Help. DX laser losing power after few secs.

Hi.
I'm new here and I've searched for the answer here and haven't found it..

I bought a "true" 100mw pen from DX (sku01372) and after few days I noticed the laser power weaken drasticaly after few seconds of continuous emission.

I've carefully disassembled the laser (following a video tutorial) looking for a toasted component or anything like that. But it seems (phisically) all right.

Wich component may cause this problem? Laser diode? Or may it be a stressed SMD on the circuit board?
I'm curious because if I find out that the problem is related to any of the components on the circuit board, I can easily replace it and the laser power will remain stable for more time.

Thank you guys, and sorry if this subject has already been discussed.

RMA it.

I`ve done it. They sent 60% of my money back.
Now I have another lasers, but I want to fix that one.

I have exactely the same problem with my 50mw.

The first days i got this laser it was working great. But now if i turn it on it light up good and after a few seconds it dim.

By 20celcius and higher it working good but by lower temperaturs it has that problem. If you turn the laser on for about 1 minute it is "warmt up" and it working good.

ALL DPSS (green) lasers are very temperature sensitive. It's not defective, it's just a cheap DPSS.. Read here for more info..

Well, ok. I understand thet the overheating is the reason, but which internal component is overheating? this is the question. I guess i could replace it using similar one with more heat tolerance.
By the way, I understand that cheap lasers gradually lose power while in operation, but mine changes the power instantly... it remains 100% for few seconds (about 15s) and then instantly goes down to 20%, and since a physical objects can't gain so much temperature in a fraction of second, I think this is most likely a IC or capacitor malfunction than a gradual diode overheating.
Any clue?

As the diode heats the housing, it expands. As the housing expands the components move around. As they move around the alignment of the pump diode to the crystals changes. As the alignment of these components change the output is effected. If the alignment changes permanently, the output changes permanently. If the alignment returns when it cools, the output returns to its former strength.

The cheaper (and smaller) the components, the more vulnerable to changes in alignment they are.

Peace,
dave