PWM + PoV display, how to determine the sufficient pwm freq?

Go To Last Post
8 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hey everyone,

This may not be so electronics related as to physics related.

Assuming I have a PoV display, which displays 8*X pixels.
How can I determine the minimum pwm frequency for the leds not to flicker ?

I can calculate the amount of time it takes the display to move 1 pixel, does the frequency of the pwm has to be (lets say) 100 times that time ?
Would that make every pixel appear as if it's flickering at 100Hz, which shouldn't be noticeable ?
Is my logic correct ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The eye is good for 0.5% resolution of TV screen chrominance/luminance misalignment and it will be tricked by at least an 18Hz screen refresh into seeing a smooth image.

Use half the power line frequency ( either 25 or 30 Hz ) refresh rate and the eye will not complain) as a minimum.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you have a light source that's rapidly moving, it will appear as a bright streak in your view. If the light is flashing, that streak will be broken up into individual segments by the off time of the light. To eliminate that effect, the distance the light moves while off needs to be less than the diameter of the light source.
As a starting point, I'd suggest working out how fast the fastest moving LED in your PoV display is moving, and select a PWM period equal to the time it takes that LED to move half its diameter.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think leds need 70Hz to not flicker. So if you want 5 chars 6 dots wide that's 30 dots x 70Hz -> ~2000 dots/sec -> 500usec per dot.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chris-Mouse wrote:
As a starting point, I'd suggest working out how fast the fastest moving LED in your PoV display is moving, and select a PWM period equal to the time it takes that LED to move half its diameter.

This will effectively be like a 2Hz flicker at each pixel wouldn't it ?

I was thinking that I'll need a much higher pwm frequency so that each pixel period would have at least 50/70/100 pwm cycles.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Be careful using TV and cinema for flicker comparison. They both assume that their viewers are sat stationary in front of the screen and eye-movement is at a minimum.
(as an aside, 24fps film is usually projected with a 4-part shutter, giving 96 flickers per second)

A POV toy, on the other hand, requires movement.

I generally try and run my LEDs at around 250hz as a minimum to avoid annoying flicker as your gaze passes them.

I think that empirical testing is probably the best way of finding the slowest speed you can get away with!

Nigel Batten
www.batsocks.co.uk

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree condemned, it is an empirical thing.
I prefer a minimum 100 Hz PWM rate because I can see flicker below that rate if I view a PWM'ed display at a shallow angle (eg around 25 degrees from where my eyes are pointed).

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I discovered the flicker effects of movement by accident. I have a high powered LED headlamp that uses PWM for dimming. I use it for night time hiking, and everything seemed just fine until it started to snow one night. The falling snowflakes were bright streaks in my vision as long as the light was on full brightness. Dimming the light resulted in the snowflakes becoming dashed trails of light.