Well, I can talk about it… some. Just can’t talk about the actual application. And no photos since that might clue into the app. It is way cool though… (but not temperature wise). Effective thermal management of big honkin’ LEDs is a royal pain in the alien probe hole…
scaru is absolutely correct. Max input voltage is 25V, although that number is a LITTLE conservative. The module will not blow up at 25.1, but it’s not guaranteed to function for any length of time above 25V. 33.6 is way out of spec, and you can expect things to start blowing up at that voltage.
the 8 primaries (8x3=24) would be perfectly safe as long as they really are 3V MAX cells.
Which do you reckon would be more efficient, a 3S li-ion pack at 11.1v nominal, or a 5S li-ion pack at 18.5v nominal? The 11.1v will be closer to the vf of the LEDs but as you're not running a linear device, does it matter?
If this module’s driver is similar to other buck drivers, then 3S Li-Ion pack will drop out of regulation at some point, while with 5S, cells will get drained completely at stabilized brightness.
I would have thought a 4S pack would be perfect Especially a 4S2P pack if you want it to run for any length of time on full…
As for surface area, 1sq.in per W is really cutting it fine. Perhaps not so much in winter as the external temperature is lower, but summer night rides will really push it. I think both my lights are around 1.5sq.in per W and they start to struggle only on the hottest night rides at low speeds. Then again, trying to hit 2sq.in per W on a 30W bike light is really hard work without it becoming huge. Still, at least you should have a good thermal path if you’re machining the housing out of solid Alu. Good luck and don’t forget the build thread
Efficiency of the driver is a tiny bit higher at 11.1 than it is at 18.5, but the difference is completely insignificant. We’re talking about a percent or two difference.
When deciding on a cell configuration, I wouldn’t worry about efficiency differences. Use whatever best fits your application.
Thanks for the replies. That's good to know as I have a 5Ah 18.5v waiting to be used.
Pilot, I have figured out the Flow Sim at last! My mesh was too coarse, not allowing the internal surfaces of the cooling fins to be taken into account. I had left it on Automatic and didn't check the Narrow Fin box!
My initial housing stabilised at 85*C. Not bad.
Matt, of course I'll start a seperate build thread
Hey Pilot, got a design sussed in Solidworks. Head on air flow at 6m/s at 30W is 71*C. I'd probably be going faster than that with the light on max anyway. Oh yeah, ambient temp is 20C. Probably be cooler than that in reality, having a rubbish summer here in England!
But I was thinking, 30W is what the light consumes. What percentage of the output is heat? Some of that power must be light right?
Some energy is, of course, converted to light. The percentage of it, however, is quite minimal. White LEDs are still not very efficient at turning electricity into lumens - they’re far more efficient at turning electricity into heat. It’s not 30 watts of heat… probably more like 26 or 27… For thermal analysis, however, I usually just assume it’s all heat - gives me a little margin or error on the favorable side anyway.
Some colors are more efficient than others. Royal blue, for instance, is quite a bit more efficient than white. I can’t speak intelligently about other colors, however, I just don’t know…