# Max pulse curent for led's?

Discussion in 'Electronic Design' started by jtaylor, Sep 15, 2005.

1. ### jtaylorGuest

Just was thinking...

Is there a limit, other than heat generation, to how much current you could
shove through an LED if you could cut the pulse width (single pulse) enough
to stop it from heating up?

Could you get enough through in a thousandth of a second to get a decent
camera flash out of it?

2. ### Guest

There is a limit. You will find it in the data sheet of any LED, as the
absolute maximum current rating.

And you can't get anything like a camera flash out of a regular LED.

IIRR someone published a scheme - in Review Scientific Instruments -
for getting near-UV out of a blue LED by significantly exceeding the
absolute maximum current rating for about a microsecond, but the LED
died after 10^7 flashes, which is a total on time of ten seconds.

3. ### Don KlipsteinGuest

Get LEDs made for camera flashes. Lumileds, Nichia and Cree come to my
mind.

In my experience, 5 mm LEDs can withstand about an amp for a
microsecond. I think they should take 200-300 mA for a millisecond - NO
GUARANTEE. Do consider that LEDs are normally not linear devices and
usually have efficiency decreasing as current increases past about 80%
of the maximum they are rated to handle continuously. Some, especially
most white ones, have efficiency peaking at current around or less than
half their maximum continuous current.

- Don Klipstein ()

4. ### Don KlipsteinGuest

I have heard of this scheme, and I got the LED to last much longer - by
using current around 3/4 of an amp. I have yet to blow an LED after
running them like that for minutes. I suspect half an amp for a
microsecond with duty cycle of .5% will be fairly safe.

Also, this trick for getting UV out of a blue LED by pulsing with high
current only works well with GaN ones with peak wavelength (at 20 mA) 450
nm and spectral halfwidth that is wide (60-70 nm or whatever). These
included Nichia's NLPB series and similar Panasonic ones that may have had
the older Nichia dice. I got better results with the Panasonics -
possibly the epoxy was more UV-transparent. All of these LEDs were
obsoleted by 1997-1998 since InGaN ones were established.

I describe this and a circuit in:

http://www.misty.com/~don/ledbl.html, for which I have not made time for
updating in the past few years.

- Don Klipstein ()

5. ### Robert BaerGuest

Pulse 5,000 amps, and i will guarantee *brilliant* illumination from
almost any diode...

6. ### Rich GriseGuest

Yabbut, won't the color balance be all off? ;-)

Thanks,
Rich

7. ### Robert BaerGuest

Who cares if the color temperature is higher than that of the sun??
Once is enough...