Electronics > Projects, Designs, and Technical Stuff
Lighting flicker meter
(1/4) > >>
IanMacdonald:
Recently suffered some rather bad headaches when working in a particular room here. Found out that the LEDs I'd recently fitted are flickering badly, and this seems to have been the cause.  >:D

Surprisingly, of four similar bulbs one was flickering strongly at 100Hz when checked with a photodiode and scope, the others not. 

This got me to to thinking that a lighting flicker meter might be a very marketable idea. It would a give the public a way to find out how much flicker is present with any given light source.

Since frequency is important in physiological effect as well as amplitude,  I guess it would need to be weighted in some way toward lower frequencies. Is there already any standard for measuring this, I wonder?
Zero999:
I don't know whether there's a standard.

A cheap and cheerful way to do this would be to add a photo diode to one of those inexpensive portable nano oscilloscopes.

I wonder if a phone camera would be capable of picking it up? If so it might be possible to write an app.
NiHaoMike:
Here's a gadget for listening to flickering light:
Siwastaja:

--- Quote from: Hero999 on November 23, 2018, 01:52:45 pm ---I wonder if a phone camera would be capable of picking it up? If so it might be possible to write an app.

--- End quote ---

Very well possible, assuming the camera API gives even the most basic controls.

Just set the exposure time to significantly shorter than 10ms (say, 2 ms or less - this isn't a problem because the amount of light input is huge, you may well go below 1ms even on autoexposure near the bulb), and take a statistically significant number (i.e., >1000) randomly timed shots. Plot the amplitude histogram.

If you want a single number, maybe
avg=0
foreach(linear_light_value)
   avg += linear_light_value;
avg /= count;

err=0
foreach(linear_light_value)
   err += (linear_ligh_value - avg)^2

rms_err = sqrt(err/count)

would do, as it would integrate the sum of the difference compared to steady light, but give more (quadratic) weigh to really low/high amplitudes.

This would be an elegant "mass market" solution since everybody (except me; my phone is 13 years old now) has a smartphone and can buy the app. And here we are giving this idea out for free :).
IanMacdonald:

--- Quote from: Hero999 on November 23, 2018, 01:52:45 pm ---I don't know whether there's a standard.

A cheap and cheerful way to do this would be to add a photo diode to one of those inexpensive portable nano oscilloscopes.

I wonder if a phone camera would be capable of picking it up? If so it might be possible to write an app.

--- End quote ---

There is actually a phone app though I'm not sure how reliable that would be. Was thinking along the lines of a cheap LCD scope myself.

Did a few tests with the bench scope on LED bulbs from around the place, and a drawer of CFLs which I'd replaced with LEDs. The results were quite worrying- The CFLs mostly did not flicker, the LEDs mostly did, having a significant 100Hz modulation on their light intensity.

This is a surprising finding, and it brings doubt upon the idea that LEDs are better for your eyes. One two types  of LED were totally free of flicker, and these weren't necessarily the most costly ones either. TCP brand were the best tested (AC component <2% of DC photocurrent) along with COBs fed from cheap Chinese inverters which were also flicker-free. Some of the store own-brands were dreadful.

All in all, a bit of an eye opener, and maybe lends weight to the argument that old fashioned bulbs (or halogen) are still best for your visual health. Or, if you want to use LEDs you need a flicker meter. :-DMM
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod