I have some light fixtures in my house that will periodically all start synchronously flickering at a slow rate (5-10 Hz) for a couple seconds and all stop at the same time. They're definitely in phase and they all use the same model of driver, which has an annoying shape that would be hard to place with swapping junction boxes and cutting bigger holes in my wall.
I tried bypassing the dimmer switch, but it has no effect. 10% brightness, 100% brightness, and no dimmer at all do it just the same, so not the dimmer switch.
Since they flicker in sync, there must be a common source of interference to synchronize them. Since they are dimmable, my theory is that some higher frequency signal on the power lines interferes with the phase cut detection circuit, causing the driver to lower the current briefly. I don't have probes handy that can check mains power, and I'm not too excited about blowing up my scope anyway, but interference on the line seems a lot more likely than RF interference.
Would something like this help?
https://www.digikey.com/en/products/detail/qualtek/851-05%2F006/739530And would it be up to code to put it in a junction box behind a light switch? Or am I thinking about this wrong?
Speaking from experience, I would suggest driving around your neighborhood, following the electric lines to the sub-station that serves your house. I wold not be surprised if you find some tree branches or vines that have grown into the higher Voltage lines. A tree branch swinging in the wind will have a natural frequency so your power can flicker at that rate.
Power companies just love to save money by cutting back on tree trimming.
Take some photos and send them to the power company. Threaten to sell them to a local TV station if they don't get on the ball.
Typical of mains browout (temp drop in V) due to large load eg industrial motor starting.
Get a mains V monitor ($15) see V during the event, use a UPS for affected lights.
Any industry eg auto body welders, food manufacturing, chemical plants nearby?
j
And would it be up to code to put it in a junction box behind a light switch? Or am I thinking about this wrong?
One of my daughters had the same problem. The problem was the dimmer switch. You seem to have determined that was not the problem. Are you absolutely sure? The lights were LED. I installed a 60W incandescent bulb and flickering stopped. That was a key clue.
As for whether modifying the box is code, that can be complicated. There are no limits on maximize size for the switch box so far as I know. There are limits on the number of wires in a box based on size, as I am sure you know. Can the current box be replaced with a larger/deeper box? There are face mounted boxes for repairs. When it comes to modifying a box, that can be tricky. Some inspectors will cite you even for drilling an extra hole. There are box extenders that are code.
Does the electricity company use ripple control for off-peak loads in your area? These are slow pulsed tones superimposed on the mains that can affect some devices. They would be present at roughly the same (multiple) times each day.
The effects are usually more pronounced with lights that are on dimmers or poor quality LED lights. You can sometimes also hear the tones in audio systems.
Definitely not the dimmer. I pulled it out of the circuit, put in a box and wired through the junction with a normal toggle switch and they do it exactly he same.
I don't think it's mains brown out, my HVAC system logs at ten minute intervals and the average voltage is stable. In addition, the driver is a 120->40V CC SMPS and not a linear driver like in cheap light bulbs, so slow changes in the RMS level should be compensated for by the current feedback.
I didn't hook up a photodiode, but by eye it looks like a few hertz periodic modulation of the drivers target current, like if someone will moving a dimmer switch from 80-100% moderately quickly. There's no shimmer or stroboscopic effect, just low frequency modulation.
Power company doing something after hours is plausible, as our trees. It only happens sometimes with no obvious patterns. I don't have a probe of the mains waveform, but we have long supply lines through a nearby forest. Short of buying mains probes, anything else I should look into?
Can you remove the PS and just install full-voltage LED's to see if they still flicker? In terms of probability, I would put the PS in the same logical bucket as the dimmer.