Pulsing lights

Status
Not open for further replies.

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
210228-1146 EST

cfg:

Your post #16 I don't have a specific answer.

Following are some results from some quick experiments:

Equipment --- my 1950 GE photographic light meter (probably selenium photovoltaic photocell), a socket assembly with reflector, CREE bulb (2200 K, 800 lumens, 9.5 W, 120 V, 60 Hz, 79 mA) in socket, 123 V and less applied via a variable transformer, end of bulb 18" from photocell, and ambient light less than a fraction of 1 foot-candle.

Experiment Last --- Because of what I saw on my pulse tests I did the following test afterwards on the CREE bulb:

120 V --- 32 ft-c
100 V --- 32 ft-c
080 V --- 32 ft-c
060 V --- 30 ft-c
040 V --- 25 ft-c
020 V --- 15 ft-c


Experiments --- Pulse tests performed at 123 V source with fixed resistor shunted by a mechanical relay contact (silver material not silver-cadmium-oxide). Pulse control of of relay is electronic with both on and off times adjustable. Load 15 W incandescent 120 V in parallel with 9.5 W CREE. Other portion of cycle was 4 seconds,

Experiment 1 --- 20 ohm, 140 mS, 3.6 V drop. 15 W flickered, CREE no flicker.
Experiment 2 --- 20 ohm, 400 mS, 3.6 V drop. 15 W flickered, CREE no flicker.
Experiment 3 --- 20 ohm, -----1 S, 3.6 V drop. 15 W flickered, CREE no flicker.

Experiment 4 --- 60 ohm, 140 mS, 10.6 V drop. 15 W flickered, CREE no flicker.
Experiment 5 --- 60 ohm, 1 S, 10.6 V drop. 15 W flickered, CREE no flicker.

Changed from a normally closed contact on the relay to normally open to change from a voltage drop pulse to a voltage rise pulse. No change in the results.

Tried a FEIT LED bulb with about the same rating as the CREE with the same 60 ohm resistor, and visually the FEIT results were about the same as the 15 W incandescent.

.
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
LED flicker is a battle we fight constantly. I’ve had problems with customers calling in saying their lights are flickering at the same time we do PLC TWACS readings, dimmer issues where the electricians blame the POCO for the flicker, and one in particular I have found weird...

Customer has a remote on their fan. Works fine but when the power goes off and comes back on the lights flicker all the time
After a few instances of this and “fixing” the issue we assumed the remote defaults to 3/4 way maximum..
Because we can now tell them to make sure the remote is on maximum output. Stopped it so far...
 

kwired

Electron manager
Location
NE Nebraska
LED flicker is a battle we fight constantly. I’ve had problems with customers calling in saying their lights are flickering at the same time we do PLC TWACS readings, dimmer issues where the electricians blame the POCO for the flicker, and one in particular I have found weird...

Customer has a remote on their fan. Works fine but when the power goes off and comes back on the lights flicker all the time
After a few instances of this and “fixing” the issue we assumed the remote defaults to 3/4 way maximum..
Because we can now tell them to make sure the remote is on maximum output. Stopped it so far...
But why have a variable control if you must use it at maximum? You didn't fix any problem, you masked it with certain conditions.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
210228-1329 EST

cfg:

A DVM like the Fluke 27 or 87 is not fast enough in AC mode, and possibly not in DC, that you can correctly detect short duration voltage changes. Thus, in my above experiments I made steady state measurements across the series resistor to determine the voltage drop. Had I put the resistor in the neutral lead I could have used the scope for the measurement.

Your flickering light problem at the customer location can possibly be analyzed or at least changed by looking at exactly where the flickering lights are located in the branch circuits.

Based on the flicker I assume the lights are on the same branch circuit that the washer is on. Assume that assumption is incorrect because you can prove the lights are on a different circuit and breaker.

Based on this new assumption take one of the flickering bulbs, put it in a socket with some sort of test leads, possibly just probes or clips or possibly a plug. I would like you to take this light to the main panel and connect to the same same phase as the washer is on, but not from the washer's breaker. Do you see flicker now? Assume yes, then switch to the opposite or different phase. From previous statements you made I do not believe you will see flicker when supplied from a different phase.

