit? Sounds crazy, but I?ve checked everything I can think of to check.

The load of the lights is reducing the voltage.

 

[LEO2854]

Two lights on the
1st set of six, which are on a completely different circuit and neutral don?t come on. (Actually one flashes on briefly, then turns off. These six lights aren?t on the same circuit and aren?t experiencing the same voltage fluctuation. Is it possible that bad ballasts right out of the box can cause this problem, affecting voltage on the other circuit? Sounds crazy, but I?ve checked everything I can think of to check.

Yes I've had bad ballasts right out of the box with these new electronic ballasts

 

[Smart $]

What kind of circuit length are we talking about here? Wire size?

Is there any load on the receptacle circuit while you are measuring?

Are you certain the lights and receptacles sharing the neutral are on different phases?

 

[storeytime]

I'm guessing about 70' to the point of the first light on that circuit, and then another 30' to the plug.

#12 THHN

No load on the receptacle.

Good point about the phase. No, I'm not certain. I picked up the circuits where they entered the building but I didn't make the connections at the load center.

One other thing. I didn't think these electronic ballasts that supply T5's got hot. As I re-checked the terminations at the fixtures, some of the ballasts seemed warmer than others. My first thought was that it was voltage drop, but the plug is on a different circuit than the fixtures, so the fixture load shouldn't affect the voltage drop on that plug. The only thing in common is the shared neutral.

 

[brian john]

Have you calculated voltage drop, that is step one.

Conductor length, amperage, voltage...............

http://www.csgnetwork.com/voltagedropcalc.html

 

[gar]

121006-0954 EDT

storeytime:

I am confused by your circuit description.

First, data from a cheap 4' T8 electronic ballast fixture with two 32 w bulbs.
123 V RMS input.
0.79 A
57.9 W
97.4 VA
0.59 PF

Assuming the same characteristics for your fixtures the calculations to estimate total current are:
4 * 54 = 216 W round this to 220 W
220 / 120 = 1.833 A for the resistive component of the input current
1.833 / 0.6 = 3.06 A for total input current

Six fixtures per circuit results in 18.4 A. This is excessive for #12 wire, but not your problem.


I got confusing information from your two posts on where the receptacle was located. Ignore that for the moment.

Number 12 copper wire is about 1.6 ohms per 1000 ft at 20 deg C. What I don't know is the length of wire to assume for resistance calculations. 70 ft is your guess to the first light fixture. Then 30 ft to the receptacle. This leaves me confused.

Need help. I?m doing a job in a metal warehouse. 12 highbay fluorescents(4 light 54W T5) on two circuits/switches. The 2nd set of 6 lights is fed through ?" conduit. The lights are on one circuit and there is a circuit for plugs. They share a neutral. (Single phase)

Size of warehouse is not specified. I will assume 100 x 100 ft. The above quote seems to say there are two light circuits, each with 6 fixtures, and a third circuit for receptacles. The source is single phase, but it is logical to assume a center tapped source exists. It appears that one light circuit shares a neutral with the receptacles. Further from your discussion I can guess the the shared neutral ends at 70 ft. That is, after 70 ft there are separate neutrals.

This seems to imply that the meter on the receptacle is reading just the voltage change along the shared neutral to the 70 ft point plus whatever hot bus voltage change occurs at the main panel as seen by the hot wire to the receptacle (the receptacle hot wire is only a long test lead back to the main panel if it is really a separate circuit with no loads except the meter.).


We need precise clarification of the circuits in the building. Also you need to make voltage change measurements at the main panel at the bus bars, and at the outputs of the breakers with all combinations of all lights on and off. No loads on the receptacles.

Make direct voltage measurements at the 70 ft point, and other appropriate circuit points.

Note: a single 70 ft length of #12 will have about 1.6 * 7/100 = 0.112 ohms resistance. A 20 A change thru 0.112 ohms is 2.24 V. If you had a 70 ft run from a perfect voltage source to a 20 A load, then the voltage at the load would be about 4.5 V lower than the perfect source.


No idea on the problem with the lights that don't work. Measure the voltage at the non-working fixture. Is it satisfactory or not? If good, then probably a bad fixture.

.

 

[storeytime]

Thanks GAR, some real good information there. I'll try to give some more info but I'll have to get some of the info you asked about tomorrow when I go back. I'm also attaching a drawing, but I made it off jobsite to the best of my memory. I will double check everything tomorrow.

Fixture amperage: I don't have the info here with me, but the lighting manufacturer sent the photometric/spec sheets before the job started and 6 fixtures after figuring continuous duty was only around 16 amps.

Size of warehouse is 44' long by 40' wide. From my drawing the last plug on the right is about 70' from the Jbox that is at the edge of the building. This is where the circuits enter the building. I am guessing at the length from there to the panel as I never saw that portion of the job. I am relying on 2nd hand info for that part of the job. I will verify that distance tomorrow.

In answer to all the voltage drop questions, I'm totally puzzled because the plug is not on the lighting circuit. I will verify all my information tomorrow and get back to this post tomorrow. Thanks for all the help.

 

[gar]

121007-2329 EDT

storeytime:

A smaller space than I might have expected.

Your 16 A estimate for the load current is not far from my 18 A so it is in the ballpark.

The receptacle circuit is the first item to check. Use an extension cord or just a long piece of wire to make a long test lead. Connect this to the neutral bus at the entry point junction box. Is this J-box just a juction box and/or a breaker box? Where are the breakers? I have to assume there are breakers in the building. The long neutral test lead should be first connected to the neutral bus at the breaker panel.

Take this long test lead to the receptacle you are making voltage measurements at. Now make voltage measurements of both the neutral and hot at the receptacle with the long test lead as the reference.

,

 

[storeytime]

I went back and personally checked everything.

2 lights not working - first one, lamps turned horizontally instead of vertically, 2nd - quick connector had loose neutral.

Plug with voltage variation - only had 1-2 volt drop. Distance was around 120 feet, so no problem there. Could not find any other problem at all.

Lesson - If you want it done right, do it yourself. Just kidding. I think my guy panicked when the lights didn't work, started checking voltage, just knew there had to be a crossed neutral. Started opening J boxes, checking voltages. Overlooked the fact that the problem is many times so simple you wouldn't think to look there first.

Thanks for all the helpful suggestions.