I recently found a source selling very cheap plastic 3 white LED
flashlights with great reflectors. I've dissected other lights for use
in other applications in the past with success. They usually consist of
3 AA type cells being replaced by a 7805 and a series diode. The lights
themselves usually have a controller built in, so in many cases they
probably handle voltage variations, as when the batteries die or are
fresh. The chip is usually under a blob of epoxy...

However, my latest find has it's battery holder for 3 cells wired
directly to the 3 LEDs (white) which are in parallel. Being so cheap,
I'm concerned that no engineering went into them, and that they are
being overdriven. At 4.0v (the best I had at the time), they drew 90 mA.
I haven't measured them with good cells at 4.5v yet. The light they put
out is very bright at 4.0v and quite impressive.

The question is, how could one characterize these LEDs to determine what
they should be driven with? Simply done though.... I would like long
life, since the housing that will be made around them will be difficult
to change in order to retrofit or replace the LEDs, and any replacement
will probably be from a totally different device. I also don't want to
arbitrarily dim them down either. Being wrong enough to reduce life from
100k hours to 90k hours is probably OK, but being so far off as to
reduce it to hundreds of hours won't do. For starters, does 30mA per LED
at 4.0v sound right for a normal (not jumbo of super bright) LED in a
Chinese product? ;)  For such a simple setup, would a constant voltage
source be OK (7805 and one or two series diodes) or is it worth the
extra to make a constant current source (3 terminal reg config'd as a
current source), given that they are in parallel? The ultimate power
source ahead of the reg is usually 12v-14v and many times an automotive
environment.

Thanks in advance,
Skip


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