Lights with a flicker

Status
Not open for further replies.
W

Wireman340

Member
Location
Chico,Ca. U.S.A.
What would cause a laundry room light to dim momentarily when I turn the 220 dryer on. The same thing happens in the kitchen when the microwave get full current . Dryer and microwave both are dedicated separately, lighting is on separate circuits. The panel looks balanced as far as the breakers go, but my next move is rearrange some breakers. Another possibility may be from a previous addition done by a handyman, he may be sharing the wrong neutral. I have yet to go on an attic mission, but tomorrow is another day. Any suggestions from experienced electricians is respected and appreciated.
 
190327-0500 EDT

Wireman340:

It may be flicker or it may be dim. If it is a momentary short duration dimming, then it is flicker. If it is change to a lower level for the duration of the added load, then it is dimming.

Why? Think about your circuit. A power distribution system has various impedances as you progress thru the system from energy source to a load. A changing load current produces a voltage drop at every impedance. Relative to your final load point the impedances likely get smaller as you progress from the final load point back to the source. But they are there and produce some voltage drop.

For many purposes you can probably assume the voltage at the primary of your distribution transformer is constant relative to changes from your final load.

Consider a very simple circuit. A 100 V battery with an internal impedance of 0.1 ohm, two different loads, a 100 W bulb, and a 1000 W heater. The bulb is on so total current from the battery is about 1 A. Bulb voltage is about 99.9 V. Heater is turned on and battery current is now about 11 A. Bulb voltage drops to about 100 - 1.1 V = 98.9 V, and the bulb slightly dims.

Because you have separate breakers, and wiring from a main panel to loads does not mean that the input voltage at the main panel or bus bars is not changed from load changes on the separate circuits. There is common bus bar impedance, should be very small, main breaker impedance, not so small, power meter, all wiring from distribution transformer to main panel, and internal to the transformer, impedances, not so small.

What you want to look for are excessive impedances in the total circuit path. Use a meter with 0.1 V or 0.01 resolution to look at voltage changes, or drops with load changes.

.
 
The one sympton you do not want to see is dimming on one set of lights and brightening on another as the load cycles. That indicates a poor neutral connection and, IMO, a more immediate hazard.

Start at the service panel panel and check for VD there. That will determine if you need to get the POCO involved.

Maybe be the first question should be ‘how long has the problem been there? From day one or just this last week?’.
 
Different voltages on bus

Different voltages on bus

ptonsparky said:
The one sympton you do not want to see is dimming on one set of lights and brightening on another as the load cycles. That indicates a poor neutral connection and, IMO, a more immediate hazard.

Start at the service panel panel and check for VD there. That will determine if you need to get the POCO involved.

Maybe be the first question should be ‘how long has the problem been there? From day one or just this last week?’.
Click to expand...
One interesting point I noticed (aside from the temporary light flicker)is that the voltages at the panel on A phase are at about 120-125, and on B phase the voltage is at around 130-135, so the utility transformer is causing this difference in voltage.... is this normal?
 
Wireman340 said:
One interesting point I noticed (aside from the temporary light flicker)is that the voltages at the panel on A phase are at about 120-125, and on B phase the voltage is at around 130-135, so the utility transformer is causing this difference in voltage.... is this normal?
Click to expand...
Measure the L-L voltage. L-N voltage should be about 1/2 that and even if both legs are loaded the same. Turn the main off, measure voltages. Use a couple hair dryers or electric heaters of the same size and load up one leg, then the other. Take voltage readings.
 
Wireman340 said:
One interesting point I noticed (aside from the temporary light flicker)is that the voltages at the panel on A phase are at about 120-125, and on B phase the voltage is at around 130-135, so the utility transformer is causing this difference in voltage.... is this normal?
Click to expand...
Typically indicates feeder neutral issues.
 
190327-0857 EDT

Wireman340:

ptonsparky is pointing you in a likely direction, a poor connection in the neutral path.

The voltage imbalance you measured is quite unlikely. Do the tests Tom suggested. There are additional tests you could run, but his are sufficient to identify if it is a neutral problem coming to the main panel.

.

.
 
Lights with a flicker"act 2"

Lights with a flicker"act 2"

ptonsparky said:
Measure the L-L voltage. L-N voltage should be about 1/2 that and even if both legs are loaded the same. Turn the main off, measure voltages. Use a couple hair dryers or electric heaters of the same size and load up one leg, then the other. Take voltage readings.
Click to expand...
I checked all the circuit breaker terminations, made sure they are all tight, as well as all neutrals on the neutral bar. individually with all other circuits open, I tested the voltages on all panel breakers, all around 126vac. I checked the laundry room light circuit while turning the dryer on, the light circuit jumped to 136 v., then dropped right back down to 125v. after dryer start up. I tested the kitchen light circuit while starting the microwave, the light circuit voltage dropped to 112 v., then right back up to 126 v. keep in mind the dryer is on a dedicated 220 vac, and the microwave is on a dedicated 120 vac. Oddly both phase voltages are even now, unlike the day before when there was a definite difference.Upon checking the utility line power at the main breaker, I found the main breaker will not go into the open position, a definite needed replacement. Can this be the problem with the branch circuits?
 
ptonsparky said:
You need more than one meter or a set of test lights as the checks are made
Click to expand...

