TEC for red diodes ?

I'm planning to build a dual red module and anticipate future more powerful red diodes to come along and take the place of the diodes I will use now. That way the module does not need to be rebuilt and only diodes are replaced for a power upgrade.

First of all, from what I read in other threads, cooling does not seem to be a big problem for diodes so does it make sense to use a TEC? Maybe for higher power future diodes?
The Rohm diodes operate at a shorter wavelength at lower temperatures but that behavior does not seem to be common for newer diodes. Meaning that a drop of a few nm can be expected but not as much as for the Rohm diodes which supposedly go to 640nm @ 0 deg C.

If a TEC is used I figured that the needed cooling capacity isn't that much for cooling a diode. Assuming a red diode is driven with 300mA at 3V producing 200mW of light will produce 700mW of heat. So a minimum TEC capacity of 0.7W is needed to keep the diode at ambient temperature. A little extra is needed to be able to cool and keep T at a lower than ambient temp if needed. With higher power diodes in mind even a 2.0W TEC should be able to do the trick. Right?

For small cooling capacity a high end TEC (such as found here: http://www.tetech.com/modules/micro.shtml) could be used operating on 1A at around 4V so the power supply doesn't need to be a beast (e.g., 4A at 3V) and overall power dissipation can be limited.

Anyway I'm trying to figure out whether I should design the dual setup with TECS, any ideas?

Matthijs

Hi Matthijs,

You are right - cooling is not a big problem on the 658nm diodes.
As long as the mount has a reasonably low thermal resistance to chassis (and is electrically isolated).

The diodes are usually rated for 70degrees C operation. You can tell they must be designed to run hot by the small size of the original heatsink in DVD writers.

From the couple of data sheets I have seen which include wavelength vs temperature for 658nm diodes the tempco was 0.15 to 0.17 nm/DegC. (808s are usually circa. 0.27nm/DegC.) So cooling a 658 by 20C only gains you a 3nm shift.

Anyone out there measured shift on the Mitsubishis?

Yes - A 2W TEC would let you shift one Watt over a 30C gradient (or 800mW over a 40C gradient) The 2W rating of a 2W TEC is pumping power over a zero gradient.

You do not really need a TEC for a dual diode set up but for more diodes you do get the secondary benefit of thermal stability of your mechanics if you TEC everything.

Good luck with the dual.

Best regards,
Robin

Hi Robin,

thanks for that info. I'll go on without tecing :) Good point on the diodes being designed for high temp - then I shouldn't worry about cooling requirements of future diodes.
Next week I'm in Hong Kong so I hope to find some good deals on DVD writers. I think I'll bring my screwdriver set and some ESD bags to limit luggage volume on the way back ;)
Thanks to Kaz's list (and contributors) I know which drives I'll be looking for. :)

cheers,
Matthijs