Lights dimming when dishwasher cycles

[Flicker Index]

Dishwashers generally use a permanent magnetic synchronous motor for drain pump and it is cycled repeatedly. The main pump is generally an induction motor but newer machines are using DC motors with rectifier.

LEDs are intrinsically extremely susceptible to flicker, because they have basically zero inertia. So little that you could modulate LEDs quick enough to not be noticed by people yet carry data in the light wave. LEDs are entirely dependent on its ballast to not pass on undesirable effects into light. Inductive kickbacks can cause LEDs to act out, especially those that is built with a dimming ballast even if it is not used with a dimmer. The LED ballast controller can interpret it as dimming signal and abruptly activate dimming.

Really all you can do is throw parts at it. Try different brands and different generations of lamps/bulbs/fixtures as each LED ballast design responds differently to interference.

 

[gar]

210110-1010 EST

Continuing my discussion.

Third --- What happens at the end of a branch circuit?

One of my benches is about 55 ft from the main panel, and fed with fairly large wire. Both phases are available, and share a common neutral.

Using the same 15 W bulb as a flicker detector, voltage detector, I get the following results with the 12 A change in current.

Phase A has the 15 W bulb. The 12 A load change on phase A produces:
Phase A drops 3.8 V.
Phase B rises 1.4 V
The 2 to 1 ratio of the voltage changes is as expected.
Flicker is barely noticeable.
Changing to a CREE 9 W LED shows no more flicker than the incandescent.

angus1 needs to see what happens to voltage at the main panel. If this is not great, then the problem does not exist prior to the main panel.

Then it is necessary to find a circuit that is not on the same phase as the dishwasher, and does not share a neutral with the washer. Put a light on this separate circuit and see if that minimizes flicker. It should..

 

[LarryFine]

LEDs are intrinsically extremely susceptible to flicker, because they have basically zero inertia. So little that you could modulate LEDs quick enough to not be noticed by people yet carry data in the light wave.

True dat. Automotive LED tail/brake lights are modulated for higher brightness with lower average power.

 

[gar]

210110-2141 EST

To add to my post #22 --- If I have no other loads on the neutral wire, and I use the EGC as a test lead back to the main panel, then I should be able to directly measure the voltage drop on the neutral wire. Doing this I also measure 1.4 V which correlates with the voltage determined in post #22.

.

 

[WasGSOHM]

Really all you can do is throw parts at it.

Or throw people at it. . .

>Automotive LED tail/brake lights are modulated for higher brightness with lower average power.

I hope they take advantage of the fact that the human eye responds to peak, not average, brightness. Maybe you said that.

 

[winnie]

>Automotive LED tail/brake lights are modulated for higher brightness with lower average power.

I hope they take advantage of the fact that the human eye responds to peak, not average, brightness. Maybe you said that.

I believe that there used to be a class of LEDs which were more efficient at a higher current level than the desired average operating level, so this modulation technique actually resulted in higher efficiency.

I don't believe that this is the case for LEDs used in the present day, but the industry is fairly vast.

Human eye response also enters into LED efficiency. A 100% efficient LED with an 800nm wavelength will not appear very bright!

-Jon

 

[gar]

210112-1546 EST

Further continuation on my previous comments.

If I look at a transformer alone, and directly at the transformer, then as an example for a "Signal Transformer" (name of company) A41-175-12 I get the following results:

This is a small transformer, 175 VA, 120 V in 12 V out. The primary and secondary are on the center leg of an EI lamination.. The primary is wound on one bobbin, and the center tapped secondary on another same size bobbin.

The test loads were 1 or 2 ohms.

A load change of 1 ohm on one half of the secondary produced a change from 7.32 V no load to 7.11 V with the 1 ohm load, or a voltage change of -0.21 V. The voltage change on the other half of the secondary, where there is no load change, from the load change on the first half is 7.32 V to 7,28 V, or a change of -0.04 V. If we assume this 0.4 V change is from voltage drop in the primary, then there is only -0.21 - - 0.04 = -0.17 actual voltage drop in the 1/2 secondary side that is loaded.

Next we load the total secondary with 2 ohms, same load current as in the loading of only 1/2 of the secondary.

