Wounded P7

[Alan B]

I'm helping a friend is building a P7 flashlight, and he apparently got the polarity wrong on the LED when he was testing it. He was using 3 each D alkaline batteries to power it at the time, and it did nothing that he could see. He subsequently decided that he had the polarity backwards.

I hooked it to a variable regulated metered power supply with the correct polarity and it acts somewhat strange. The P7 begins to light up at about 200mA. All four dies appear equally bright. As the current is increased it gets brighter, though we didn't have a way to compare it quantitatively. The supply was only 500mA so we did not test beyond that.

It acts like there are about 12 ohms of parallel resistance. This causes the forward voltage to be inadequate to light the LED until the current is about 200 mA.

Connecting the batteries in reverse across the LED probably put too much current into the static protection diode and changed it into something like a resistor.

Anyone have similar experiences?

 

[uk_caver]

Re: Injured P7

If the P7 is like the P4, it has a static protection diode built in that conducts at even small reverse voltages.

Speaking from past P4 experience, it's certainly possible for even low reverse voltages to either totally wreck the LED (burn out the bond wires?) or just burn out the protection diode until it's open circuit, leaving the LED working fine.
I dare say it would also be possible to damage the protection diode to the point where it acts as a partial short-circuit even for forward voltages.

If the LED is effectively unusable, you could always try *really* burning out the diode.

As an experiment (for use a circuit where I really didn't want the diode there), I tried connecting some P4s to a 3-cell NiMH battery pack, which could provide a lot of current.
On 5 or 6 emitters, I got a 100% success rate - the diode (tiny dot in the corner of the die) rapidly flashed out of existence.
Encouraged by that, I then tried exactly the same on some stars, and got a 100% failure rate - the diode glowed, but didn't flash and die, and the LEDs were then nonfunctional.
I guess the difference must really be down to luck or maybe some kind of batch variation, since on the emitters, the effect was so rapid that it couldn't have been due to the lack of heatsinking.

If you were actually at the point of discarding the LED, you could always try reverse-connecting it to a few NiMH cells for a short burst or two, and seeing if that makes things any better.

 

[Kestrel]

Re: Injured P7

It might be ok, but the following is another possible diagnosis:
http://www.youtube.com/watch?v=qJQwHwP0ojI

 

[Alan B]

Re: Injured P7

It might be ok, but the following is another possible diagnosis:
http://www.youtube.com/watch?v=qJQwHwP0ojI

I can positively state that the P7 in post #1 is not dead. It works but there is about 200mA lost somewhere (at 2.5V), presumably in the reverse diode. Above that current level the LED gets quite bright. Perhaps this reverse diode will fail in forward bias..

What is the reverse voltage capability of the power LED dies in the P7?? I did not find a rating for that, but perhaps Cree has one for the dies?

I suggested to the owner that he put it in a 3D M@glite with a regular switch and no controller, a basic DD light. At 12 ohms the current loss is about 15% which is still a pretty useful light. We'll get another heatsink/P7 for the D2Flex he is building.

It is quite interesting to hear other's experience here. Thanks.

 

[Kestrel]

Re: Injured P7

Hey, sorry, just wanted to send a smile this afternoon.

 

[Alan B]

Re: Injured P7

Hey, sorry, just wanted to send a smile this afternoon.

No problem, haven't seen so much compressed StarTrek in years. Brings back memories. Had no idea they had so many casualties!