I have been interested in smart home systems, automation, and sensors for a long time. I like devices without reference to any manufacturer. So I often use the excellent Shelly Wi-Fi devices. These relays, controllers, and bulbs have an open embedded web server and are controlled via MQTT or web API.

I recently made new lighting in my home lab. I used a Shelly RGBW2 controller (24 volts) for the bright white LED strip (main light) and Shelly Duo RGBW GU10 bulbs (background light).

I immediately noticed discomfort in my eyes, and by evening my eyes were watery. I thought I had pinkeye, or the air in my room was dehydrated, or the street air was suddenly polluted. I agonized for two days.

On the third day, I experimented with LEDs and a spectrometer for one of my future posts. I was getting extraordinary results. During one of the experiments, I pointed the spectrometer at my illuminating LED strip in the lab for comparison with samples. Bingo! I saw a scary pattern of light ripples.

I also measured the light pulsations from the bulbs and saw a similar pattern.

The Shelly RGBW2 controller and Shelly Duo RGBW GU10 bulbs produce 600/1000 Hz pulsations (accordingly) and deep modulation (almost 100%).

But maybe it's normal, and my eyes are abnormally sensitive?

As an engineer, I looked into the current/existing regulations and found what I needed.
IEEE 1789-2015 «IEEE Recommended Practices for Modulating Current in High-Brightness LEDs for Mitigating Health Risks to Viewers»

It is easier for the average person to read a presentation from the U.S. Department of Energy, «FLICKER: Understanding the New IEEE Recommended Practice»

Briefly.
«Max % Flicker ≤ Flicker Frequency x 0.08»
For a frequency of 600 Hz - the modulation must not exceed 48%.
For a frequency of 1000 Hz – the modulation must not exceed 80%.
For a frequency of 1250 Hz - the modulation can be any.

My spectrometer detected an amplitude of 98-99%... And that's too bad.

How are the light modulation (amplitude) and the light pulsation frequency related?

A little check on the other side. I built a small circuit to test the voltage modulation from the controller (Shelly RGBW2 output) to the light source. I connected the device to a 24-volt power supply. I attached a load (a piece of LED strip) and an oscilloscope at the channel's output.

Additionally, I will check several device operating modes: 1%, 50%, and 97% of maximum brightness, and I want to confirm my guesses about the PWM modulator operating mode.

My results.
For 1% power:

For 50% power. The main light in my lab works in this mode.

For 97% power.

The minimum PWM regulator voltage is less than 10 volts in all cases.

This is not zero. Could the LED strip emit light at this voltage? If yes, then the light pulsation amplitude (the difference between the minimum and maximum level) may decrease noticeably. Is my spectrometer wrong, and the light emission amplitude does not start from zero?

I assembled another test bench to test the minimum supply voltage for emitting light from the used sample strip. Simple. I connected the LED strip to an adjustable power supply.

At 10 volts, the strip does not work - there is no light emission!

I gradually increased the output voltage and got the initial result (weak glow of the LED strip) at 15.4 volts.

This means that the LEDs do not emit light at the lower/minimum PWM regulator voltage of Shelly RGBW2 (see text above - no more than 10 volts). The light pulsation amplitude starts from zero.

More information on the Shelly Duo RGBW GU10 bulb light pulsation.

I checked the light modulation at different operating modes (at different light brightnesses).

At 100%:
The modulation is around 58% (it should not be more than 48%).

At 50%:
Almost 100% (it should not be more than 48%).

Results.


I don't know how a person, a buyer, can protect his health from dangerous devices without specialized knowledge and instrumentation.
My measurements showed that seemingly beautiful devices can be dangerous. My eyes did not fail me and immediately reacted to the defect and hazardous light property.

I have written a letter to the manufacturer describing the problem and await their response. Stay tuned for a follow-up.

Update 06/08/2023.

I received a reply from Shelly (see below in the text).

Unfortunately, I do not recommend using Shelly Duo RGBW GU10 bulbs and Shelly RGBW2 LED controllers.

The built-in PWM controllers in these devices:
- have unacceptable operating frequencies,
- do not meet IEEE 1789-2015 requirements and recommendations,
- and are hazardous to the human eye.