...I needed coffee and ran out of edit time...

You need to see the affects at the panel while the loads are cycled from L1-N and L2-N.

A typical dryer is a combination load of a 120v motor and a 240v heat strip. Yes, an individual circuit, but it affects the voltage of the whole house as your test proved. Most likely while your voltage raised on the lighting circuit (L1-N), L2-N dropped a similar amount at the panel or before. The heating element will affect your voltage readings as it will cause VD across L-L but also acts as voltage divider and would tend to even out the drop of L-N. Why we want 120v loads when doing these checks.

The dimming of lights as the micro cycles would mean they are both on the same leg at the panel. Both L1-N for example.

The main breaker is another problem but checking VD across it while testing will give you more things to think about. Line to load as well as line to line.
 
190329-0647 EDT

Wireman340:

You have been given testing information that you have not followed, and it appears that your background does not include much circuit analysis theory.

I will suggest the following procedure:

1. Turn off all breakers except the main.

2. Get two 15 W 120 V incandescent bulbs, and associated sockets, and wires. 15 W bulbs are easier on the eyes. These are to serve as instantaneous voltmeters.

3. Use a good digital voltmeter with at least 0.1 V resolution.

4. Get a 1500 W space heater. When hot has a resistance around 10 to 12 ohms. Produces about a 10 A load at 120 V. At room temperature the heating element is about 10 ohms on my test unit, and rises to about 12 ohms with 120 V applied.

5. Mount the two 15 W bulbs side by side so you can simultaneous see them.

6. Connect one bulb between the neutral bus bar, and the phase A bus bar. You can do this via an existing breaker by removing whatever is presently connected to the breaker. And do the same with the other bulb and phase B. Most people can see flicker in an incandescent at a change of 1 to 2 V. When looking for a neutral problem you are looking for an increase in one phase voltage and simultaneously a decrease in the other phase voltage. The two bulb approach is an easy test method because you can simultaneously see both bulbs.

7. In some fashion make it possible to connect and disconnect the space heater between neutral and one or the other phase. Basically you want a moderate load that easily can be switched on and off. Note: your 240 V dryer may have a 120 V motor and that would be why starting the dryer causes the motor inrush current to flicker a light. Can also be caused by high resistance in a hot line, but you already have evidence that seems to imply a neutral high resistance.

Procedure:

Watch the two light bulbs. Connect the heater between the neutral bus, and phase A. Does phase A bulb dim, and phase B brighten? If so, then there is at least a neutral problem before the point where you connected the heater load. If only the phase A dims, then there is a phase A hot line problem.

Assuming that we see evidence of a neutral problem, then it is clear that it is somewhere before the point where you connected the heater load.

You probably have no access to measure voltage before the main panel. So using the digital meter probes put the probes directly on the wires coming into the main panel, I mean on the wire directly, not on the lug.

8. Measure phase A to neutral and cycle the heater. Record the voltage change. Should be a drop.

9. Move the meter to phase B and neutral, but keep your changing load on phase A. Again record the voltage change and direction.

10. Move the meter to read between phase A and B and cycle the load. Record this change. If no major hot wiring resistance, then there won't be much change compared to what you see with the measurements to neutral.

11. If convenient put a screwdriver in the earth somewhere toward the power company transformer and away from the ground rod and/or where a water line may run. Measure voltage between this screwdriver and the main panel incoming neutral wire. A power company neutral problem should probably show as a noticeable change.

With this information you can probably judge whether the problem is before the main panel, a power company problem, or whether you are looking within an area where you can make other measurements.

.
 
Simple point: loads can cause momentary voltage sags, but only a poor neutral can cause voltage rises.
 
Wireman340 said:
One interesting point I noticed (aside from the temporary light flicker)is that the voltages at the panel on A phase are at about 120-125, and on B phase the voltage is at around 130-135, so the utility transformer is causing this difference in voltage.... is this normal?
Click to expand...

LarryFine said:
Simple point: loads can cause momentary voltage sags, but only a poor neutral can cause voltage rises.
Click to expand...

You can't get more simple than what Larry said.There is no reason to complicate things.

Once you know you have a bad neutral many times you can find it just with a visual of the cable from transformer to weather head. These days you need the power company to let you check out the meter connections ( and do a meter reseal ) so may as well have them check out the neutral.
 
growler said:
You can't get more simple than what Larry said.There is no reason to complicate things.

Once you know you have a bad neutral many times you can find it just with a visual of the cable from transformer to weather head. These days you need the power company to let you check out the meter connections ( and do a meter reseal ) so may as well have them check out the neutral.
Click to expand...

