William Chops Westfield wrote:
>>> Certainly sounds high for a T1 3/4 LED - if that is what you are
>>> using - at
>>> 100% duty cycle. I have always worked on 20mA at 100% duty cycle.
>
> Yeah, and some of those chinese knock-off LEDs have very "optimistic"
> ratings. There were some studies over in candlepower forums that
> showed some of the no-name white LEDs having drastically shortened
> lifetimes (brightness-wise) when overdriven (compared to, say, nichia
> LEDs with the same nominal ratings.)  I've had some LEDs fail in the
> way described here, especially UV and white.  The interesting thing
> to me is that the LED seems to continue to work as a diode with
> similar voltage drop; you just stop getting light out of it.  (eg one
> in series string of LEDs will go "dark" while the remaining LEDs stay
> lit at the same brightness as before the failure.)

I wonder if they're changing wavelength.  I had a bunch of old LEDs that 
I've had for a long time and experimented with some of them.  I exposed 
them to extreme amounts of current (like 5V with no resistor) to see 
what would happen.  Some blinked out instantly, while some would light 
up and stay lit for a good while (not really minutes, but surprisingly 
long).  I noticed that those would dim....some permanantly like the OP 
and you describe.  They still worked as diodes (I didn't check Vf), but 
they just didn't put out as much light.  Some were really dim.  One red 
LED changed color to an orangy-yellow color as it got really hot.  It 
seemed to retain some of this color shift permanently.  Maybe they start 
emitting IR instead? 

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist