Any practical limit to multiplexing LED minimum duty cycle?

That is good information, thanks :-)

Cheers

Klaus

Huh, I did some testing on TH led's with lots of light output. (well f'ing lumens on the DK site.) ~5-10 years ago. (I was looking for photon sources) All at DC, the two or three I looked at carefully all had a threshold. Below which light (current from a photodiode) was less than linear with led current, (a 2/3rd's power law) and above which was ~ linear.. until heating or resistance things took over at high current. (I_threshold ~1mA, or so.) If you care about such things it's not a difficult measurement. I've used the uA input on a DMM with a PD as a meter. (fluke DMM) I'm not sure how deep a DMM will go into the 2/3 region...

George H.

I also measured the photometric output vs. forward current for various LEDs.

It's been too long for me to remember which LEDs had thresholds--I think they were low-efficiency greens (GaP?) and or yellows. But the InGaN blue, green, and white LEDs' outputs were impressively linear over the three decade range I measured them carefully. And, separately, failure to light at 1uA generally meant an ESD-damaged unit.

Lowering the forward current on the GaN LEDs /improved/ efficiency substantially, down to fairly low currents, due to i^2r losses in the LEDs' high ESRs. (I'd have to look up the data to give exact figures.)

For similar measurements I used a small solar cell as a PD, a small test jig with an LT1014 TIA, into an old IBM analog interface card, and logged by a BASIC program.

Since the pulse measurements were to determine perceived brightness, for those I flipped the A/B switch and adjusted the currents or pulse width for "no noticeable change" when flipping the switch.

Cheers, James Arthur

Right! Green and yellow.

Thanks, I'll have to look at GaN someday. In my case I wanted something Orange or yellow..(green) a little above the band gap of the red LED that I was using as a detector.

It'd be fun to look at all the different LED's.. it'd be an OK high school science fair project.

George H.

Nitride LEDs tune a lot with bias current, unlike phosphide ones. A bit of a trap for young instrument-builders, that.

Cheers

Phil Hobbs

I've wondered why the auto makers use this type of circuit and I have to assume they are trying to save those pennies from an ever so slightly more complex current regulation circuit. Then I wonder why they don't just cut off half the LEDs and I expect it is because they have designed the lenses to work optimally with the minimum number of LEDs, so again, counting the pennies. Damn those pennies!!!

The circuit is used since it takes less connections (read smaller microcontrollers), less parts (shared resistors). So yes, pennies saved

Cheers

Klaus

No savings on MCU pins, just the $0.50 inductor required for the switched current supply. I guess there is a diode as well.

Osram makes some really beautiful true-orange LEDs.

John Larkin wrote on 8/30/2017 10:49 PM:

I recall my customer's equipment having a bi-color LED that was orange and purple, very pretty.

Searches Digikey for orange osrams.. 15 found.

Most photons from this one (@50 mA!)

(same device shorter url)

George H.