Overdriving the Cree R2 @ 1400mA??

I am thinking of building a D26-module with a Cree R2. I plan on using the AMC1735 1400mA driver since I want all light possible from this setup.
However I want to know how this affects the LED, since lifeexpectancy probably goes down a bit.
I can live with a minimum of 100 hours of use on this one, but is it possible or will the LED just burn??

Use a brass pill and fill the empty space with something like silicon based thermal glue (eg. Fujik from DX and others). I'm running 1.25-1.3A in my Derelight DBS without any problems. It might reduce the leds life some, or not.

I´ll try that and report back... will probably use either the thermal gooey stuff from DX or some AA..... I´ll start with a R2-module from DX, so the pill is brass from start...
What does the specs say about output from a Cree R2 @ 1400mA?

isn't aluminium a better thermal conductor than brass?

Der Wichtel said:

isn't aluminium a better thermal conductor than brass?

Click to expand...

Yup, it is http://en.wikipedia.org/wiki/List_of_thermal_conductivities

Greetz Johan

Yes, it is. But brass is denser and heavier and that seem to work better for me in the small dropin pills, the brass pills also have thicker construction than the corresponding alu ones I have used. But in theory alu should be better..
Stefan

StefanFS said:

Yes, it is. But brass is denser and heavier and that seem to work better for me in the small dropin pills, the brass pills also have thicker construction than the corresponding alu ones I have used. But in theory alu should be better..
Stefan

Click to expand...

How about copper than, is also heavy and dense but with better thermal properties than alu.

Johan

SUBjohan said:

How about copper than, is also heavy and dense but with better thermal properties than alu.

Johan

Click to expand...

Copper would be better, but then you run the risk of galvanic corrosion between the copper and alu if an electrolyte or acid water or whatever comes into the light. Inside a light that's not likely, but I think manufacturers want to be safe rather than sorry. There are also a LOT of different brass alloys available, some with up to 85% copper in them so it's not an easy topic.

As an example the MRV has a massive heatsink made from brass. Compared with the Tiablo A8/A9 that has aluminium heatsinks of about the same volume, the MRV is more effective at keeping an SSC P7 cool and has slightly longer runtimes in identical setups, the MRV design seem to simply be more efficient when it comes to lead heat away. It's also heavier because of the brass..

Scattergun said:

... What does the specs say about output from a Cree R2 @ 1400mA?

Click to expand...

335 lumen


tl

ive been running my aspheric mag with a dhs heatsink and a AMC1735 1400mA driver for a while now.....without any problems just....the two 3V rcr123 "fireing the driver become ugly hot.........and after 5 min the thermal protection kicks in :duh2: ....drivin beyond specs :shrug:

but what can i use instead? with 18650 the driver doesnt regulate to 1,4A.......
only if i feed it with more than 5V :thumbsdow

Yes the current is only limited by the die temperature and the bond wires.

So if you can keep the led cool enough, then you can pump as much current as you wish through the led until the bond wires can't hold the current.


PS @ NOS: Na, auch bei Ledstyles unterwegs :thumbsup:

I have a similar setup going now with a SSC P4... I know the SSC is not as good as the Cree when it comes to heatsinking, but has anyone tried using it at 1400mA??

Scattergun said:

I have a similar setup going now with a SSC P4... I know the SSC is not as good as the Cree when it comes to heatsinking, but has anyone tried using it at 1400mA??

Click to expand...

Sure, I run 70 XR-E's at 1800mA plus in the DB70. It has a few hours on it and still going strong. Has it lost max lumen output since being run at that power, I bet it has, but how much, I don't know.

Cheers
Dave

SUBjohan said:

Yup, it is http://en.wikipedia.org/wiki/List_of_thermal_conductivities

Greetz Johan

Click to expand...

Not to be cheeky, but diamond has awesome thermal conductivity, just sucks the heat away...

wechnivag said:

Not to be cheeky, but diamond has awesome thermal conductivity, just sucks the heat away...

Click to expand...

Wonder why my wife thinks that the 10 carrat diamond on order is for her. :lolsign:

taschenlampe said:

335 lumen

Click to expand...

Two driven @ 700mA each = 1400mA = 421 lumen

The lumen figures you guys are throwing around is with the LED die temp at 25C. I guarantee you that you'll be far from that when you drive the sucker at 1.4A, especially with a setup like a drop-in where you get very little heat transfer. I bet you will not see any increase in output driving at 1.4A since the heat generated will diminish the light output so much that it will probably be no brighter than driving it at 1A.

I'm sure to get flack for this but Surefire has done the right thing driving their drop in moderately. I think even 1000mA is on the verge of too much.

CM said:

The lumen figures you guys are throwing around is with the LED die temp at 25C. I guarantee you that you'll be far from that when you drive the sucker at 1.4A, especially with a setup like a drop-in where you get very little heat transfer. I bet you will not see any increase in output driving at 1.4A since the heat generated will diminish the light output so much that it will probably be no brighter than driving it at 1A.

I'm sure to get flack for this but Surefire has done the right thing driving their drop in moderately. I think even 1000mA is on the verge of too much.

Click to expand...

And I'll probably get flack for this but Surefire has done the wrong thing in even using those drop-ins for LEDs. Like you say their heat transfer ability is less than optimal. I know people like to be able to easily switch them out but there is a much better solution. It is not a big deal with incans where the heat goes out the front of the light but LEDs are a different animal and need a different solution.

saabluster said:

And I'll probably get flack for this but Surefire has done the wrong thing in even using those drop-ins for LEDs. Like you say their heat transfer ability is less than optimal. I know people like to be able to easily switch them out but there is a much better solution. It is not a big deal with incans where the heat goes out the front of the light but LEDs are a different animal and need a different solution.

Click to expand...

LOL, I wasn't going to go to such extreme but I hear you. At least Surefire drop-ins seem to behave like they have thermal protection since they drop back to a lower (regulated) level in plastic lights like the G2's. You can greatly enhance the heat transfer of the cheap modules by using foil. I do this with the ones driven to 1A.

Now to all that want to drive the ****ens out of these drop ins, listen up. I modified one of the cheap R2 drop-ins from 1A to 465mA. The output was much flatter and runtime went to 2 hours 45 minutes compared to one driven at 1A. The 1A one had nice output the first few seconds (about 2x brighter) after which it dropped off big time and settled to about 1.35 times that of the one biased at 465mA after a few minutes of cooking itself. My guess is that you'll get similar results when driving it at 1.4A since the driver is even less efficient and you'll heat up much faster, driving the output lower faster. Using foil helps a lot but you still won't improve efficiency and you'll just delay the drop off in output.

The SSC P4 I plan on driving at 1,4A is going to be mounted to a H22A heatsink... will this do any difference? And will three Alkaline D-cells be able to sustain a regulated 1,4A to the LED through a 7135-circuit?

I gave a light to a friend that runs an SSC P4 at 1.2A on max, and he hasn't reported any problems yet. It's on one of H22A's DHS heatsinks in a D Mag. It's not quite 1.4A, but pretty close.

The alks might be able to do that, but I'm not confident of that. If you want, you could try an unregulated light with three D NiMHs. And heck, if you're doing that, you might as well use a P7 instead of a P4 and get about four times the output. However, the P7 will then want a glass of milk...