Lights blink every time a motor starts

[bcaward]

I've been called in to solve a problem in a house I did not wire. The lights all blink whenever a motor starts.

At first the complaint was "lights blink when my AC kicks on", so I checked size of conductors to AC equipment, grounding in panel, etc. Nothing.

Maybe it's load-related, I thought. The guy has like 250 lights in the house. But then he showed me that, with most lights off, he could still get the little stair-tread lights to blink when he turned on his disposal! I mean, come on.

I've checked every connection in the panel. . . all good. The local utility has been out and checked the distribution connections . . . all good.

The guy is willing to pay for a new, separate service, in case he's underpowered (which he's not), but what would I hook to it? Every motor in the house?

He asked a very good question: Why did this not happen in my old house? Why does it not happen at any of my friends' houses?

I don't have an answer . . . PLEASE HELP! Suggestions?

 

[Buck Parrish]

Did you check the main connections, in the meter base?
It sounds like a loose neutral or main phase.

Some times the poco will miss a loose connection. Esspecially if it is intermintently doing this.

If a cable from the poco is rubbing a tree, this could happen.

 

[hillbilly1]

How many houses does the utility have running off the transformer that feeds his house? How long is the service lateral or drop from the transformer? What KVA is the transformer? Service size? Info, we need more info!

 

[brian john]

One thought:

Light flicker (dimming and restoration of light to near preflicker level) of lights from motor inrush) is a byproduct of several factors and generally is difficult to over come.

AC motor loads (in particular compressors) but all motor loads have inrush currents. These inrush currents result in voltage drop in the branch circuit, panel bus, service laterals, utility transformer and possibility the utility HV feeders.

Flicker is in addition user dependent; some people are more susceptible to flicker. For some reason flicker seems to bother women more that men (based upon personal experience).

A Voltage Drop (VD) as little as 3 volts on a nominal 120 VAC system is noticeable (by me), 2.5% VD.

We have had cases were the VD was all a byproduct of primarily single phase 120 VAC loads, in these cases we were able to put the lighting on L1 and all the motor loads on L2, the HVAC compressors were something the homeowner had to live with. This minimizes the flicker.


To test for this VD and resulting flicker to determine the source of the inrush, use a min max amp clamp and min max multimeter at the main service watch the lights(it may take two workers) and correlate to the meters.

 

[beanland]

Voltage Dip

Voltage Dip

Whenever there is an increase in current there will be a reduction in voltage. The reduction in voltage is from impdeance. The impedance at a service is the POCO high voltage, the POCO transformer, the POCO service drop, the service entrance cables, the panel and breakers and the branch circuits. Most of the drop in a normal installation (no loose connections, no long runs) is at the POCO transformer. So, check each item. If the POCO transformer is small (15kVA) and the motor is large, the only fix is to get the POCO to install a larger transformer.

 

[hillbilly1]

My uncles automotive shop had such a severe voltage drop that when the air compressor started the ups unit for the computers would go into alarm. I Convinced the POCO that their 10 kva transformer that run his house, his neighbors house and his shop was a little too small! Up north they tend to load up as many houses they can on one transformer. Once they changed out to a larger transformer, he had no more problems with voltage drop. In town they would have as many as 5 houses on one small transformer.

 

[ptonsparky]

You need to put a VOM to this problem. Take your own voltage readings all the way out to the utility meter. Only then will you be able to suggest to the utility their transformer is to small. A new or additonal service won't help a thing it is wrong from the start.

We had lights on a radio tower that would cause the building lights to dim. We were several hundred feet away from the utility xfmr. but when I showed the utility we were getting 10 volt drop from one side of the road to the next, they increased the xfmr and overhead size to the meter.

Then again our utilites are public owned and seem to be very responsive, at least to electricians.

 

[charlie]

. . . they would have as many as 5 houses on one small transformer.

Five homes is no problem, neither is fifteen if the system is built correctly. You are correct in that the serving electric utility has to be involved. Calling the service dispatcher to get a trouble man out to do a cursory inspection is a good start. If he doesn't find anything, he should request a test to be run with a recording chart (actually, it records the data to a memory card that is downloaded so a chart can be produced).

If this doesn't happen, it is time to call serving electric utility's engineering department and ask for a test to be run. If it is obvious to the engineer what the problem is, he may write a work order to have the transformer replaced, the transformer moved closer to the home, the secondary size increased, split the secondary and add a transformer, the service size increased, or all of the above. :smile:

 

[winnie]

Two other pieces of the puzzle:

1) You need to educate your customer a bit. There will _always_ be some amount of flicker when a motor starts. As has been mentioned above, there will always be some impedance to the electrical supply system. If you change the current flowing through the supply impedance, then the voltage being delivered must change. If the voltage changes, the lights will change brightness. This is unavoidable.

At the other houses that the customer has experience with, this dimming has not been noticed, but it is still there.

The service that you can provide to your customer is to determine if the amount of dimming is greater than expected given their electrical system, and you can try to find the weak links in their service.

If the 'weak link' is a loose connection, then obviously that needs to be fixed.

But if the weak link is a 500 foot 120/240 service lateral because they've just bought a new house in the country, then perhaps the dimming is within the _expected_ performance of their electrical system.

If the weak link is an overloaded POCO transformer, then you might be able to exert some pressure to get the service improved.

But no matter what you do, if your customer goes looking for evidence of voltage drop, they will find it, because it is an inherent part of the system.

2) Certain forms of lighting, in particular _dimmed_ lighting is more susceptible to flicker with supply voltage changes. You might consider changing particularly sensitive loads.

If you find that the voltage drop is being caused by normally functioning components (eg. a long service run), then you might look at some sort of voltage regulation for sensitive loads.