Depending upon what the above tells you, then I might look at the washer branch circuit right at the washer breaker in the main panel. If there is no flicker at this point, or greatly reduced, then I would believe it necessary to search along the branch circuit for somewhere where the lights have wiring in common with the washer. Might be on common (neutral) and not the breaker.

You need to look from different directions to try to find what wiring, hot or neutral, that is shared by both the washer and the lights.

.
 

kwired

Electron manager
Location
NE Nebraska
210228-1329 EST

cfg:

A DVM like the Fluke 27 or 87 is not fast enough in AC mode, and possibly not in DC, that you can correctly detect short duration voltage changes. Thus, in my above experiments I made steady state measurements across the series resistor to determine the voltage drop. Had I put the resistor in the neutral lead I could have used the scope for the measurement.

Your flickering light problem at the customer location can possibly be analyzed or at least changed by looking at exactly where the flickering lights are located in the branch circuits.

Based on the flicker I assume the lights are on the same branch circuit that the washer is on. Assume that assumption is incorrect because you can prove the lights are on a different circuit and breaker.

Based on this new assumption take one of the flickering bulbs, put it in a socket with some sort of test leads, possibly just probes or clips or possibly a plug. I would like you to take this light to the main panel and connect to the same same phase as the washer is on, but not from the washer's breaker. Do you see flicker now? Assume yes, then switch to the opposite or different phase. From previous statements you made I do not believe you will see flicker when supplied from a different phase.

Depending upon what the above tells you, then I might look at the washer branch circuit right at the washer breaker in the main panel. If there is no flicker at this point, or greatly reduced, then I would believe it necessary to search along the branch circuit for somewhere where the lights have wiring in common with the washer. Might be on common (neutral) and not the breaker.

You need to look from different directions to try to find what wiring, hot or neutral, that is shared by both the washer and the lights.

.
Neutral is shared, if anything back at the feeder or service level of distribution. Same possible with hot - if single phase source and on opposite lines still a common coil to some extent at the source and some other loads can have impact on both hots.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
210228-1508 EST

kwired:

True that you can find common neutral all the way back to the pole transformer. However, its impedance is much lower than probably most anything else in the home. I have provided data on this in the past. But here are measurements from today:

1. Variable load --- 1500 W 120 V space heater 12.5 A.
2. Change in voltage at main panel at input lugs for 12.5 A load change on phase 1:
Phase 1 decrease of 0.8 V
Phase 2 increase of 0.3 V.

Multiply by 4 for possible washer starting current and these voltage changes would be about 3.2 V and 1.2 V.
Now assume a 50 ft branch circuit path of #14 copper, and the added drop is about 0.25 * 50 = 12.5 V. That is 0.25 ohms, and 50 A. So a total drop of about 16 V at the washing machine.

A 3.2 V change you start to see as a single flash, somewhat lower for a short time repetitive flashing. You will be very aware of a 16 V flash, even a single flash, and even not expecting or looking in the direction of the flash.

If two phases of a three phase source are the inputs there is virtually no interaction at the transformer. If two phases of a single phase center tapped secondary, then there is a little interaction from primary impedance, and this slightly lowers the voltage of the unloaded secondary half, but in my case the neutral drop more than compensates for this drop, and net result is an increase at the main panel of the unloaded secondary.

.
 

Hv&Lv

Senior Member
Location
-
Occupation
Engineer/Technician
But why have a variable control if you must use it at maximum? You didn't fix any problem, you masked it with certain conditions.
Variable control ( dimmer)is what the customer owns. We don’t supply it. The variable control cuts off part of the wave and causes the pulses to be below a noticeable pulse rate. If I’m not mistaken that is below 100 pulses per second.
We just told them how to get the pulses to stop for this one instance.
The TWACS pulsing is still a mystery though...
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
210301-1208 EST

cfg:

I suggest that you obtain a CREE bulb like I described above from Home Depot. Put this in place of one of your flickering bulbs. Does it flicker?

