Interesting, Jay. That makes sense that there is something else going on to exacerbate the flicker with the “dirty” unit running PWM at 200Hz. Seems like flicker should not be perceptible to my eye at that freq. It was not quite seizure inducing but it was noticeable to me out of the corner of my eye, which is where the display sits all of the time, so yeah, not good.

I ran some quick tests on the IC-7300 with the square wave “clean” supply. I don’t see any interference with it on down at the AM MF band or elsewhere on HF. My kitchen LEDs are a different story though. Maybe I will attack them next!

LCM

Levi C. Maaia, K6LCM

www.levimaaia.com


On January 4, 2024 at 11:40:38, Jay Hennigan (jay@west.net) wrote:

On 1/3/24 09:50, Levi C. Maaia, K6LCM wrote:

> Anyway, I know that LED drivers can be a source of RF noise, especially
> cheap ones that use PWM to dim the light from LEDs. The one that came
> with this piece was visibly “dirty” to the point that I swear that I
> could see the flicker at certain brightness settings. I put my portable
> scope on the DC LED dimmer/driver output and it was very ugly.

They are pretty much all PWM. Resistive dimmers are horribly inefficient.

> I decided to do some research and see what I could learn about PWM and
> LEDs. Turns out that the frequency at which these drivers pulse the
> light varies from model to model with some pulsing the light on and off
> at as slow a frequency as ~200 Hz. Apparently that was enough for be to
> perceive flicker.  I found another unit that was more that three times
> faster (~600Hz) and swapped it in. The flicker went away and the picture
> on the scope looks a lot cleaner now.
> In the attached video I show the “clean" output, then the “dirty"
> output. I can’t figure out how the dirty output  manages to even dim the
> light. It barely budges off of 12 volts while the clean driver has nice,
> neat square pulses.

The "dirty" one is likely just filtered with a capacitor across the
output. LEDs are current driven. If the series LED string has a forward
voltage drop of just under 12 volts and there's a fat capacitor across
the output of the PWM dimmer it will look like 12 volts with a bit of
ripple at the switching frequency on a scope. Ripple will be more
evident at lower brightness levels. While you've labeled it "dirty" it's
likely going to be cleaner from an RF standpoint than driving the LEDs
with square waves over a length of wire that can and will act as an antenna.

I suspect that the flicker is an artifact of the dimmer control
circuitry missing pulses at certain brightness levels. Maybe poor
line-side filtering causing power line "beating" with the internal
oscillator at particular brightness levels.

200 hertz isn't going to produce visible flicker. Old-school TVs are 30
hertz with interlace that makes it effectively 60. 35-mm movies at
theaters are/were 24 hertz, usually with a shutter that mimics 48 hertz
to reduce flicker. Even many gaming monitors are 120 hertz or less. The
fact that it only flickers at certain brightness levels says that the
200 hertz refresh rate isn't the cause of the flicker.

> I didn’t notice any noise on the ham bands with either, which is good. I
> have other LED drivers around the house that are noisier on the RF side.
> Maybe I’ll try swapping those out.

I think you'll find that the squarewave version is noisier. Try the low
end of the AM broadcast band.

--
Jay Hennigan - jay@west.net
Network Engineering - CCIE #7880
503 897-8550 - WB6RDV

_______________________________________________
SBARC-list mailing list -- sbarc-list@lists.netlojix.com
To unsubscribe send an email to sbarc-list-leave@lists.netlojix.com