PWM - What is it, How does it work and how to detect it.

[jorn]

The loudest light i have owned is a lumapower lm21. The noice from the pwm filled the room... The funny thing is i could hear it used way different frequencys on med and lo mode. Lo mode was growling in a lo pitch noice (crappy low frequenzy), medium mode was screaming in a higher frequency. A simple shaketest comfirmed it visually.

 

[RI Chevy]

No. They can only hear it when it is close to their heads. But at a distance of about 2 to 3 feet.

I have no PMW on high, PMW on medium, and PMW on low. But the thing is it is louder on medium than on low, and their is a different, higher pitch tone on medium than on low. Low is quieter and has a lower pitch tone.

Even on high? If so, then I'd say there is a faulty inductor, or bad solder/trace inside...

No PMW on high. Only on Medium and Low.

 

[Cataract]

The loudest light i have owned is a lumapower lm21. The noice from the pwm filled the room... The funny thing is i could hear it used way different frequencys on med and lo mode. Lo mode was growling in a lo pitch noice (crappy low frequenzy), medium mode was screaming in a higher frequency. A simple shaketest comfirmed it visually.

Ouch! room-filling PWM noise is as bad and even worse to the ears as low frequency PWM to the eyes!
Because of the square wave's shape, it should sound and look like higher frequency -in most cases- when using a higher output, but the frequency is normally the same. This is due to the time off being shorter at higher output, making it seem like the flicker happens more often (thing of a faster strobe, but in this case, it is only our perception that is tricked into believing that). Some light could use a different frequency for each mode, but that's not a very economic way to go as it means having more complex or just more IC's (chips) on the board.

No PMW on high. Only on Medium and Low.

I know it might have seemed like I wasn't paying attention, but a high pitch whine on high means an inductor is whining, not having anything directly to do with PWM as high is most often direct-current or current driven (think of the straight line in my graphs.) It still could come from a bad solder of faulty inductor (small copper wire coil) in your light but, seeing that it might not be an isolated case, it couldt be a side-effect of low frequency PWM as well... I never heard that in person but I take both your words for it. I definitely don't like PWM in all its forms, period, but PWM noise must be a total wretch!

So, there ya go: Another bad side effect of PWM could include annoying noise (perfect or faulty board and/or components and/or soldering or whatnot)

 

[jorn]

Frequency is not the same on lo and med on the lm21. Lo mode uses a really slow and easy to see pwm. Med is faster and harder to detect.

 

[Cataract]

I'm trying not to get too technical to keep the thread as simple as possible in the spirit of the original post:

They probably use 2 different chips with fixed frequencies and pulse width.
Changing the frequency instead of pulse width is also a means to change the perceived intensity: less "beats per second" will be perceived as dimmer than a higher (faster) frequency, but that also means it is guaranteed you will perceive the strobing effect very easily on lower modes. That is not really PWM, though, but it is the same basic principle. It makes for a much simpler circuit and possibly cheaper, but not necessarily smaller.

 

[RI Chevy]

I know it might have seemed like I wasn't paying attention, but a high pitch whine on high means an inductor is whining, not having anything directly to do with PWM as high is most often direct-current or current driven (think of the straight line in my graphs.) It still could come from a bad solder of faulty inductor (small copper wire coil) in your light but, seeing that it might not be an isolated case, it couldt be a side-effect of low frequency PWM as well... I never heard that in person but I take both your words for it. I definitely don't like PWM in all its forms, period, but PWM noise must be a total wretch!

So, there ya go: Another bad side effect of PWM could include annoying noise (perfect or faulty board and/or components and/or soldering or whatnot)

I got you. I am not sure I could check the soldering on the M361N now though. Funny thing is I can't see the flickering, but I can hear the whine. I have 2 distinct whines on medium and low only. Just my luck.

 

[Cataract]

I got you. I am not sure I could check the soldering on the M361N now though. Funny thing is I can't see the flickering, but I can hear the whine. I have 2 distinct whines on medium and low only. Just my luck.

Ah, things are clearing up now...

