buckpuck question

[mofiki]

I'm using a Lux-Drive 3021 1000mA buckpuck to drive one new K2 led. The driver seems to be getting pretty hot after only one minute and it seems to keep getting warmer. I don't remember it ever doing that when I had it connected to a four led circuit (two series & two parallel). I have a 7.4v 4000mA Li-Ion battery that I used in both circuits. It's peak volts measure 8.4v. I was wondering after reading the spec sheet on the drive about the margin input spec. It says Input Margin (350 unit, add to LED Vf)...2v. I understand the driver also uses 1.5 to 2 volts but the way this is stated in the spec and the specs do mention the 2v drop of the driver later in the discription so I was wondering if this ment something else.
Does this mean that the battery voltage should be close to the LEDs forward voltage plus 2v. Does the battery voltage matter at say 5 or 10 volts higher than the Vf of the LED even though the maximum input of the driver says it can take up to 32v? I'm just worried about running this thing while it's getting so hot.

Well, no worry now because the led just smoked. I pulled everything apart to check for shorts and when I was sure I had none I appied power to the led again and poof. I don't know if this was because it wasn't heatsinked now. I pulled the led off the heatsink because the meter did show something ( short beep and a value readout) when set to diode test but only in one direction. When I reversed the test leads nothing. I set the meter to ohms and got nothing either way. Went back to diode test and the leads would cause a beep again. I would guess that the heatsinking had something to do with that. That is why I removed the led from the heatsink. Sorry for the length of this post but I hope I gave a good enough explaination of the problem.

 

[mofiki]

Well, sorry to waist anyone's time. I check the V out voltage of the drive and it measured the same as the battery. I would imagine that means that the driver is screwed right? Obviously there isn't an led out there that can take 8.2v so I shouldn't have gotten that high of a reading right? I'm not sure. I never checked voltages to compare before things went bad.

I seem to be answering my own questions here. I found some information on how drivers work and learned that I would read the battery voltage across the Vout pins of the driver. The driver limits the voltage based on the current requirements and duty cycle so I hooked up the drive to an old luxeon emitter, had to add a resistor to the cntrl -ref pins, and everything appears to be normal. I'm using 4 RCR123 Lithium cells so back to the original start of the topic; it must not matter about the input voltage versus the output voltage either.

What the heck would of caused the driver to get so hot???

 

[Mel_PL]

I've used many buckpucks and never had any problem. They are very reliable. Did you pay attention for proper wiring?

-- M.

 

[mofiki]

As I mentioned in the last post, I think it has been determined that the driver is ok. I'm leaning more towards improper heatsinking. I think there is a reason why so many builders use the Artic Alumina and not JB Weld. My meter picked up a reading in the diode setting too. I can't figure how that would have mattered when none of the body parts of the lights housing had contact with the battery. I wonder now about thermal run-away if the led doesn't disapate the heat properly.