The results are 14.67 V no load, and this drops to 14.19 V, or a change of 0.48 V. Of this drop we can expect 0.48 -0.08 = 0.4 V to be the drop associated with the secondary side.

An 0.2 V drop for 1/2 of the secondary from this approach is close to the 0.17 V drop from the first approach.

Note: the size of a transformer does not have much effect on the concept of what happens in the transformer for the same general design approach.

. ..

 

[jim dungar]

Note: the size of a transformer does not have much effect on the concept of what happens in the transformer for the same general design approach.

In my experience the size of the transformer impacts its nameplate and how it behaves.

Small transformers, like for door bells, often have a nameplate output voltage based on no-load conditions ( i.e. 16V nameplate for use with 12V loads). However a similar sized control power transformer has a nameplate showing its full load voltage. Small transformers, especially power style, often have compensating turns added to their windings.

 

[LarryFine]

I don't believe that this is the case for LEDs used in the present day, but the industry is fairly vast.

Next time you're behind a vehicle with LED taillights, sweep your eyes quickly from side to side.

 

[winnie]

Next time you're behind a vehicle with LED taillights, sweep your eyes quickly from side to side.

I know that they are modulated on and off, but I don't believe this increases the light output of the LEDs per watt. It may increase the perceived brightness of the LEDs.

-Jon

 

[LarryFine]

I know that they are modulated on and off, but I don't believe this increases the light output of the LEDs per watt. It may increase the perceived brightness of the LEDs.

I believe the main reason is brighter perceived light with less heat to dissipate.

I guess this is a way of saying more light per watt over time, i.e., averaging.

 

[gar]

210113-0047 EST

jim dungar:

You don't understand the point I was trying to make about the similarity of transformers of different sizes..

A power transformer can be modeled as a series input of a resistor and inductance. Then at this midpoint is a shunt inductance and resistor, then from the midpoint to the output is another series inductance and resistor.

For example see: https://www.electricaleasy.com/2014/04/equivalent-circuit-of-transformer.html

.

 

[jim dungar]

You don't understand the point I was trying to make about the similarity of transformers...

My comments were in relation to your experiment and measurements. When trying to duplicate results it is important to start from the same place The comparison of values from a doorbell transformer maybe different than those from a HVAC control transformer because of the slight differences in their construction and labeling.

 

[angus1]

Finally got back to that job
Measured voltage at both panels
New 320 continuous service I installed
125/126 volts on both phases in both panels
All connections tight
Swapped phase of dishwasher and kitchen lights stopped flickering
Didn’t have much time so I left
Now I’m told a different lighting circuit is flickering that was not before
Must be due to phase swap
Just a pain in the neck as homeowner thinks because I did a service upgrade all pre existing branch circuit problems should go away
Going back soon to investigate further

 

[LarryFine]

Swapped phase of dishwasher and kitchen lights stopped flickering
Didn’t have much time so I left
Now I’m told a different lighting circuit is flickering that was not before
Must be due to phase swap

Put the dishwasher circuit back where it was and move the first lighting circuit to other phase.

 

[angus1]

Good thinking
Thanks

 

[ptonsparky]

Good thinking
Thanks

I didn't go back and reread everything. Did the POCO update anything on there end?

 

[angus1]

Have not contacted them yet and they are very difficult to get in touch with and get a respence here in Boston

 

[gar]

210113-2208 EST

ptonsparky:

There is no need to bother the power company if there is little voltage change at the main panel on the phase the washer is on when the motor cycles. I have already provided information on the amount of voltage change to an incandescent that is needed to get some degree of flicker. From experiments I have performed before it is about the same for both incandescents, and LEDs.

.

 

[gar]

210114-1044 EST

jim dungar:

Whether you have a 1 W or 1 MW 60 Hz power transformer the "T" equivalent circuit is a fairly good approximation to a real power transformer. The size and specific construction of a particular transformer will determine what values to use in the equivalent circuit.

.In a lot of cases one ignores the center leg of the "T" in an analysis. Doing so reduces the circuit to an input series impedance, and an output series impedance. If there is only a single secondary, then both impedances can be added to make it a single series impedance reflected to either the primary or secondary side as desired.

If you have more than one secondary coil, then primary and secondary impedances need to be separate. Usually you would reflect the primary impedance to the secondary side for easy visulation, but they are still separate.

.