445 dimming issue

Hi all, i tried searching for a solution to my problem. It seems others have had similar issues but with different solutions. So heres my issue...
I built a driver using a lm338, and am using it to power a m140 diode, first time i tried this, it worked great for a little while but then, it stopped producing the same amount of brightness and power. I figured, dead diode, sucks to suck. So i bought another one. Made sure that it was well heatsunk, and that the driver was working well, i also replaced the lm338 with a new one since it was not heatsunk at all last time and i thought maybe that was part of the issue.
So i turn in on, it works great, i never have it on for more than about 25-30 seconds at a time, but after about 3-5 minutes of runtime, same issue. It will flash bright for a split second, and then dim to probably no more than about 100mw. If i keep trying, it stays bright for less and less time before dimming. If i leave it alone for a while, and try, it will stay bright for that initial time before dimming again.
So my question is this, is my diode dead again? Have i killed two of them? Or is it more likely the regulator thats dead? Or maybe something else i have overlooked. Thanks for reading and for any help you can give!

Not sure what the problem is but it's "heat sinked" not "heat sunk."

Sounds like a ship sunk because of heat.

Yeah I've had that happen a few times to me. You need to have a really good heat sink.

Don't run your laser for more than 20 seconds at a time.

Diode is damaged but might still be usable.

Have you check the battery voltage as it sounds like the battery is sagging.

BTW is it true cats taste like chicken :thinking:

Oh yea, thats something i didnt mention, im almost certain its not battery because when i have it hooked up to an adjustable power supply, the results are the same. And Im pretty dang sure that my heatsink for the diode is more than sufficient. Are you saying the chip needs a really good heatsink as well? Or just placing emphasis on that fact that the diodes do?

Just like chicken. A little stringier.

For long duty cycle yes your driver needs a heatsink

so we're thinking its the driver then. thats what i was hoping for. still not totally convinced thats the issue, but its the cheapest place to start. i will post a reply when i get the new chip in. any other suggestions would be greatly appreciated until then though!

What current is your driver putting out? and how did you test it before connecting it to the diode?

Alan

I had it hooked up to a test load before the first time to make sure it was working at all, and it was putting out far lower than the diode can take. I also put a fuse in the driver to make sure that it doesnt go above 2A just to make sure.

You could try hooking the driver up to the test load and run it at least as long as you ran the diode. Since the output didn't drop and stay low, as you said it would return to normal after it sat awile, it sounds like a power supply issue. Try hooking the driver back to the test load, with the diodes sinked good, and see if the drivers output changes.