Ran the following experiment this morning:
Said CREE bulb directly across input to a 1/3 HP 120 V cheap induction motor with a switch between the bulb and the motor so that bulb is on continuously. From this switch point to my power source is an extension cord plus some branch wiring to add series resistance.

The voltage drop at the motor during starting is about 10 V RMS with a starting current of 30 A RMS. About 6 cycles of 60 Hz starting time. No flicker of CREE LED bulb. Measurements made on an oscilloscope. Peak values were measured and converted to RMS via 0.707 .

I might point out that I have never seen inrush current on an induction motor. I do see inrush current on a transformer. I do see starting current on a motor, and that is what you people erroneously call inrush instead of starting.

Inrush is mostly a 1/2 cycle thing, while starting current is many cycles.

.
 

Fred B

Senior Member
Location
Upstate, NY
Occupation
Electrician
Is light pulsing in time to the agitation pulse of the washer? I've seen that, and simply swapping the line at breaker panel by moving lighting breaker one space as suggested below had helped. LEDs seem to be more susceptible to "interference" from motor loads. That didn't get to the bottom of why it did it but it stopped the flicker.
Try moving the lighting circuit or washer circuit in the panel from L1 to L2 or L2 to L1.
 

Flicker Index

Senior Member
Location
Pac NW
Occupation
Lights
The flicker would be just the simple response to VD in the case of it cycling with the washer. No doubt they would respond equally with other similar loads.
maybe yes, maybe no. Even when the voltage is ramped slowly, the effect on output and variation in input voltage is entirely dependent on the design of the luminaire or lamp's LED ballast.

For notches, spikes and swells, the effect is sometimes due to glitch out in the LED ballast controller. Early CREE LEDs have had issues with the LED ballast controller rebooting when a ventilation fan was switched off. The inductive kickback that is known to sometimes cause GFCI to trip or cause a popping noise in radio was causing the lamp to shut off for about 1/3 of a second.
 

hanklazard

Member
Location
Alabama
Occupation
Electrician
Shared neutral with washing machine? I have many cheap led's flicker and continue to flicker after thunderstorms, the only solution was to swap the led bulbs out. After the next thunderstorm the new led's started flickering as well. It appears that many led's have really sensitive drivers. Dimmable Leds seem to have less problems.
 

kwired

Electron manager
Location
NE Nebraska
Shared neutral with washing machine? I have many cheap led's flicker and continue to flicker after thunderstorms, the only solution was to swap the led bulbs out. After the next thunderstorm the new led's started flickering as well. It appears that many led's have really sensitive drivers. Dimmable Leds seem to have less problems.
Whole house surge protection could possibly help with what you described.
 

hanklazard

Member
Location
Alabama
Occupation
Electrician
Whole house surge protection could possibly help with what you described.
Thanks, I've been meaning to try the surge arresters out. One square d surge arrester connects to neutral and the other square d model connects to ground and neutral? Is there one you prefer.
 

kwired

Electron manager
Location
NE Nebraska
Thanks, I've been meaning to try the surge arresters out. One square d surge arrester connects to neutral and the other square d model connects to ground and neutral? Is there one you prefer.
Neutral only won't do anything, you must be misunderstanding something.

Typical whole house protector will connect to both ungrounded conductors as well as to the grounded conductor. That way it can absorb any transient voltages between A-B, A-N, or B-N.
 

Another C10

Electrical Contractor 1987 - present
Location
Southern Cal
Occupation
Electrician NEC 2020
a lot of great thought out answers ....

I've resolved the issue at times with simply replacing the LED with a different or known better quality Luminous, and made sure they were dimmable, the non dimming seem more sensitive to surge.
 

tortuga

Code Historian
Location
Oregon
Occupation
Electrical Design
What about adding a capacitor to the LED lighting circuit (perhaps at the lights?) to fill in the sags in voltage?
 
Status
Not open for further replies.
Top