Keep frying my p7's.....why?

[nailbender]

I would expect that if the heatsinking was marginal it would show angry blue, and possibly be damaged rather than so quickly destroyed. Since it goes completely out rather quickly perhaps the current is waay too high, and it isn't a heatsink issue. The OP didn't mention heat, if the light was used for short periods then overcurrent is more likely. If for long periods then heat buildup is more suspect.

In either case the "J" bin will reduce the current and likely solve the problem. One graph I looked at showed current doubling for 0.2V of change (with zero circuit resistance).

Here's what I'd do:

Install a J bin. Make sure the thermal epoxy is thin, and double check the electrical isolation.


Measure the current at the tailcap. If over 2.8A add some resistance (probably won't be needed).

I agree with alan, I think the most important thing is to make sure the AA is thin and the LED is well attached. I use three 10000 mah NIMH batts all the time and never have any trouble. They are all direct drive and I test all for at least 45 minutes. I just sent one to CA and the member said he ran on high for over two hours. If there is the smallest crack between the Led and the heat sink or the adhesive is too thick it will heat up and turn blue.
I have seldom measured more than 3amps for extended periods of time from three NIMH.

 

[Amandrew]

I have to admit, the I bin that I had in there did turn really really blue. I checked the led and there was a gap between the led and the heat sink. I only ran it for 2 seconds though before I took the led off and re-adhesived the led. That could have been a partial reason for the led going bad. I get now what your saying steve about I bin with 3.6 volts going to it, of course will fry it. I guess I over looked the 3.25 to 3.5 voltage rating on the I bin, I thought not a big deal with .1 volts but I guess I proved myself wrong. (two times):(. Anyways. I did have a really thin layer of adhesive the second time and no more blue and never got hott after that. I also have a third p7 on its way, J bin this time! ;) I'll keep everyone posted on my findings with the amps and volts going through the led. Thanks everyone for your help, if it wasn't for cpf i'd be on my 3rd fried I bin. lol:grin2:

 

[Alan B]

Thanks for the update. The spec sheet says a few seconds with no heatsink is OK, so that probably did not hurt it. Good that you caught it quick.

After epoxying them I ramp them on a power supply to 2.8A, watching the color. So far I have not had one shift. The epoxy gap needs to be really small, so I made some tools to clamp it tight while it sets. Also check to see if it is electrically isolated from the heatsink just in case. Do that while it is setting up so if you have to remove it you have a chance before it gets too hard.

How did you remove the bad LED from the heatsink without damaging the heatsink?? Which epoxy did you use??

It is good you posted this thread, we can all learn from your experience. By doing this you saved a few other LEDs from a quick death, and perhaps encouraged a few more to try this mod with more understanding of how easy it is to do, and fix if the LED goes bad.

 

[Amandrew]

I removed the p7 pretty easily actually. I looked at the led and it had a gap between the led and the heat sink. dont ask me how that happend but I kept pressure on it the whole time. I must have not kept pressure long enough. The adhesive I used was the alumina adhesive from 4 sevens along with the anodized heatsink so I hope between the adhesive and the anodizing it wont short. Your advise has been more then helpful towards a succesful direct drive p7, thank you. Also anything else that is a common mistake or common problem please let me know. I have been doing nothing but reading on p7's for the last two days.:laughing:

Yeah I did a search on my problem and couldn't find anything. It could be my amature search abilities or there just isn't a topic like this. I hope my 40 bucks went towards helping others from frying thiers. I guess I can look somewhat possitive at wasting $40.00.:)

 

[yellow]

working on the wrong "errors":

1st: when the sink is anodized (which the ones of 4sevens' site I just checked are), You dont have to take ANY additional measure to isolate. Anodizing is electrical isolation.

2nd: pease, guys, I am breaking my fingers typing this FOR YEARS now: do not use arctic to APPLY/FIX the led.
It does not work good + needs much more time to cure as anyone thinks (if it ever cures when the outest layer - exposed to the air - cured).
What happens is, that the led gets powered too early and the non cured internal part starts to "cook", pushing the led away from the sink.
no connection to sink --> dead within seconds with getting blue
(if Your led turns blue, pls tell this with the damage report)
use thermal paste on the slug to apply the led, then glue it in with expoxy around the body (or use screws with a star)
all the parts, normal epoxy and thermal paste, are cheaper than arctic and give the opportunity to move the led around in position for endless time, before finally glueing it in. Arctic gives its 5 mins. This alone were the reason not to use Arctic.

3rd: I dont like the idea of parallel driving multiple die/led, but 3 A (or much less with a direct drive 3 Ni-Mh application) is a joke,
when everything else (=heatsinking) were ok, nothing happens
att: please give figures on current draw in the future


--> get another led, mount it w. paste + glue around led body, wait 24 hours for the glue to cure, check hadness of glue on its remains, fire it up and check draw