Any practical limit to multiplexing LED minimum duty cycle?

Hi

I am working on a LED driver

Standard LED with average 2mA current. I am drinking it at 10mA, so duty cycle is 20%

Just wondering, is there any lower limit for the eye for minimum duty cycle?

Cheers

Klaus

Reply to
Klaus Kragelund
Loading thread data ...

That would probably be frequency dependent.

Reply to
DemonicTubes

Let's say we use a frequency so the eye cannot detect flicker, 50Hz

And then drive the LED within the specs at 20mA, 10% duty.

The figure of merit of best efficiency point may be non optimal, but I do not consider that right now

Reply to
Klaus Kragelund

Data sheet:

formatting link

Reply to
Klaus Kragelund

Klaus Kragelund wrote on 8/25/2017 11:30 AM:

If you drive with a fixed 10% duty cycle it will appear 10% as bright. What is your question?

I have used duty cycle to reduce intensity before, but the blink rate was

*much* higher. At 50 Hz you can see the blinking when you move your eyes. I find this effect very disturbing.
--
Rick C 

Viewed the eclipse at Wintercrest Farms, 
 Click to see the full signature
Reply to
rickman

LED driven with drop resistor with the following current and duty cycles:

40mA, 5% = 2mA avg 20mA, 10% = 2mA avg 10mA, 20% = 2mA avg 5mA, 40% = 2mA avg

Is that clear enough?

Cheers

Klaus

Reply to
klaus.kragelund

I worry that at 50Hz flicker would be detectable and annoying at duty cycles under 50%, especially since there is no phosphor to help out between pulses. It would be more noticeable in your peripheral vision and/or when moving your eyes.

Reply to
DemonicTubes

Yes, this is clear. What is your question? A lower limit to the duty cycle? I believe tests have shown there is no practical limit. I think you reach the max current through the LED first.

--
Rick C 

Viewed the eclipse at Wintercrest Farms, 
 Click to see the full signature
Reply to
rickman

IIRC the eye does get less sensitive for very short pulses, but I forget how short is short.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

FYI: please use a high frequency, even just for displays. The flicker of a low duty cycle at 50Hz is irritating to many people; at 100Hz, it may be passable, but is immediately obvious as soon as one moves their eyes.

It can be a safety issue. Every time I'm driving at night, I see cars with pulsed tail lights. Every time I scan my surroundings, bip-bip-bip-bip-, there's a line of dots across my vision and in that instant, I don't know where that car is.

My own experience has shown that over 1kHz is quite acceptable: it's hard to even see the pulsing, and when it can be seen, it's more like a dotted line than a series of overlapping frames.

I mean, it's not like it's hard to pulse LEDs at any speed. Unless they're being turned on and off from the CAN bus. *Shudder*. (You can even pulse them at >100kHz, add an inductor, and get higher LED efficiency! w00t!)

As for frequency, duty and current: AFAIK, at frequencies above the persistence of vision, the eye is a perfect integrator. At least up to pulse widths where transient vaporization occurs (TW, fs pulses?). :)

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

Like Phil H, I'm not sure about the relation between pulse duration, duty cycle and apparent brightness. One thing worth noting though is that there's an SOAR-type limit to how far you can reduce the duty cycle to get the same average current.

To take an extreme example for illustration, you cannot safely run a common 5mm LED (T-1 3/4) at 2A peak at 0.1% duty cycle to get 2mA average even if it has a continuous rating of 20-30mA. The maximum peak current is probably in the region of 100mA. It's best to check the manufacturer's datasheet if you're thinking along those lines.

Reply to
Pimpom

At least on the CREE web site, there was a white paper about running LEDs in pulsed mode.

Reply to
upsidedown

Tim Williams wrote on 8/25/2017 1:46 PM:

YES! The first time I saw this I was merging onto a road that can be busy and it appeared there was only a single car as I prepared to merge. On moving my eyes back to forward it suddenly looked like a dozen taillights were beside me and my instinct was to twitch the wheel away. Fortunately I didn't as there was no place to go. That was a Cadillac and since then I have seen this on many cars. I can't believe the auto makers haven't seen that this is a problem and upped the frequency.

--
Rick C 

Viewed the eclipse at Wintercrest Farms, 
 Click to see the full signature
Reply to
rickman

If you stay below safe pulse power levels for the light output I believe there is no limit in the percentages you are looking at. PWM-dimming (which is duty-cycle dimming) is done down to factors of 0.001 and less.

Assuming you want to keep a reasonably high brightness with fairly harsh multiplex duty cycles: For the LED itself you'll have to request the peak forward current rating from the manufacturer. It is often stated in the datasheet with infra-red LEDs but less so with LEDs for the visible range.

A typical rating would be ten times the continuous max and if you go above the rated pulse limit the lifetime begins to deteriorate. I believe the mechanism of wear is degradation via local electromigration. When you push the current to duty cycle ratio to extremes you could also smoke a bond wire.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

Yes, I wrote that in the original post. I wont drive them higher than the 25mA rated current, 60mA peak. Mind you, that is a thermal limit the first one, the second a pulse current limit

Cheers

Klaus

Reply to
Klaus Kragelund

Good info

I will contact the manufacturer for more information, just to be sure

Cheers

Klaus

Reply to
Klaus Kragelund

Given the responses that you are getting here, maybe you should try it.

--
John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Can be risky depending on what the product is. For example, it might all work fine on the bench and even for a few weeks. Then in the field LEDs start to dim or fail way too early. We've something similar that at a client where a contract design & assembly house had not protected an RF input well enough. They claimed it was fine and that everything was bench-tested. Problem was, after a number of thunderstorms rolled through more and more transceivers became desensitized to the point where end customers considered them deaf. Ranges went from hundreds of feet to tens of feet. The customers wanted new units on warranty, lots of them, and every one of them required a truck roll to install it.

I've got some more stories like that.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

The graph "Luminous intensity vs Forward Current" in the middle of page

5 on your datasheet goes to 60mA at 10% duty cycle which is encouraging. I guess the dashed part means "only pulsed" but Everlight should be able to put something in writing for you.

formatting link

Mind the temperature derating curve on page 5. For that to start dropping already at 25C is a bit disappointing.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

Klaus said he wouldn't exceed the LED max current spec, so it should be safe to play with duty cycles and see how it looks.

--
John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.