Malkoff drop-ins are potted, so you can't open them up. If you can't see any flickering, even using my technique or any other, then you might just have a noisy inductor. Doesn't mean your module is not using PWM for med and low (I don't know Malkoff devices all that well); the sound frequency from an inductor will change with different pulse widths, with a higher pitch for a wider pulse - or higher output as in medium mode. I could bet this is it, as I do have a hard time believing Malkoff was cheap enough to use a slow frequency modulation. There is however another possibility: a fit that is not even in the host could also cause high pitch noise on certain modes. Basically, the "vibration" can be transmitted through smaller contact points. The physics of this can be complicated to explain, so I prefer to leave it out at this point, but it has happened to me once and was cleared up with a good maintenance and rotating the dropin a bit (sounds like shameless self-publicity and cross-ref between my own stickies) Try these steps to see if it changes anything: Basic Flashlight Troubleshooting Guide (I should have started with that before elaborating, but I'm leaving all the good info on the thread.)

If that doesn't change anything, it could be interesting to start a thread on your specific drop-in and ask if anyone else has noise issues. Something tells me your case is exceptional and you might be able to get a replacement - provided the problem has not been solved with maintenance by now.

 

[RI Chevy]

Thanks for the informative response. I have it in a Surefire bored 6P. The fit is flawless inside the host, but I'll try a few of the tips and see if I get a different outcome. I have many, many Malkoff modules, and I really do not want to make a big deal out of it, as I have to much respect for the Malkoff's. All of my others are absolutely perfect. Thanks again.

 

[Cataract]

My pleasure, I think people deserve to know.

I know about the Malkoff quality, so I find it a little odd. It is most likely a very exceptional case, but do let us know if maintenance makes any difference. If it does I would be grateful if you would mention it in the troubleshooting guide to share with others.

 

[RI Chevy]

I experimented with the M 361 N drop in and hosts that I have. Unfortunately I hear the PMW in any host I place it in while on low and medium power. So I think it is the drop in, not the host. The PMW is pretty consistent regardless of what host I place it in. I am running an 18650 battery.

 

[Cataract]

I'm pretty sure you have it cornered to the dropin. I'd take a chance and write Malkoff about this, they might just replace it.

 

[dml24]

Cataract, thank you for a simple to understand explanation. Excellent photos and graphics!

 

[Cataract]

Thanks, I did this all in the good spirit of CPF knowledge sharing :twothumbs:

 

[weaver]

First, thanks Cataract for a very informative description of PWM!

Or, using a photograph, you can calculate frequency by counting the number of pulses and dividing that number by the picture's exposure time in seconds. In the first image in the OP, there are 8 pulses shown in a time of 1/8s for ~64Hz.

Great tip, this may be the easiest method to measure PWM with fair accuracy with just a camera. With this, I was able to find the PWM of an Everest headlight to be about 110 Hz, which I find quite annoying in some situations, while the higher PWM of Lumapower LM31 on low mode is higher at about 675 Hz, which is not that noticeable in most situations.

The loudest light i have owned is a lumapower lm21. The noice from the pwm filled the room... The funny thing is i could hear it used way different frequencys on med and lo mode. Lo mode was growling in a lo pitch noice (crappy low frequenzy), medium mode was screaming in a higher frequency. A simple shaketest comfirmed it visually.

Interesting, I bought a Lumapower LM31 recently, and there I can only hear a faint whine if I hold the light right next to the ear.

 

[agnelucio]

Any frequency above 5kHz is pretty much undetectable, unless you're using a slow motion camera.

Although, if the frequency is 1kHz, and the duty cycle is 5%, then 1/1000=0.001, and 100/5=20, and 0.001*20=0.02, and 1/0.02=50.
So the light may appear to be flashing at 50Hz. Just thought I'd mention it.

Nice explanation by the way. :thumbsup:

 

[weaver]

Although, if the frequency is 1kHz, and the duty cycle is 5%, then 1/1000=0.001, and 100/5=20, and 0.001*20=0.02, and 1/0.02=50.
So the light may appear to be flashing at 50Hz.

If that was the case, then my LM31, with a PWM frequency of 675 Hz and a duty cycle of 1% on low, would appear to flash at about 7 Hz, which it doesn't. It still appears to flash at 675 Hz, while during every flash cycle the light is only on for 1% of that time.

