Very annoying heat problem

[The Baron]

Hey, i have a jayrob maglite build that has been having some problems. It is an M-140 that I got from Rocket689 with a FlexmodP3 set at 1.75A. First, when I got the diode, it had a crack in the window, causing it to have 2 beams. it worked like a champ though and would stay bright for minutes on end and then I would turn it off after about three just to be safe. I sent the diode back to Rocket and he decanned it for me to take off the window that had a crack in it. after i have it put together now, it will peak at 800mW after 8-10 seconds and the focus ring will get VERY hot. The output will exponentially drop. It gets so hot in 15 seconds that the dot gets extremely faint. Can somebody help me find out the problem?

EDIT: BTW, those readings are with a jayrob G1 lens

 

[ApexProxy]

Maybe when the diode window was removed it might be possible that the diode itself became slightly unseated from the module. You may want to verify that the diode is still nicely recessed. You could always cut the wires to the driver and give the diode a press and then re solder the leads if it appears to have come out slightly.

My guess is when Rocket was decanning the diode the pressure on the diode pushed it very slightly out of the module. Or if he had to remove the diode, its possible he might not have pressed it back in fully? I can't really imagine anything else being the issue short of something between the module and the heatsink blocking heat transfer. I could be wrong though

EDIT: On second thought, what if you are getting overshot beam from the lack of can and its hitting the inside of the lens/module or if the diode is slightly crooked and its hitting the lens/module?

 

[The Baron]

I dont think its overshot but I need to look into the issue you talked about. Too bad I dont have a press

 

[The Baron]

I just checked, the diode is pushed past flush with the module

 

[AnthoT]

I just checked, the diode is pushed past flush with the module

If you don't have any press method you can put some arctic silver ( not the glue kind!!! ) to help with the heat dissipation... It's not as good as fully pressing it but it should move the heat better than just letting it sit on the diode

 

[The Baron]

I want to know why the diode is doing this though. Any ideas?
About the Arctic Silver, do i put it on the diode and around the module or just on the module?

 

[ApexProxy]

My best guess is the diode was slightly pushed out of the module when it was being decanned so now the heat isn't being transferred correctly. It might explain the super low duty cycle. Either that or your module isn't making good thermal contact with the heatsink. You might want to start by looking at the diode and verifying that its still all the way in and recessed slightly like it should be. Then you could put some arctic silver around the module where it contacts the heatsink. If you wanted to use some thermal grease on the diode itself so it makes better contact where it is pressed in the module make sure you use the ceramic kind that isn't electrically conductive. I have no idea if this would really help much but it could be worth a try if you simply cannot get a press.

If it came down to it I'm sure you could send your diode in its module to someone with a press and they could make sure its fully seated. You could also just send the whole laser to someone to take a look.

 

[The Baron]

I can guarantee that the module is making good contact with the heatsink because before it was decanned, it worked PERFECTLY fine. I just looked, it is neatly recessed. I will attach a picture in a sec to show you

Here are some pics that I attempted to macro with my glass lens
http://d.pr/i/p2l3
http://d.pr/i/lRbq

 

[ApexProxy]

Its hard to tell definitively from the pictures but it looks ok to me. I would try adding some arctic silver 5 between the module and heatsink and go from there. Careful not to use too much though. The goal is to make up some of the little gaps with silver where you wouldn't have had any contact before, not to make all the contact you get be from the AS 5. When you put it back together I would suggest putting the set screw in the heatsink a tiny bit extra tight so it increases the contact through the compound. Not overly tight so you bend the module though XD

 

[The Baron]

I just arctic silver'd the shit out of this thing and it now stays steady at 1W(@1.75A!). I am leaving it on right now for 5 mins and it seems fine. The output is just half of what it should be. Pic: http://d.pr/i/K8Zu

EDIT: Oh yeah, the whole host feels nice and warm now! I would say the heat problem is FIXED but my problem is still not solved. If it didnt do this before(the decan), why now? and the other obv problem is why is it half of its potential?

EDIT2: I am pretty sure this is producing WAY too much heat because in a matter of a minute the whole head of my Maglite will feel really warm. When it worked fine, I wasn't even able to feel the warmth at all after a 2 minute runtime.

 

[ApexProxy]

It could also be possible that the trauma of the decanning and then extreme overheating has permanently decreased the diode's output.