Author Topic: Smoothing pwm flicker after the fact...  (Read 1975 times)

0 Members and 1 Guest are viewing this topic.

Offline PurduephotogTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Smoothing pwm flicker after the fact...
« on: May 16, 2018, 02:07:38 am »
I'm very sensitive to flicker. Extremely. Painfully, pukingly sensitive.

I've recently convinced my company to invest in a flicker meter that helps quantify just what I see and educate some of the folks I support. It's making progress at getting them to invest in better gear for their people.

I've been searching for ways to modify or bump up the cycle on electronics that have already been made. For instance I have some absolutely beautiful color correcting LED lights with a 300hz optical flicker. The led modules themselves are fed and could be separated... If there was a clever way of interesting an additional circuits that could take the chopped DC feeding them and convert it to current regulated... I'd be very happy. Searching for high current pwm smoothing didn't hit the right terms yet.

Would you have suggestions or already made products?

Thank you!
 

Offline JS

  • Frequent Contributor
  • **
  • Posts: 947
  • Country: ar
Re: Smoothing pwm flicker after the fact...
« Reply #1 on: May 16, 2018, 02:58:47 am »
  Adding a cap in parallel with the diodes will be a step in the right direction I guess.

  What you are thinking I guess is an inductor in series, much harder to add plus you might need to take care of the spikes, depending on the PWM output type (driving high and low or just high?) If driving high and low you don't need the diode to clamp the spikes but might want to add a series diode with your caps before they blow the output stage if only a cap is used and no inductor.

  I love halogen lights and the old fashioned incandecent, hot wire light is so much better! I think I'm more concerned about CRI than flicker, but both are a problem. CRi is a bitch and bites you quite often, dark/mate surfaces are the worse as you won't distinguish black from rust with a low CRI LED. Today I found CA rating in my LED bulbs, kind of a  :palm: when I found out what it was, just a way to put a higher index than the CRI  will be in bad lights. I think 300Hz fliker is fine for me, unless I'm using a machine running at 18000rpm in which case it could be bad for my hand toes.

JS
If I don't know how it works, I prefer not to turn it on.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22030
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Smoothing pwm flicker after the fact...
« Reply #2 on: May 16, 2018, 06:44:30 am »
Gluh, flashing tail lights, amirite?

I'm not sensitive to general flicker, but this example comes to mind as I notice it every time, and it's a potential safety hazard!

Very cheap LEDs do it too, because they're only lit half the time.  Less than half, actually.  Same as some failing fluorescents (when they're flashing, not just making that lumpy fluctuation).

For lights that plug into the wall and are lit less than half the time, in principle you could power them by DC and be done.  I don't know how safe that is.  You certainly want the output fused and limited.

For lights that are installed and driven by a controller (like tail lights), the lamp itself could smooth out slow pulsing, but this affects turn-on/off speed, which might not be a good idea with regards to, say, the safety of brake lights.  This would require an inline module to be designed, and probably an extra power or ground wire to be ran.  So, not too cheap/easy, unless done in quantity.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: tooki

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 6577
  • Country: fi
    • My home page and email address
Re: Smoothing pwm flicker after the fact...
« Reply #3 on: May 16, 2018, 09:31:07 am »
Gluh, flashing tail lights, amirite?
Rapidly (a few Hz) flashing blue-white bicycle front LEDs directed horizontally when it is dark, here. Gives me a headache every single time. They're illegal, but neither the cops nor the bicyclers don't care. My upstairs neighbour is an avid bicycler, and always uses that flickering seizure-inducing device. I've seriously wondered about building a 13 Hz flashlight, to retaliate with. (It wouldn't work, because they usually only look a couple of meters in front of them.)

I'm only sensitive to fluorescents when tired, so I'm not particularly sensitive. I do have a floor/reading lamp conversion project in mind, with LED COBs and a PWM-controllable (up to 1000Hz) CC drivers, but I'm still waiting for my rotary encoders from eBay. (Replacing all the electronics and the bulbs. I hope it won't flicker at low duty cycles...)
 

