Re: 12x Blu-ray SLEDS GB! 450mW!! Sanyo 12x!!
Ace82 said:
[quote author=brtaman link=1223438970/960#971 date=1227116789]Actually ace from what I gather its output (optical flux or something in that are) that kills diodes (COD) and not so much heat. I mean heat accelerates the process to some extent.
Today I was testing a new ddl driver I built myself by powering an 803t@125ma, long story short I forgot about it after I started working another project, after around 40 minutes i finally realized that the diode was still on (without any heatsinking except for the front part of the aixiz) touching the aixiz casing was painful it was blistering hot yet the diode is still putting out the same power as it was before the "freak" experiment.
Yeah, I know, but to some extent, I'm sure that the heat and COD are related. I mean, if your LD is cool, then there's probably not enough photons to cause COD. And if it's hot, there might be. :-?[/quote]
There is some relationship between heat and COD, but it's quite the opposite. Heat reduces the efficiency.
For example:
- At a constant current (in a laser pointer), heat reduces the efficiency, and the optical flux. In a laser with a heatsink, there is little heat. The power drops a few mW as the temperature climbs a few degrees. The tiny amount of heat that developes in a heatsinked laser can not harm the diode in any way. Theoretically it reduces the strain, due to reducing the optical flux at the die. But it's minimal so it doesn't really matter, if you use any kind of heatsink.
- At a constant power (in a disk drive), heat reduces the efficiency, causing the driver to drive the diode harder in order to keep it at the desired power. That's ok with the diode, as the optical flux won't increase, but will stay the same. But the increased current required to keep it at the constant power causes it to age faster. Heat itself also causes them to age faster.
The two conditions are completelly different from the diode's point of view. If you look at a diode datasheet, towards the end it will show an expected lifetime, in relation to temperature. But the condition is a certain power, not a certain current.
But heat doesn't kill diodes. At least not heat from the current flowing through it. It's only optical flux at the die (output power + reflections), that causes COD. It causes the weaker points on the end facets to burn up. Once they burn, they don't emit light, but convert it to heat instead. This heat is low in comparison to the heat created by the diode's operation, but it's only at a single point, causing increased stress on the area around.
The heat from normal operation is evenly spread out and doesn't really harm the diode directly. The point heat can, and the defects start spreading. Once such an imperfection developes, it's pretty much over for the diode. It can only go worse, and does so exponentially.
I used to think that heat is the problem too, and that if you cool a diode enough you can push it harder without damage. I thought that if you cool it activelly, you can safelly push it even further. Low temperatures do increase efficiency, causing more power to come out, but for the same reason the stress on the diode is actually increased.
In reality a diode can survive a higer current when HOT, while it can die at that same current when cold, simply because so much more light will come out.
Or if you shine a laser beam into a mirror so that it goes back in - it will kill a diode on the spot, even if it's cold, because it almost doubles the optical flux. (if your aim is good).
The mechanics of COD are pretty well understood in science and heat doesn't play such a big role in it. At least not the heat we are all so scared of.
Unless you meant the examples, where power seems to climb with heat. That's another story, however that is not normal for diodes... And theoretically, it could be wavelength drifting up, causing more light to pass the plastic lens, while the power behind the lens actually drops a little...
I'm still looking for another such diode to test what is actually going on.
:'(
It lasted about 10hrs I'd estimate, at 160mA, I never even made it to Dave's LPM but using an educated guess, was around 150-170mW. I almost thought this one would last :'( now, I need to get the next best diode and turn down the pot just a hair... The funny thing is that this one LEDified while it was lasing, I've never seen that before. And, it's not 100% LEDified, as it still projects about 2mW of violet light but it doesn't seem coherent, and it doesn't lase. It actually makes a big blob of light with two small dots side-by-side, and of coarse you can see that faint "green" color in there as well.
;D