 

[reppans]

I think this thread should have a disclaimer "if you want continue enjoying all your lights, please do not read." Before CPF, I never knew what PWM was, and I never saw it. Then I learned what it was, started looking for it, and even found very easy ways to detect high frequency PWM in every day scenarios and without equipment.

Damn it, now I can't stop seeing it and it has all but ruined my taste for any light with it. It doesn't make me nauseous or anything, but it constant comes into my view and my brain instantly notices something unnatural that makes me do a double take... "Oh yeah, PWM" and then everything is back to normal. It's just plain annoying now, and has made quite a few shelf-queens of some good lights.

Ignorance is bliss.

 

[Cataract]

Any frequency above 5kHz is pretty much undetectable, unless you're using a slow motion camera.

Although, if the frequency is 1kHz, and the duty cycle is 5%, then 1/1000=0.001, and 100/5=20, and 0.001*20=0.02, and 1/0.02=50.
So the light may appear to be flashing at 50Hz. Just thought I'd mention it.

Nice explanation by the way. :thumbsup:

Thanks!
I'm not 100% sure on your math, though (no time to pick up the calculator but I think you skipped a step), but yes, the shorter the duty time, the more it will appear to flash, mostly because it is easier to detect. I think the correct interpretation of your math is not really that it will look like it is flashing at 50Hz, but that it should be as easy to detect as if it was...

I think this thread should have a disclaimer "if you want continue enjoying all your lights, please do not read." [...]

LOL! I think you're right, though...

[...]Before CPF, I never knew what PWM was, and I never saw it. Then I learned what it was, started looking for it, and even found very easy ways to detect high frequency PWM in every day scenarios and without equipment.

Damn it, now I can't stop seeing it and it has all but ruined my taste for any light with it. It doesn't make me nauseous or anything, but it constant comes into my view and my brain instantly notices something unnatural that makes me do a double take... "Oh yeah, PWM" and then everything is back to normal. It's just plain annoying now, and has made quite a few shelf-queens of some good lights.

Ignorance is bliss.

I noticed that with medium speed PWM you can enjoy blissing ignorance for quite a while. With slow PWM lights, though, you notice it very easily and it becomes annoying over time when you get used to have lights with non-detectable or no PWM at all. I was lucky enough to notice it early, L0D being one of my first lights, and have paid attention from there on, even though it does not make me sick either, but I understand the pain or shelving a perfectly functioning light. I use my L0D as a keychain light now, and my older Zebras will become beaters when I upgrade them as they only annoy me near reflective surfaces.

 

[reppans]

I noticed that with medium speed PWM you can enjoy blissing ignorance for quite a while. With slow PWM lights, though, you notice it very easily and it becomes annoying over time when you get used to have lights with non-detectable or no PWM at all. I was lucky enough to notice it early, L0D being one of my first lights, and have paid attention from there on, even though it does not make me sick either, but I understand the pain or shelving a perfectly functioning light. I use my L0D as a keychain light now, and my older Zebras will become beaters when I upgrade them as they only annoy me near reflective surfaces.

Interesting... my old Zebra H51w uses a PWM that's about the slowest I've ever seen on any light - it was the first to become a shelf queen. The Foursevens Minis uses a high frequency, but just as you have mentioned here and elsewhere, the reflections off glass, faucets, etc, are what constantly catch my eye for that annoying double-take - I can't use them any more. It's really not an issue outdoors with few reflective surfaces, but indoors, there always seem to be something that reflects back.

At a quick glance, I also didn't see any mention here of the cellphone PWM test. Here's an example of a current regulated 47s Quark, high frequency 47 Mini, and and low frequency H51w all taken with an iPhone (they're all on their lowest modes):

 

[jorn]

I notice outddors because it rains alot, so any wet surface can be reflective..
Hate the pwm flicker on tail lights on the newer cars. Slow pwm led lights should have been banned on cars. The faster the car in front of you moves, the more visible the pwm gets. In every curve, or bump in the road, the flicker from the tail lights grabs my attention. The roads around here are made of bumps and curves only