Offline PurduephotogTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Smoothing pwm flicker after the fact...
« Reply #4 on: July 08, 2018, 04:34:29 pm »
I realize this is really late in the game, but I just measured a buck-toot (7027 LED driver module). The sucker had no flicker at all. I tried it with a couple of different PSUs inline and it delivered brilliantly.

So if you really don't need dimming OR possibly need 2 levels, a pair of these (and they're VERY inexpensive) might solve your needs like mine.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Smoothing pwm flicker after the fact...
« Reply #5 on: July 08, 2018, 04:48:06 pm »
Gluh, flashing tail lights, amirite?
Rapidly (a few Hz) flashing blue-white bicycle front LEDs directed horizontally when it is dark, here. Gives me a headache every single time. They're illegal, but neither the cops nor the bicyclers don't care. My upstairs neighbour is an avid bicycler, and always uses that flickering seizure-inducing device. I've seriously wondered about building a 13 Hz flashlight, to retaliate with. (It wouldn't work, because they usually only look a couple of meters in front of them.)

I'm only sensitive to fluorescents when tired, so I'm not particularly sensitive. I do have a floor/reading lamp conversion project in mind, with LED COBs and a PWM-controllable (up to 1000Hz) CC drivers, but I'm still waiting for my rotary encoders from eBay. (Replacing all the electronics and the bulbs. I hope it won't flicker at low duty cycles...)


Those bicycle lights are annoying.

Funny you mention the tired part, I've noticed that when I'm tired my eyes are far more sensitive to flicker too. For example I have an older Tek TDS series scope with the NuColor display and normally it looks fantastic but if I'm sleepy I tend to see a pronounced flicker from the sequential color. I've seen some computer monitors with PWM dimming that cause a very pronounced eye strain, especially when I'm tired.

Now the really annoying part is I've encountered people who steadfastly insist that the human eye can't perceive flicker above x Hz and that I'm making it up. Frustrating because I can sit there and *see* the flicker, there would be nothing gained by faking it. I honestly wish I couldn't see it.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 6577
  • Country: fi
    • My home page and email address
Re: Smoothing pwm flicker after the fact...
« Reply #6 on: July 09, 2018, 04:31:15 am »
I realize this is really late in the game, but I just measured a buck-toot (7027 LED driver module). The sucker had no flicker at all. I tried it with a couple of different PSUs inline and it delivered brilliantly.
I'd like to check out analog dimming using AL8805, e.g. the FemtoBuck drivers. With the 33uH inductor, the switching frequency is above 300 kHz, so not only should there be no flicker, there should be no coil whine, either, I'm hoping.  It looks like connecting a 4.7uF capacitor in series after the on/off switch and voltage divider dimming pot (the analog dimming input range is 0.5V to 2.4V) should give a nice half-second soft-on, too.

Or perhaps use a Teensy LC to implement a capacitive touch pad, with a slider for dimming and on/off; AL8805 dimming input via filtered 3% to 75% PWM duty cycle (it's a 3.3V MCU). Dunno yet.
 
The following users thanked this post: Purduephotog

Offline PurduephotogTopic starter

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Smoothing pwm flicker after the fact...
« Reply #7 on: July 09, 2020, 11:25:29 pm »
I'm sorry I never got back to this... Yes, I'll take a look. Thank you for the suggestion... always wanting to find new drivers!
 

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2796
  • Country: us
Re: Smoothing pwm flicker after the fact...
« Reply #8 on: July 09, 2020, 11:49:21 pm »
Rapidly (a few Hz) flashing blue-white bicycle front LEDs directed horizontally when it is dark, here. Gives me a headache every single time. They're illegal, but neither the cops nor the bicyclers don't care.
They'll care when some guy driving a semi has an epiliptic seizure and runs them over!  But, then, it will be too late.

Jon
 

Offline User01

  • Newbie
  • Posts: 3
  • Country: us
Re: Smoothing pwm flicker after the fact...
« Reply #9 on: February 25, 2022, 06:32:38 pm »
The worst ones... the Cadillac Escalade tail lights. I don't know how or why they picked the perfect frequency to make a torture device!
« Last Edit: February 25, 2022, 06:39:48 pm by User01 »
 
The following users thanked this post: tooki


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf