Lowering temperature and increasing curret vs lifespan

[BartPL]

High,
I've searched the forum but I haven't found any info.

Did anyone try to use 500mW diode cooled to bellow -20*C, and double or triple its power?

I want to cut a 3mm aero-depron and my guess is that I need 1W+ power to cut it at reasonable speed. Since the cost of diode increase exponentially with power, I figured I could use cheap 500mW diored cooled to around -20 ... -40 deg C and increase the power (hopefully 3 times ).

I was thinking about a DVD diode with a 130W Peltier plate for 15$ that would pump-out around 15W at -30* or two such plates to get even lower temperature.

This guy increased power ten times (relative to max operating conditions) but fried it in couple minutes :/
https://www.youtube.com/watch?v=94w-_LCb6Ao

Regards
Bart

 

[Lotus_Darkrose]

From the video description:

"This diode definitely seemed to be outputting over 1W of power, but I cannot say for sure. (I need a laser power meter)"

*Is* and *seemed to be* are two very different things.

This test is in the video you linked to is completely irrelevant without a laser power meter.

If you're curious, please check the following thread, as there is a link to a video with a 638nm diode cooled with liquid nitrogen.

http://laserpointerforums.com/f42/possible-tec-orange-build-89704.html

 

[BartPL]

Lotus_Darkrose, thx for reply

Yes, I understand that this guy was very excited, but he didn't exactly know what is he excited about.

But as to facts, this diode of his was 250mW with efficiency of 0,97 W/A and he delivered 650mA. I'm not sure what is the linear range of this efficiency but lets assume that it is linear. So, he got around Po=630mW. Which is 2,5 times the power of the diode. (I don't know why I wrote 10 times :banghead: ).
I also assume that he didn't really cool it down with dry ice due to low contact area, co2 vapours between diode and dry ice, etc.

Now, I've seen The Cryogenic Laser video before and what's interesting for me, is that he got+50% efficiency at -20C. That would give 750mW from 500mW diode. Assuming that this effect apply to all diodes. I guess it does apply to at least all AlGaAs (that's what he used). However, he didn't increase the current.

Now to get 1W, I need another 30% and that doesn't seem like much.
I red about COD and my question is, how does it work in practice. Will the semiconductor material melt from increased power density alone or it has to reach some temperature, in which case lowering temperature should allow for increase in power density. I know that materials melt from temperature but in nano scale it might work quite differently.

In short, did anyone run a diode at 2x rated current at less than -10*C for a longer time?

What does CW stand for in datasheet? Continuous Work?

 

[Benm]

I never have, and doubt it would work very well.

The thing with cooling laser diodes is that it also does another thing: it decreases wavelength. If you start with a 660 nm laser at room temperature and cool it down to -20 celcius or so, the wavelength may move quite a bit - perhaps all the way to 640 or 630 nm. This would make the beam appear a LOT brighter, at equal current and even at equal effieciency (in terms of electrons flowing through it vs photons emitted).

But the key thing we do not really know: does it actually increase lifespan? Chances of COD would be reduced at lower temperature, but the question is by how much. If this is a function of absolute temperature (as expressed in kelvin) the difference may not be spectactular at all.

To us the difference between +40 and -20 celcius by feel like tropic vs arctic, but in absolute terms that equates to 313 and 263 kelvin, not even 20% in absolute temperature.

 

[Cyparagon]

The limitation with single mode diodes tends to be the power density on the die's out put facet. It doesn't matter how much you cool them, they'll still suffer catastrophic optical damage.

 

[BartPL]

Thx for reply and I have another question...
To get above 1W, I need 4x 250mW diodes, which I could cheaply extract from DVDs.

As far as I remember from primiary school, if I put parallel rays trough a magnifying glass, they are focused in a point, depending on the focal length of the glass.

So, assuming I could adjust LDs to shoot in parallel, and they would have more or less collimated beams, it should work just fine.

Did anyone tried it and how did that work out in real world?
Regards
Bart

 

[Cyparagon]

The hardware to properly align such a system would be far more expensive than just buying a 1-2W blue diode.

 

[Benm]

In all fairness, they would actually focus on a single point, which could be useful for burning/etching and such.

Problems arise when you want te beams to combine over any distance from the lasers, not just in one specific spot. It is virtually impossible to combine the beams of over 2 lasers (of indentical wavelength) if you want to do that. There are some techniques (like 'knife edging') that approach it, but the results are rarely perfect and aligning things is a highly complex job.

 

[upaa27]

I have done it before with a nice 445 9mm and after about 15 minutes (turned it on and off when power drop occurred to cool) it silently died

 

[RayJay]

Why not just buy a $45 M140 diode? You can push that to well over 2000mW easily without any active cooling? That would be much much easier and cheaper then what you want to do.