lights dim with a/c unit

Status
Not open for further replies.
W

wishmaster68

Member
Location
Houston,Texas
I was at my sister-in-laws house yesterday and noticed that when the 5 ton unit outside comes on the kitchen lights dim. The unit is right outside the kitchen and next to the panel. Being 220 volt and no neutral, why and is there a way to fix this? Thanks in advance.
 
I don't see undersized wire as a cause for a voltage drop on the other circuits. If the wire is undersized, you limit the current flow and decrease the voltage drop that the other circuits on the system see. You do however increase the time of the high current inrush and that may make the light dimming more noticable.
 
The house in about 4 years old and has had no extra work done to it. The panel is right beside the unit which back right up to the kitchen where the problem is. Is this maybe because of the locations of the unit and panel. There is no other light dimming issues in the rest of the house.
 
090413-1045 EST

wishmaster68:

You need to analyze the circuit.

You can assume that the primary supply to the utility transformer is a very low impedance relative to the transformer and service drop, and that for short time periods the primary supply voltage is constant relative to the timing of the air conditioner turn on. Also I will guess that the A/C inrush current at turn on might be on the order of 200 to 300 A.

The circuit with which you are dealing consists of a transformer, service drop, meter, and various connection impedances. My guess is that these total to maybe 0.1 to 0.5 ohms up to the main panel. These impedances are common to both the A/C and the lights. 200 A thru 0.1 ohms produces a drop of 20 V. 20 V will produce a very noticeable dimming of an incandescent. 1 to 2 V is probably noticeable.

Put your voltmeter between neutral and the light circuit hot lead and observe the voltage change when the compressor turns on. You will not see the full change because it is likely short relative to the time constant of your meter.

You asked why this occurs even through you have no neutral to the compressor. The compressor current is still flowing thru the transformer impedance and service drop hot wire that is common to the light causing a voltage drop in the supply seen by the light. However, it is not causing any voltage drop on the neutral.

The problem could be worse than my estimate if there are bad connections.

You need to determine a realistic source impedance for the transformer and service drop, and the inrush current to the motor. Then some calculations will tell you if the measurements you make are to be expected. But you probably need better instrumentation that just a typical voltmeter.

Edit: The no other light dimming issue does not correlate with what is generally expected.

.
 
The proximity of the kitchen light and the AC unit is not likely to be an issue (unless something weird like the light being fed from the AC circuit was done.) It's more likely that the dimming is noticed because the AC unit can be heard starting at the same time.

Some dimming is normal. It's hard to give an opinion without seeing it happen, but more than once, I've been called to price a service change because of this, and after seeing the dimming, advised the customer that the minute amount was normal and no issue.
 
I want to thank everyone for their quick and professional responses. I have been reading this forum for some time and thankful for all of you here. I will go back to her house later and check it out. Thanks again to everyone.
 
gar said:
090413-1045 EST
Also I will guess that the A/C inrush current at turn on might be on the order of 200 to 300 A.
Click to expand...

yeah. what he said.....

two other things that might affect the perception of the dimming....
lighting type, and quality of dimmers on the circuit affected.

incandescent lighting is affected by voltage sags more than other types of
lighting. I squared R losses are noticeable.

i high quality dvm like a fluke 87 or similar will allow you to capture spikes
and dips as brief as 1 MS. it's amazing how much a 2 hp motor will spike
a current flow when it starts. sampling at the 100 MS sample width, will
usually give a 10 fold increase in the running current draw of the motor.
1 MS is even higher.
 
090413-1221 EST

If you replace an incandescent with a GE "dimmable" CFL, then you may not notice any dimming. These CFLs, with sine wave excitation, have almost constant light output from 130 V down to about 100 V, then just quit. Whereas, a standard CFL has rather good dimming characteristics with sine wave excitation, and very poor dimming capability on phase controlled dimmers.

.
 
wishmaster68 said:
. . . the 5 ton unit outside comes on the kitchen lights dim. . .
Click to expand...
A 5 ton unit is considered to be a large AC unit. The normal electric utility transformer installation and secondary distribution normally is not sized for larger than 2 to 3 ton units. Rather than spending lots of time trouble shooting the entire system looking for trouble that is not there, call the serving electric utility and complain about the voltage flicker and dip. They will send a trouble man out to do a cursory check and you need to insist on a recording voltmeter for a longer period of time.