Hush your mouth! That is Blasphemy!! How can I justify all these toys, er ah tools, when you say things like that.:p
 
190329-2050 EDT

Data from my working system using a Fluke 27 and a 1500 W 120 V heater.

My supply is a 50 kVA transformer, wimpy wire from transformer for about 70 ft, 0000 copper for about 50 ft, bolt on meter, to QO panel with 200 A main fuses, and copper bus bars.

The transformer neutral is only grounded in three places --- a ground rod at base of pole that transformer is mounted on, at my incoming 1.25" over 100 ft copper water line, and at my neighbor's water line. My neughbor's water line is not connected by city water pipe to mine by less than 5/8 of a mile.

From prior knowledge the heater is close to a 10 A load.

Following are all within my QO panel.

At an instant of time phase A was 122.9 V and B was 223.1 V. A to B was 246.0 and equal to the sum of the two separate readings.

With 10 A load from A to N the voltage change for A to N was -0.9 V while the B to N change was +0.4 V. Loop impedance for A to N is about 0.9/10 = 0.09 ohms. Looks like neutral drop is about 0.4 V. Voltage drop across main fuse is about 0.03 V and we can expect some drop in the power meter. Watching A to B there is about an 0.4 V drop with the 10 A load from A to N. Everything looks consistent.

Voltage drop change across the A phase 200 A fuse is about 0.031 V for a 10 A change. Voltage measured from input lug to bus bar.

Voltage drop across QO 20 from bus bar to output terminal at 10 A about 0.077 V.

If there is a neutral only problem it will clearly show up. If only a hot line it will clearly show up. A combined hot and neutral problem will also clearly show up.

.
 
190329-2344 EDT

Little Bill:

Sure was a typo. Easy for me to do, and many times hard to find. Obviously should be and was 123.1 V. And I had double checked by adding the two, but it still escaped me. I guess I did not look close enough when adding 1 and 2. or did not actually look at those two numbers.

.
 
190330-1016 EDT

To add to my comments on measurements.

A transformer can, and is often times, modeled as a series resistor and inductor in the primary of an ideal transformer, plus a series resistor and inductor in each secondary. This also means that for a two coil transformer, a primary and one secondary, that the model can lump all the series impedance in either the primary or secondary. For a multiple secondary transformer the distribution should be between both the primary and secondaries.

In my measurements where the sum of the two individual secondary voltages exactly equaled the line to line measurement there could easily have been readings that might have differed by possibly +/-0.1 V or so from the digital meter alone, and of course there is always the power company voltage fluctuation. Thus, I was quite surprised at the exact match.

For simplicity I will model the transformer as two identical voltage sources connected in series (identical frequency and voltage, and a fixed phase relationship so as to add), and no internal impedance.

For this equivalent circuit I will label the terminals as P1, N, and P2.

P1 to N is 100 V, N to P2 is 100 V, and based on the above assumptions P1 to P2 is 200 V.

Assume three distribution wires from the three terminals. Each wire has a resistance of 0.1 ohm. And a test load of 9.8 ohms is used.

Label the ends of the distribution wires as P1L, NL, and P2L.

Open circuit P1L to NL is 100 V, P2L to NL is 100 V, and P1L to P2L is 200 V.

Place the 9.8 ohm test load from P1L to NL, and the current thru the test load is 10 A, and the voltage drop along P1 to P1L is 1 V. So there is also a 1 V drop along NJ to N. The load voltage is 98 V.

There is no load on P2L, and thus no voltage drop along P2L to P2. Thus, the voltage from NL to P2L is 102 V. The voltage from P1L to P2L is 100-1+100 = 199 V.

Change the neutral resistance from 0.1 ohm to 1.0 ohm. Keep no load on P2, and load P1L to NL with the 9.8 ohm resistance. Load current becomes 100/(0.1+9.8+1.0) = 100/10.9 = 9.174 A. The voltage across the test load is 9.174*9.8 = 89.91 V. The voltage from NL to P2L is 100+9.174 = 109.174 V.

The voltage from P1L to P2L is 89.91+109.174 = 199.08 V. This added to 0.9174 = 199.9997 or 200 V. Thus, a check.

Had there been more drop in the P1 to P1L line it would have shown up in these measurements.

Had there been a high resistance in the P1L to P2 wire it would not show up here unless greater than possibly 10,000 ohms. Thus, the test load test needs to be applied to both phases.

.
 
LarryFine said:
Simple point: loads can cause momentary voltage sags, but only a poor neutral can cause voltage rises.
Click to expand...
but Line to Line is still L1 to N plus L2 to N. If he has more than 250 line to line the POCO may have regulation issues.

Long service runs or even too small POCO transformer can cause severe enough unbalancing to look like bad neutral issues also, but line to line voltage will also drop when loaded in this situation.
 
Status
Not open for further replies.
Top