Reliability of overdriven XR-E?

Has anyone overdriven the cree XR-E (e.g. more than 700mA) for a long period of time now? Are there any ill effects? With the low thermal resistance, junction temps should be well below the maximum when mounted on a typical heatsink. Is the 700mA a limitation on the bond wires or is more likely due to damage caused by localised hotspots when running at greater than 700mA? Has anyone killed localised phosphor

?
Ta

I haven't seen any tests, but I do have a Downboy 1000 enroute to run an XR-E :-D

I have read on this forum that they have been commonly driven by DB825s - any feedback from someone using this approach?

Well if anyone has some dead Cree XRs or XR-Es I'd love to have them! I'd like to experiment with some handling/machining/mounting techniques that I know will destroy a few in the process. I'd rather be destroying already dead ones.

I have two Cree XR-E leds and one 5Watt SSC Z-Power in a headtorch, all in series and I drive it at about 950mA (3021 BuckPuck rated 1Amp).

Before this I have had 3 Z-Powers (capable of 1400mA) in the light, and I just didn't replaced the driver.

It was turned on max. 3 hours continually, but many times, total time about 30 hours. I don't think the light output has changed. Reference light is 20 Watt Silva halogen (comparable output). But I have to say that the efiiciency is probably much lower than within 0-700mA range. The gain of light from 700 to 950mA is detectable but small.

BTW, the SSC company uses the same LED chip (Cree EZBright) up to 1A:
http://ledsmagazine.com/articles/news/3/12/2. The chip maximum DC current is 1A. It's all the matter of cooling.

Well either these are tougher than the earlier Luxeons or not enough crazed CPFers have had a chance to play with these yet. Early on with the Luxeons, folks here would kill them by the bucket load. I'm still trolling for some already killed ones if anyone is willing to part with them. The earlier Cree XRs with the same package would work for my needs too.

Sounds like VoiToi has put the most hours run on the Crees so far... Rechecking the output would involve careful Before/After testing with a light meter.

I was testing light output by bouncing the light off the ceiling - when the Cree is driven at 1200 ma, it is maybe 70-80% or so more light than the readings at 600ma.

The Heatsink design used while running the LED will have a big effect on it's life/output.

Newbie may have some statistics here, as I remember seeing a chart showing the vf shift of a cree he was testing over, if I remember correctly, a 30 day period of time.

I haven't used my A19 Cree (DB825) for more than 30 minutes total so far (still on the original cell) so I don't have any data to share yet. I have about 30 minutes on my Cree'd MilkySpit ML-1 driven at 350-400 mA.

The a-19 from the shoppe is driven at 819mAh. Somewhat higher than the emitter was designed for.

I don't understand why they picked this drive level. Why not just run the emitter at it's maximum rated power and go with that?
In any event I have run several batteries through mine with no apparent ill effects ...yet. Too small a sample set.
Yaesumofo

I'd say going up to 1000ma isn't going to make a big difference on them, at least not for flashlight use. Luxeons last a long time, and if the Cree is as durable or better, then no worries.

The die is rated for 2A (or more? cant remember) newbie said, but the package isn't. I asked if the bond wires in the current package can take 2A and he said yes, and has tested at 2 amps.

I plan to run my spotlight (bin P2 WG) XRE at 1.5-2amps. It's on a decently sized pentium heatsink. At 1.3amps continuous the heatsink barely warms up, and after disconnecting power it cools off almost instantly. So I know I'm nowhere near the limits of my heatsink. So I think I'll go for 2amps, as even then I doubt it will be even as warm as most led flashlights.

LightBright said:

Sounds like VoiToi has put the most hours run on the Crees so far... Rechecking the output would involve careful Before/After testing with a light meter.

I was testing light output by bouncing the light off the ceiling - when the Cree is driven at 1200 ma, it is maybe 70-80% or so more light than the readings at 600ma.

The Heatsink design used while running the LED will have a big effect on it's life/output.

Click to expand...

What was the voltage on that test?

About 3.40V at 600 ma and 3.65V at 1.2 amps

LightBright said:

About 3.40V at 600 ma and 3.65V at 1.2 amps

Click to expand...

2.04 & 4.38 Watts respectively...Cree is rated for 2.63 Watts
Millermods L1p w/ a L2p body would be 3.9 Watts, sounds like fun but a L1p head is a far cry from a Pentium heatsink
Might be safe for short bursts...

I'm still looking for some dead ones. Folks, don't make me have to go and kill some myself

If you do, make sure you try to bracket bond wire fuse current while you're doing so....

Doug S said:

I'm still looking for some dead ones. Folks, don't make me have to go and kill some myself

Click to expand...

I may have to bump this one weekly until some of you guys fess up to killing some. But you have to kill'em first.

Doug S said:

I may have to bump this one weekly until some of you guys fess up to killing some. But you have to kill'em first.

Click to expand...

Whoops, I let this go two weeks without a bump. I am still in search of a few dead but mechanically intact XRs or XR-Es. In fact, it you have two or more you will part with I'll send you a new in package Chinese ARC 1AAA clone for every pair. I'll verify the clones to be good as these have a known defect rate.

Why not just buy a couple of Cree and Seoul emitters glue them to a decent heat sink connect a bench power supply and crank up the juice and run them until you get a failure.
NEWBIE has been running tests on these emitters and has cranked them up.
It wouldn't surprise me one bit if he had a couple of "dead" emitters. Maybe you could ask him for some.

Yaesumofo

Doug S said:

Whoops, I let this go two weeks without a bump. I am still in search of a few dead but mechanically intact XRs or XR-Es. In fact, it you have two or more you will part with I'll send you a new in package Chinese ARC 1AAA clone for every pair. I'll verify the clones to be good as these have a known defect rate.

Click to expand...

yaesumofo said:

The a-19 from the shoppe is driven at 819mAh. Somewhat higher than the emitter was designed for.

I don't understand why they picked this drive level. Why not just run the emitter at it's maximum rated power and go with that?
In any event I have run several batteries through mine with no apparent ill effects ...yet. Too small a sample set.
Yaesumofo

Click to expand...

Actually, they were 917mA and mine measured 940mA to the emitter.

yaesumofo said:

Why not just buy a couple of Cree and Seoul emitters glue them to a decent heat sink connect a bench power supply and crank up the juice and run them until you get a failure.
NEWBIE has been running tests on these emitters and has cranked them up.
It wouldn't surprise me one bit if he had a couple of "dead" emitters. Maybe you could ask him for some.

Yaesumofo

Click to expand...

I don't have dead CREE yet, but I have a dead Seoul P4 that had delamination problems at first power up.

Doug S said:

Whoops, I let this go two weeks without a bump. I am still in search of a few dead but mechanically intact XRs or XR-Es. In fact, it you have two or more you will part with I'll send you a new in package Chinese ARC 1AAA clone for every pair. I'll verify the clones to be good as these have a known defect rate.

Click to expand...

Time for my weekly bump of this one. I'm concluding that these are either really tough or the knowledge and skills of the average member working with these has greatly increased over the early days of the Luxeons.