The serving electric utility needs to replace the transformer, the secondary, the service drop, or a combination thereof. :smile:
 
charlie said:
A 5 ton unit is considered to be a large AC unit. The normal electric utility transformer installation and secondary distribution normally is not sized for larger than 2 to 3 ton units. Rather than spending lots of time trouble shooting the entire system looking for trouble that is not there, call the serving electric utility and complain about the voltage flicker and dip. They will send a trouble man out to do a cursory check and you need to insist on a recording voltmeter for a longer period of time.

The serving electric utility needs to replace the transformer, the secondary, the service drop, or a combination thereof. :smile:
Click to expand...


LOL thankfully the OP doesn't come under our local POCO..it'll take years to get any action..we're still waiting for our AMR meters that were promised 4 years ago...:rolleyes:
 
Gar,

I like your 'ball park' approach to calculating undefined affects. :smile:
Your scenarios are so believable. You listed the whole chain of events.
You think .1 Ohms, etc.?
I could believe that, and it certainly gives life to the calculations. :smile:

But,
> ""1 to 2 V is probably noticeable.""
In my experience, this is way too small, but your 20 Volts should show.

""
>Put your voltmeter between neutral and the light circuit hot lead and observe the >voltage change when the compressor turns on. You will not see the full change
>because it is likely short relative to the time constant of your meter.
""

I use a clamp-on with an analog meter movement. Very responsive.
I also have used my O'Scope on one job, and it nailed the inrush voltage drop.
When I placed the O'Scope across the panel Bus to the Branch Circuit line,
I could read the Current of the inrush across the CB unit itself.
It was a bad, old breaker that had been tripped so many times
that it had arc pits on the contacts.

I wonder if the Dis-connect (outside in weather)
could have built up some impedance?
Need to clean the contacts??
I had one where water (from the roof with contaminants) had seep in.
Then during a Spring Cleaning, the water got onto the contacts.
 
Last edited:
090414-0758 EST

glene77is:

I will revise my figure slightly. A 2 V change on a 100 W bulb is detectable, a 4 V change is more noticeable. At some time I will create a more controlled experiment.

Two things I like to try are:
1. View a problem from an energy perspective.
2. Try to simplify the problem to the some essential points with appropriate assumptions.
Dr. J. G. Tarboux, my prof in DC and AC machinery, provided a true insight into the operation of these machines with these techniques.


wishmaster68:

A very useful experiment to run would be the following:
1. At or near the main panel find two circuits, one on phase A and the other on B. Hopefully these are circuits on which there are no changing loads. Neither of these circuits can be from the breaker that supplies the compressor.
2. Connect a 50 W bulb to each circuit. 25 W might be better because it is not as bright to look at. Or add grey filters to the lights.
3. Set the bulbs side by side.
4. Observe the relative change in brightness between the two bulbs when the compressor turns on.
5. If one bulb dims more than the other, then the phase associated with that bulb probably has a high resistance joint.

If one phase has a high resistance joint, then that might account for the comment that other lights do not dim.

.
 
charlie said:
A 5 ton unit is considered to be a large AC unit. The normal electric utility transformer installation and secondary distribution normally is not sized for larger than 2 to 3 ton units. Rather than spending lots of time trouble shooting the entire system looking for trouble that is not there, call the serving electric utility and complain about the voltage flicker and dip. They will send a trouble man out to do a cursory check and you need to insist on a recording voltmeter for a longer period of time.

The serving electric utility needs to replace the transformer, the secondary, the service drop, or a combination thereof. :smile:
Click to expand...

For once I agree with Charlie:smile:. Most of the time you will find that the majority of the starting voltage drop is in the utility transformer. The utility transformer is too small and may be 1 or 2 spans away. The VD in the transformer and secondary can be a significant part of the total VD. With the right type of meter you can capture the starting VD at the main. If it is high then the only way to correct the problem is call the utility.
 
Last edited:
I attended a Mike Holt class and Mike told us that he was having this problem at his own home. He checked all of the wiring and insisted that the POCO install a larger transformer. The lights still dim when the A/C units starts. With a 5-ton unit, I suspect some dimming will be noticeable.
 
bob said:
For once I agree with Charlie . . .
Click to expand...
Just once, Bob, my feelings are hurt.
icon9.gif


haskindm said:
. . . With a 5-ton unit, I suspect some dimming will be noticeable.
Click to expand...
As will any service with a fairly large motor, in relation to the transformer, starting, even with the transformer on the nose of the service. :)
 
Status
Not open for further replies.
Top