7
\$\begingroup\$

My company installs ZigBee-based phase cut dimmers for bulbs and LEDs (those which support phase cut dimming). Recently, we encountered a site where we faced a flicker issue. The light is slightly flickering. When the same set of LED driver + LED is being tested at our test lab, everything is working fine. This issue seems very site specific.

Site details - Some active construction is still happening at the site. Machinery like stone cutters, drill machines are being used on a continuous basis. It's also possible that the site is being powered by a generator set and not connected to grid at all (I am not sure of this part yet).

Here is the site waveform:

Site AC

I noticed a repetitive glitch in the AC waveform which could be a reason behind the flicker issue.

AC waveform as seen at our test lab:

Lab AC

This one looks different from the site waveform.

What is causing this distortion at the site? Is there a way I can correct it so that dimmers don't cause any flicker?

\$\endgroup\$
  • 2
    \$\begingroup\$ Can you put an LDR or photodiode on channel 2 to see when the light triggers? I think you can use an LED as a photodiode if you're stuck. Check the web. (It's hard to be original these days.) \$\endgroup\$ – Transistor May 27 at 22:08
  • \$\begingroup\$ I'll try it out. Alternatively I can hook up the second channel to LED driver output and monitor the waveforms. \$\endgroup\$ – Whiskeyjack May 28 at 3:13
  • 3
    \$\begingroup\$ If you think those ripples are bad, check this guy who noticed that his utility company injects 9% 1 kHz ripple on purpose. \$\endgroup\$ – pipe May 28 at 8:03
  • \$\begingroup\$ I am hoping that different ICs in led drivers will have different response to this glitch. I am planning to test a different matching LED driver. I will post an update once I find something. \$\endgroup\$ – Whiskeyjack May 28 at 10:07
12
\$\begingroup\$

The first waveform looks fine for mains, it can get much worse.
I suggest you use a rent a power corruptor, or use a lab with an operator, to see how vulnerable your dimmer is.

And work with those results to improve the design of the phase angle dimmer.

\$\endgroup\$
  • 1
    \$\begingroup\$ TIL about "power corruptors". It's like integration testin, but for hardware. Neat! \$\endgroup\$ – Alexander May 28 at 2:19
7
\$\begingroup\$

The distortion is due to diode rectifiers, phase angle rectifiers. The distortion is not such big that should cause any malfunction. Possibly, the LED power supply misinterprets these false glitches with expected phase angle voltage from the dimmer.

\$\endgroup\$
  • \$\begingroup\$ Will active PFC in the LED driver lead to the said misinterpretation? \$\endgroup\$ – Whiskeyjack May 27 at 9:51
2
\$\begingroup\$

Flicker: as dimmed, the LEDs have longer dark time than with full light. The light from the fluorescent material decays more. The human eye can well see the flicker, at least if the LED is a little aside from the center of the field of the vision. It has been the same with fluorescent lights and old CRT televisions.

Another thing: If you have another non-linear load which also distorts the voltage waveform and does it in somehow a complex and different way, the summed distortions can produce mixing results in nonlinear components. Some of them can be in the visible frequency range. Without proper spectral analysis this all is only speculation.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.