LED1501 & LED1502

Werner Almesberger werner at openmoko.org
Thu Apr 17 09:44:10 CEST 2008


[ Copying the list again. ]

Allen Chang wrote:
> LED1501==>LED Ultra Bright Amber/Blue ??p=609/468nm SMT 3.2*1.0*1.5mm 
> 3PIN HT-210UD/NB HARVATEK
> LED1502 ==>LED Ultra Bright Red ?????=636nm SMT 3.2*1.5*1.0mm 2PIN 
> HT-110USD5 HARVATEK

Thanks a lot ! I put the data sheets into our internal document
repository.

I examined this because I found that the power button LEDs seemed a bit
dim. Here's what I found:

All three LEDs are being supplied through 220 Ohm from 3.3V. If we look
at the typical forward voltages, we get these currents:

red:	(3.3V-1.9V)/220R = 6.4mA
amber:	(3.3V-1.9V)/220R = 6.4mA
blue:	(3.3V-3.3V)/220R = 0mA

Okay, that's obviously not the right result for blue. We're too close
to the non-linear part of to curve there. Next try: if I draw a line
for (3.3V-x)/220R into the Vf vs. If diagrams, I get roughly the
following intersections:

amber (UD):	1.78V / 7mA
blue (NB):	2.9V / 2.5mA

Taking this to the intensity vs. current chart, and multiplying with
the minimum and typical luminous intensity, I get

amber:	7mA -> 25%	14-30mcd
blue:	2.5mA -> 15%	5.4-9mcd

Interestingly, on my GTA02v5, amber looks less bright than blue. Red,
which should be quite similar to amber, looks fine.

We could safely reduce R1561, R1563, and even R1562 to 100 Ohm, which
would make the LEDs roughly twice as bright (and double their current
consumption, which, in the case of GTA02v6, would be of course still
far below GTA02v5's).

Do we care to make such a change ?

- Werner




More information about the openmoko-kernel mailing list