PWM is Pulse Width Modulation, the dimming method of the most LED light. If I stare in a 0.1% dimmed LED torch, it blinks hundred times per second with the maximum power into my eyes. I am wondering if it has serious health risks. Does anybody know about it?
Flickering light strains the eyes and the visual cortex.
There are technical limits with higher intensity flashes, that’s commonly agreed upon, because you can’t dim a high intensity LED to very low output with PWM without spacing puses too wide apart to remain indiscernible.
The other issue is intensity. A quantum within coherence length transports a defined energy portion. It’s only less number of quantums in a defined time, but still a single PWM pulse can produce an avalanche of quantums delivering together enough energy to harm the eye.
So yes, PWM is harmful to the eye, if the LED is strong enough.
Eye safety standards allow high energy, but very short laser pulses without protection. It depends on the wavelength and total energy.
The total energy is very low and shouldn’t cause problems. So while it won’t harm to look into a low intensity pulse width modulated light, it can turn on at full power any time, so it’s safer to not look directly into the light.
For some people very slow PWM can cause eyestrain, headache, dizziness or nausea.
@Sombrero, just curious why do you want to stare into a LED beam—is there some therapeutic benefit? How about if your eyelids are closed?
Only a few lights PWM at 100Hz. Which is really a terrible PWM Freq.
And at a low duty cycle can cause me instant headaches.
A 100Hz PWM with a 90% duty cycle is not bothersome.
Many start around 3K and shift to something higher as the brightness increases.
What actually seems to be bothersome is the time spent in the off part of the duty cycle.
There are several threads about PWM on the BLF.
I think the only health risk is eye strain or a splitting headache.
This is amplified if using the light to look at moving objects.
Heck, just moving my eye across a room lighted by a crappy PWM light really bugs me.
All the Best,
Jeff
Oh, and welcome to the BLF!
All the Best,
Jeff
That explains a lot, actually. All these years, I thought it was just K-mart in general having those effects on society.
I hope you enjoy your stay, Sombrero!
I guess that looking at the phone and monitor screens for several hours a day has a more negative impact on eyesight. This light also flickers very quickly.
Thanks for the kind greetings! 
I am new in this forum and the LED flashlight universe, I got my first serious led flashlight yesterday (Sofirn Q8 Pro), and I love it!
You all wrote interesting answers, but the conclusion is quite controversial for me. I think the main question is that how much the energy of a fast blinking light is “integrated” in the retina. A stupid example: If I push my finger to a 300 degree hot iron for 3 second, it will burn my skin. But if I touch it in very short impulses, It won’t hurt at all. But maybe the light exposure is not analog with my example.
kennybobby: No, I don’t (yet) use the light for therapy, but in the lowest brightness it has so faint light, that I can easily look directly into the LED chips. I tried this, after this I red a more technical review on net of my light, there turned out than the brightness level is implemented with PWM.
Sorry for my bad english…
Many PWM controlled lights have an additional channel running at a few hundred mA only. Quite safe if you don’t do silly experiments.
Kmart blue light specials - good one!!! 
no it does not
For those who are interested, I connected my oscilloscope to the Q8 Pro led cables, here is a GIF with some frames captured, from the minimum toward the maximum brightness:
what you want to know, would take an eye PhD to tell you
if they even know
wle
sorry but ALL ‘white’ LEDs use blue light to make the yellow that mixes to seem white
white isn;t white
that article barely says anything
wle
Here’s a video of the BLF LT1 lantern showing the various PWM and waveforms.
0.00: Middle Tint Setting, Ramping the Brightness. The PWM Duty Cycle reaches 54% Duty Cycle. On Low, the Frequency changes to 3.9kHz, with a very short Duty Cycle.
0.46: All cool LEDs only, Ramping the Brightness. Notice the light at full brightness has no PWM. It’s running a DC Offset. As the brightness decreases, the Duty Cycle lowers. The Frequency stays at 15.9kHz. At the lowest brightness, the PWM again drops to 3.9kHz.
1:40: A Middle Level Brightness setting, Ramping the Tint. Very unique PWM waveforms are created by the Duty Cycles of the 2 types of LEDs Interacting with each other.
2:15: Ramping both the Tint and Brightness.
I changed the time frame on the scope to zoom in on the waveforms. Starting at Warm, Low Brightness. Ramping to Full Brightness, Cool Tint.
Notice that at Full Brightness, the light is not running pure DC Offset.
There is still some Duty Cycle being seen from the LEDs.
I miss-spoke in the video. The full cool bright is a DC offset, but there is some artifact left from the way the Duty Cycles of the LEDs are interacting.
There are some really interesting waveforms being generated by the Duty Cycle of the warm and cool LEDs stacking on top of each other.
Taking a look at the ends of the tint ramp – The single square wave Duty Cycle, is (I assume) an indicator as to why some have reported that the light has the longest runtime when at those tint settings.
All the Best,
Jeff
That was really cool Jeff, thanks for sharing.
Where was the scope probe connected, or how was that signal generated/captured?
It’s probably about suppressing melatonin at night. Early studies involved nurses IIRC. There is something, but you cannot be sure it’s the light (alone).
Relevant information regarding flicker and its health effects can be found in the publication: IEEE 1789-2015.
The side effects, such as headaches, vary from person to person, but also vary depending on the brightness of the light, and the depth of the modulation. Dimmer light is less consequential.
There is a graph on page 29 which suggests that at 100% modulation, above 1150 Hz, the side effects go away.