[Cyborg] 200 laser diodes

Eric Boyd mrericboyd at yahoo.com
Sat Jun 5 17:12:48 UTC 2010

I'd love to help make that testing rig.

It'd be especially interesting if we could test the unit-to-unit 
variation in optical output.  We wouldn't need a rig capable of 
measuring absolute power, just one capable of monitoring relative power 
between exposures.  I'm thinking any old photo diode should be able to 
do that, yes?  Just hook up the laser & a photo diode to an arduino, do 
a brief pulse of light, record the reading, do the next one, etc... The 
hard part might be the laser-diode alignment, we probably need to build 
a mechanical rig of some sort.  We could also monitor the electrical 
power consumption, I wonder if these lasers vary much in efficiency.

I'd be pretty upset if there are more than a couple of duds in this lot; 
I think 5% would be outrageous.  At any rate, it's too late to tack on 
more now :-(.  What makes you think the failure rate will be so high?  I 
am wishing I'd ordered a couple extra dozen - I didn't mostly because I 
didn't want to get stuck with more than my 16.

I have an interesting safety question: do you think that PWM controlling 
the lasers would actually make them eye safe?  For instance, let's say 
we discover that the true optical power of the lasers is something like 
10mW.  Would it actually make them any eye safer to PWM control a laser 
to 50% duty cycle?  Theoretically that reduces the "power" to 5mW, but I 
expect that in practice it's the peak intensity on your fovea that 
causes the damage, and PWM will not reduce the peak intensity.  Thoughts?


On 6/4/10 1:48 PM, Andy Isaacson wrote:

> I wouldn't bet on the order getting here in a hurry, but that's OK with
> me. :)
> By my math it works out to $2.65 per diode, so I get 18 for a total of
> 47.60.  But we should probably assume that 5% of the order will be duds,
> so maybe we should tack on a bit to cover losses.  Thoughts?
> Maybe we should build a test rig to test out the shipment?  I'd be down
> for helping with build and testing if anyone else wants to participate.
> -andy

More information about the Cyborg mailing list