-Jon

 

[bcaward]

More info:

More info:

Sorry. It's a 200 amp underground service. The lateral is about 200 feet long, but it goes to a hand-hole; the transformer is some distance farther. It's in a new development, and this is the only house currently served by this transformer.

If the extra distance to the transformer (the distance I didn't measure) were a contribuing factor, wouldn't that mean that the poco engineers screwed up and gave us a connection point that won't work? That's why I didn't measure it - I assume that the distribution system is solid. (If it's not, then with each new house they hook to that leg, it'll get worse and worse, right?)

The only thing I haven't checked yet is the terminals in the meter can. But I was reasoning that if there's a good enough neutral connection to carry the substantial load of air conditioning, hundreds of lights, etc., under normal operating conditions, then with almost everything turned off it should be a good enough connection to allow a small motor (disposal) to start without making one small light blink, which is what's happening.

I might be able to move all of the lights over to phase A and move as many motor circuits as possible over to Phase B, if anyone thinks that will help. Actually, I could try just moving the stair tread lights over to the opposite phase that the disposal is on, and see if they still blink, couldn't I? If it's going to work at all, it should work for that load and that motor, right?

 

[beanland]

Receptacle Tester

Receptacle Tester

I suggest that the next step is to test the impedance from various locations back to the source to isolate the problem. An Ideal receptacle tester or some such method for measuring source resistance is needed. Get a value for source resistance at the main panel and compare it to others in the house. If the source resistance at the main panel is high, then the problem is the POCO service lateral (200 feet is long) and secondary distribution system.

Basically, measure the voltage, turn on a known load (toaster?, hair dryer?), measure the voltage again. Divide the voltage drop by the load amps. That gives source resistance.

 

[winnie]

If the extra distance to the transformer (the distance I didn't measure) were a contribuing factor, wouldn't that mean that the poco engineers screwed up and gave us a connection point that won't work? That's why I didn't measure it - I assume that the distribution system is solid. (If it's not, then with each new house they hook to that leg, it'll get worse and worse, right?)

Unfortunately you are running into different definitions of what 'works'.

The POCO is permitted a fairly wide swing in delivered voltage. You should check, because there may be something broken and out of spec. But it is very likely that the flicker being seen is caused by voltage swings that are in the allowed range.

Things _might_ get better with more houses hooked up, if the voltage is somewhat on the high side and you are seeing some sort of saturation effect on an unloaded transformer. But my bet is that things will be unchanged unless the transformer is undersized for the development. And yes, things might get worse. (Sorry for the useless paragraph)

If you can convince the POCO that the voltage variation is out of spec, then you might be able to get them to change the transformer or the lateral.

You won't be able to move all motors to one leg, since any 240V motors will need to be on both legs.

How much total connected lighting load is there? (Just lights, not general purpose receptacles.)

-Jon

 

[Besoeker]

Maybe it's load-related, I thought. The guy has like 250 lights in the house. But then he showed me that, with most lights off, he could still get the little stair-tread lights to blink when he turned on his disposal! I mean, come on.

If the motor dips the voltage on starting it wouldn't matter how many lights were on.

 

[weressl]

If the motor dips the voltage on starting it wouldn't matter how many lights were on.

Of course it could. Elementary my dear Watson.

 

[kjfelec]

I had this problem just yesterday. Customer called to say that the lights were blinking when the washer was running. I went to the call and sure enough they were. Actually they almost looked like they were strobing.

Tested voltage at the panel and had a lower voltage from hot to neutral on phase A than phase B. It was 106 volts on A compared to about 122 volts on B. After going thru every thing inside I went outside to check the meter I noticed a freshly fallen tree limb. It looked like it came down right on the service drop from the pole to the house. I looked over at the pole connection and saw the messenger/nuetral broken right where they hi-pressed it.

Called the local POCO for an emergency trouble call and they came right out (believe it or not) and re-terminated the conductor. That solved the problem.

After talking with the customer a little more she said that it all started a few nights earlier when we had a nice rain and wind storm but thought the problem would go away.

Kelley

 

[Besoeker]

Of course it could. Elementary my dear Watson.

Care to expand on that?

 

[weressl]

Care to expand on that?

If you have load modeling program you can demonstrate it to yourself.

Try a 5kVA transformer.
Now add about 15-100W lighting load,
Now add a 3kW electric stove,
Now start a 2 HP motor on it and see the voltage stability at each point.

Reverse the order.

It is really elementary.

 

[Besoeker]

If you have load modeling program you can demonstrate it to yourself.

Try a 5kVA transformer.
Now add about 15-100W lighting load,
Now add a 3kW electric stove,
Now start a 2 HP motor on it and see the voltage stability at each point.

Reverse the order.

It is really elementary.

Okay........
Exactly how much difference would it make to the voltage dip transient as the motor is started?

 

[Besoeker]

And yes, I do realise that the transformer isn't rated for that continuous duty.

 

[robbietan]

Unfortunately you are running into different definitions of what 'works'.

The POCO is permitted a fairly wide swing in delivered voltage. You should check, because there may be something broken and out of spec. But it is very likely that the flicker being seen is caused by voltage swings that are in the allowed range.

If you can convince the POCO that the voltage variation is out of spec, then you might be able to get them to change the transformer or the lateral.

here in the Philippines, utilities are permitted to have a voltage variation of plus 10%/ minus 10% of the nominal voltage. If your contract voltage is 120 volts, expect a voltage anywhere between 108V to 132V when you measure at the main point of connection. we also have a flicker standard which means that you can complain of flicker if you have measured data from your house.

however, I side with the others that say the inrush current of the motor when starting that causes the voltage sags. since transformer capacity is not the problem, terminals comes to mind. and a separate circuit for the motor with correct wire size.