Re: [ATM] LED's and Diode Electronics

["Nils Olof Carlin" <nilsolof.carlin@telia.com>]

Richard,



>>>Last night's PPC/CHHU meeting was mostly about LED's.   Richard Nelson
brought some samples that he picked up at ALL Electronics, and we looked at
the I-V curve of one such device.    The equation for current through a
diode is I = Io*(exp(mE/nkt)-1), but I am not certain if that is the
appropriate equation for a LED.    LED's seen to operate at higher voltages
than ordinary forward-biased silicon junctions.   I guess the first thing to
do is curve fit the data we have...

Several years ago, I tested red LEDs at low currents (a few mA or less) and
found a pretty accurate exponential relationship, with (perhaps, but I am
not sure I recall correctly) 90 mV or so increase per factor 10 in current
and a voltage of 1.5V or slightly more at 1 mA. However, at currents close
to maximum ratings, this is not even closely valid - an internal resistance
may make a crude model, I don't know.

>>>Particularly, there was interest in bright LED flashlights, and the
question came up of how to supply power to a LED.   The usual method is
through a current limiting resistor, and that is wonderful for you if your
name is Eveready or Duracell.   But why not a capacitor?   The initial
current would be very high, but it would be for only an instant.   The idea
is to put all of the power into the LED and not put some of it into a
resistor.

But could you charge a capacitor without a resistor, or at least some
resistive element, somewhere in the way of the charging current? For that
matter, I bought some Chinese and Japanese (the latter may have been Chinese
too, but my Kanji isn't good enough to tell) NiMH cells- so far they work
well and may save money in the long run.

>>>Along with this question, there is another one:   How does the photon
output of the LED vary with current?

I compared the illuminations of two fgreen LEDs illuminating a paper target,
varying the distance of one, the current of the other, setting the
illuminations equal. Trusting the square law, I found to my surprise that
the light output (for some 2-3 mA and less) was proportional to the 3/2
power of the current! I'd love to have this verified (or falsified). This
relationship obviously must break down at some current, but it might explain
why it makes sense to pulse LEDs for low illuminations.

>>>All of this is highly dependent on temperature.

I'd love to know more about temperature dependence. Does the LED, current
fixed, give more or less light in the cold of an observing night than it
does when I study it indoors?

Nils Olof


_______________________________________________
ATM mailing list http://www.atmlist.net/