Residential Pulsating LED Lighting - Neutral to Ground Voltage Ringing Transient

[bbwilson81]

I am hoping that someone here can help point me in the right direction. I collected power quality data at a newly constructed residence that is experiencing continuous pulsating LED lighting. There are many different brands and types of LED lighting throughout the home, some of which are unaffected. Looking at the standard line-to-neutral RMS voltage and current and voltage and current THD data points, the results appear to indicate that the incoming power is normal and within standards. Upon further investigation, the neutral-to-ground voltage and current (Ch3) was measured to reveal a 2V ringing transient that fires off on the voltage zero crossings. The monitor was set at the utility transformer, and the condition is present with and without the service connected. The utility changed out the transformer with to no avail and tested their driven ground at 7 ohms. Any gurus out there have any thoughts on the matter? The service is in a very low-density rural setting. The only unique condition is 2-2 MW wind turbines that are interconnected to the utility's distribution system a couple miles away although the homeowner advised that the condition is present even when the turbines are offline. Thanks! (ref. attached waveform)

 

[Speedskater]

The lower case word "ground" has many meanings, is it?
a] EGC (Equipment Grounding Conductor) Safety Ground circuit
b] GEC (Grounding Electrode Conductor) circuit
c] Planet Earth

 

[bbwilson81]

The lower case word "ground" has many meanings, is it?
a] EGC (Equipment Grounding Conductor) Safety Ground circuit
b] GEC (Grounding Electrode Conductor) circuit
c] Planet Earth

(b) GEC (Grounding Electrode Conductor) circuit

Attached is a picture identifying the location of the power quality monitor connections. It was connected in the meter pedestal rather than "at the utility transformer".

 

[ELA]

I am hoping that someone here can help point me in the right direction. I collected power quality data at a newly constructed residence that is experiencing continuous pulsating LED lighting. There are many different brands and types of LED lighting throughout the home, some of which are unaffected. Looking at the standard line-to-neutral RMS voltage and current and voltage and current THD data points, the results appear to indicate that the incoming power is normal and within standards.

The sample rate of your monitor may be too slow to capture inrush current to LED drivers when turning on at 62.5V.
It does capture the resultant (longer duration) voltage ringing caused by the sharp rise time of the inrush current.

 

[bbwilson81]

This particular waveform was captured with the main disconnect in the open position. The oscillatory transient is present with or without any load on the service. A LED from the house connected on the source side of the main disconnect switch with the switch open has the same pulsating effect as it does in the house. But yet not all brands of LED’s in the house pulsate.

Monitor Sample Rate: 1 MHz Voltage (16666 samples/cycle), 250 kHz current (4166 samples/cycle)

 

[winnie]

Your diagram isn't a picture of the actual installation, but rather a schematic drawn on a generic image of the meter pedestal, correct?

Are you measuring the 2V spikes between the wire circled in green (the GEC) and the wire circled in white (the neutral to the residence?)

The power log data is from when the residence main is off; what does it look like when the residence main is on?

2V, even as short transient spikes, across a short length of bus bar doesn't really make sense to me, so I am wondering if I understand the meter connections correctly, or if there is a metering error. Was the GEC disconnected to make the measurement? Is there a separate GEC at the residence such that the 2V spikes were measured between two different grounding electrodes?

Keep in mind that LED lamps all depend upon their driver circuit to set their output. Dimmable LED lamps are _not_ (in general) responding to the supply RMS voltage, but rather are trying to detect the dimming waveform and adjust accordingly. Some dimming circuits do a great job of detecting electrical noise and responding with flicking LED output.

I suspect that the 2V spikes you are measuring are _not_ the direct cause of the problem, but may themselves be a symptom of the electrical noise causing the LED flicker, or may simply be a red herring totally unrelated to the problem.

 

[bbwilson81]

Your diagram isn't a picture of the actual installation, but rather a schematic drawn on a generic image of the meter pedestal, correct? You are correct.

Are you measuring the 2V spikes between the wire circled in green (the GEC) and the wire circled in white (the neutral to the residence?) Yes, and it does not make sense to me either. I typically do not monitor CH3 voltage or current for residential analysis, so I am lacking any frame of reference. Is it possible that the transient could be sourced by a UPS driven load in the home? I agree that this could be a red herring, but I am grasping at straws with all other results being "normal/typical".

The power log data is from when the residence main is off; what does it look like when the residence main is on? Refer to the attachment.

2V, even as short transient spikes, across a short length of bus bar doesn't really make sense to me, so I am wondering if I understand the meter connections correctly, or if there is a metering error. Was the GEC disconnected to make the measurement? No. Is there a separate GEC at the residence such that the 2V spikes were measured between two different grounding electrodes? The house does have a separate GEC at their main panel 80-100' away from the utility-owned meter pedestal/disconnect. The URD is 3-wire between the meter pedestal/utility disconnect and house panel. Both locations have a driven ground bonded to neutral. It all remained intact while the data was collected.

 

[winnie]

IMHO there are several plausible ways that you are getting that Ch3 voltage measurement.

1) The connections to the ground/neutral bus bar are not solid, so you are actually measuring voltage drop caused by current flow. If you have current flowing from (say) the transformer neutral to the transformer local EGC, and there is a bit of oxide coating on the GEC contact, then you might see the voltage spikes without current flow from the house neutral. If this is the case, then changing probe locations on the neutral/ground bar will change the measured voltage.

2) The connection to the ground/neutral bus bar is solid, and there is no actual voltage difference between the probe contact points. But the probe wires form a loop, and a local magnetic field (say leakage flux from the transformer) is inducing voltage in your probe loop. If this is the case then changing the way the probe wires are arranged will change the measured voltage spikes. Twisting Ch3 and common together should eliminate the measured voltage.

3) There is something else going on creating line synchronized electrical noise which the metering system is picking up. Any electric fence chargers around?

Sadly, other than guesses, I've got nothing.

-Jonathan

 

[bbwilson81]

I appreciate your review of my study. In terms of a customer centric solution to the pulsating LED lighting; would you have any ideas beyond having them try different LED bulbs? Based on your comments it seems that any dimmable LED bulb in service that is not behind a dimmer switch could be replaced by a less sensitive, non-dimmable bulb. I have no experience with LED lighting to push them in the right direction. Thanks again!

 

[ELA]

I appreciate your review of my study. In terms of a customer centric solution to the pulsating LED lighting; would you have any ideas beyond having them try different LED bulbs? Based on your comments it seems that any dimmable LED bulb in service that is not behind a dimmer switch could be replaced by a less sensitive, non-dimmable bulb. I have no experience with LED lighting to push them in the right direction. Thanks again!

Here is a link to another discussion on LEDs blinking for an unknown reason. I tested and suggested a "Jasco" device in that thread that may help with LED blinking in some instances.

https://forums.mikeholt.com/threads...ss-distortion-led-bypass-experiments.2575014/

 

[Elect117]

Short of this just being a simple LED manufacturer issue where those are very sensitive to waveform distortions, I believe you have a nonlinear load creating a current harmonic.

On the second image, you have 20A on Ch1 and 50A on Ch2. The current waveform on Ch2 is distorted by a harmonic. I think it is a 3rd harmonic. A 5th harmonic would make it look more wavey at the peak. That or there is a switched power supply.

The first image is an almost perfect power supply.

Is there solar or some kind of switched supply in the house? Anything with a modulated supply? Or rectifiers / inverters?

The current distortion is impacting the LEDs. Are all of them on one leg? Can you split them between the two?