Cree XR-E vs SSC P4 (at 1A drive)

This is inspired by the P4 evaluation thread, but is going a bit OT for that, so I thought it best to start a new thread.

I've seen testing of Cree XR-E at 1A and over. This is of course outside the spec of the LED, but given sufficient cooling is it feasible to run the LEDs at 1A, or will it do them harm? Since the SSC P4 with the same die appears to be specced for 1A, is there some other issue (slightly worse heat transfer?) which makes the Cree unsuitable for driving at 1A?

If it is feasible to run a Cree at 1A, ignoring the current non-availability of Cree Q bins and any issues with optics (as I think I have options for both) which would be the best for running at this power level?

Finally back on the P4, for driving at 1A do I have to have really perfect heatsinking? Given the slug is not neutral, it seems I need to put a thin layer of thermal epoxy on the base to insulate it before mounting (I'm running multiple series LEDs on a single heatsink). Is this a really bad idea for running at 1A, and is there a better alternative?

The Cree's bond wires will fuse open at or above 2A. If you can keep it cool, that is the only limitation. Understand that the flux/drive current curve indicates diminishing return above the rated 700mA. It still gets brighter though. Newbie projected just under 220 lumens at 2A using a P3 binned Cree XR-E HERE. I imagine a QIII binned XR-E will be astonishing at 2A!

If I can find a driver that will push close to 2A at 7.4Vin, I will make a wicked weapon light for my Glock.

Anyone have any ideas for a ~2000mA driver?

Bernie

I've got a XR-E being driven by a Downboy1000 right now. Hopefully later this week I'm going to do some long runs and check temperature. But in normal intermittent use it never heats up noticeably.

Not projected, measured- yes, I did run it that hard for the measurements. I had it direct soldered to 2" by 3" by 0.165" thick copper sheet metal. No other heatsinking for that test, which might have helped at the upper end. I am not recommending anything whatsoever, you go beyond the specifications, you are entirely on your own. The McGizmo A19 XR-E light ran them at 917mA (I think it was 940mA actually measured on mine? It is in one of the threads around here).




From what I've seen on the Seoul P4 so far, I'd keep it under 1A, it is made a bit different than LEDs we are used to.

.

Wow Newbie! Are you still seeing blue spots? It must be nice to have a bench power source. I have a similar copper plate mounted to a fan-cooled finned aluminum heat sink. I am thinking of making a LED test station using it. What features do you consider mandatory for even a "basic" variable power source when testing LEDs? Understand that I am cost conscious, and therefore not considering a full-featured power source.

Bernie

Humm, voltage adjustment and current limit adjustment, with 10 turn potentiometers (or mod yourself).

Mr. Al had one he did a while back, with all the design details in the thread, with schematics.

Not sure how to locate it in the archives though...

I'd also consider Kelvin connecting to the LEDs, or you will get errors in your measurements. Two contacts for current on LED pins, plus two more to LED pins to measure Vf. As you get higher in current, the loss in the leads and the contacts will make the power supply meter inaccurate- a common error people make.