Re: [ATM] LED's and Diode Electronics

[Mike Lockwood <melockwo@uiuc.edu>]

Hi,

Of late, LEDs have gotten ridiculously bright.  Many traffic signals 
now use red, yellow and green LEDs to conserve energy.  (I find them 
TOO bright, in general, and distracting to the driver.)  LED lighting 
for general household use looks like it's on the way eventually, given 
their efficiency, extremely long life, easy dimming control, and less 
waste when they are disposed of.  But I digress....

Here's some practical info for those who might want to try using an 
LED in their testers.  I've done the leg work for you, all you have to 
do is buy a few parts.

If you want to pick out a candidate LED for a tester, my first pick 
would be a bright cyan LED.  (I like testing in light blue light for 
some reason.  Your eyes may prefer green, yellow, etc.)

Here's a datasheet for some bright Agilent LEDs:
   http://cp.literature.agilent.com/litweb/pdf/5989-0169EN.pdf

I picked one out of the data sheet and found it on the Newark 
Electronics site.  This LED is generally around 12cd (12 candlepower) 
in brightness using 20ma of current.

Go to this web site:
   http://www.newark.com/

and in the search box, enter "96H6752", and you will get the price and 
info for this LED ($1.85 apiece, US).

This is a cyan LED with a viewing angle of 15 degrees, and a forward 
voltage of 3.3V.  That is, the voltage across the operating LED is 
3.3V.  This is the minimum voltage that you need to get light out if 
it.  In this case, two AA batteries won't cut it - so let's use my 
favorite - a 9V battery.

For a 9V battery, if we want 20ma of current from this LED, we can 
choose the limiting resistor to be:

   (9V-3.3V)/R = 0.020A,   R = 5.7V/0.020A,  R = 285 ohms

So, this LED, with a ~280 ohm resistor and a 1 Kohm potentiometer in 
series with it, would likely make a nice, idiot-proof, dimmable light 
source for a Foucault tester.  In this circuit, the LED disipates 66mW 
   of energy (3.3V*0.020A), and the resistor disipates 116mW 
(5.7V^2/280 ohms).  Make sure you use a 1/4-watt resistor.

Yes, using a 4.5V or 6V battery (lantern-style) would be more 
efficient, since the resistor would disipate less power.  9V batteries 
are so small and easy to get, though.

Currently I am using a much dimmer light blue LED in my slitless 
testers.  I think I got them at Radio Shack.  It's bright enough for 
an uncoated mirror.  We used it to test our 30" F/3.8 mirror.  I'm 
fairly sure that it's an order of magnitude dimmer than the LED that I 
gave info for above!

Some knife-edge photos and a picture of the tester (using the old LED) 
can be seen at

   http://www.mbm30.org/MBM8/Primary-Mirror-Figuring-T3.html

	Mike Lockwood

PS:  Davey, please stay on the list.


Dominic-Luc Webb wrote:
> I tried a common LED and it was too dim. For aging
> eyes, intensity is a problem. For later tests at polishing, some
> people are fussy about specific wavelengths in which case LEDs seem
> a pretty natural choice, provided one chooses a relatively monochromatic
> one (many are not).


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