Hyper-bright light: feasibility and potential interest

[Fallingwater]

I got an idea for an interesting project.
I'd like to build a 4 to 7 (still have to decide) emitter overdriven CREE light with forced air cooling (read: tiny fan), with power provided by a RC-grade lithium polymer battery. I still have to decide if there will be any regulation or not (mainly depends on whether I find the appropriate pre-made circuits for non-exorbitant prices).
Size would be roughly 2-C, but with the design I have in mind I'd need a custom-machined body (it very definitely won't be a Mag retrofit or similar).

I only have the rough idea and design in mind, but I think it might turn out quite interesting.
I don't have all the information I need to proceed, though. For example: how hard can one overdrive CREEs while still having a decent bulb life? How efficient are the various converter boards, or to put it better, how much of the current draw on the battery can I expect to be wasted as heat by the converters? How expensive is it to have someone vaguely competent machine a flashlight body? (this could be a serious problem if it turned out to be more expensive than I can afford...)

Another thing: the project is still only in my mind and the only sure fact is that it's going to be damn bright, but just in the case it turns out to be a decent product, do you think there might be any interest to buy some? Mind you, it wouldn't be a light for casual use, more like a mini lighthouse in the palm of your hand for those who want to SERIOUSLY illuminate stuff.

Don't take this as a serious business interest, though; I'm just probing the idea, and yes, I do know selling lights is not as simple as stuffing a few emitters and a battery in a tube and putting the result on eBay.

 

[Lighthouse one]

I think your idea already exists in several different forms. Someone sells a heat sink adapter for 3 or 4 cree's into a mag. The cree's can run 1200 ma with good heatsinking....but 1000 ma with 3 or4 cree's is real strong already.

I've also seen DX and Kai are selling regulator boards for over 1000ma. Good luck!

 

[Gunner12]

Something like this(Elektrolumens Stunner)? Or this?

 

[2xTrinity]

I don't have all the information I need to proceed, though. For example: how hard can one overdrive CREEs while still having a decent bulb life? How efficient are the various converter boards, or to put it better, how much of the current draw on the battery can I expect to be wasted as heat by the converters? How expensive is it to have someone vaguely competent machine a flashlight body? (this could be a serious problem if it turned out to be more expensive than I can afford...)

Anyway, if you're doing 2xLiIon, your most efficient way to go would probably be to use a separate regulator on each emitter, and wire those all in parallel -- that way, you would be using a buck (voltage-stepdown) to do the job. Those tend to be quite efficient, up to 95%. Buck will also give you the flattest regulation.

The 800mA drivers from DX are not bad at all, they're not the most efficient, but they're pretty good (near 90%), and cheap -- $7 for four. That's not an overdrive as you'd like, but with the money you save over anyone else's drivers, you could always afford to make up for it by adding more emitters. Running 7 emitters lightly will be a lot more efficient and reliable than driving 4 emitters like crazy to get the same amount of light.

I've also seen DX and Kai are selling regulator boards for over 1000ma. Good luck!

The ones over 1000mA are linear regulators, which are limited in voltage so they can only power one emitter at a time -- if you try to feed them with 2xLiIon, they will be extremely inefficient, wasting nearly half the power. If oyu wire the batteries in parallel, and essentially set up all the LEDs in parallel as well in a direct-drive scheme, that might work. The regulator will prevent the emitteres from being overdriven too much while the batteries are fresh, and then the light will be direct-drive for the majority of the runtime, gradually getting dimmer.


Though using a separate regulator on each LED might sound kind of excessive, one plus this gives you is that you coudl be able to switch these off and on separately, perhaps in two banks, to give multi-levels. Fact is, you can buy 10 DX drivers (always get spares of DX stuff, in case you get a dud in the bunch you don't want to be held up 2 weeks to order more) for far less than the price of one driver that could handle an application like this.

 

[Fallingwater]

Something like this(Elektrolumens Stunner)?

Similar in concept, rather different in design. And way less expensive, if the machining cost doesn't ruin the plan.

Or this?

No, not THAT insane. Mine would be much more portable. And dimmer, of course, but still much brighter than the average 2-C light...

Anyway, if you're doing 2xLiIon, your most efficient way to go would probably be to use a separate regulator on each emitter, and wire those all in parallel -- that way, you would be using a buck (voltage-stepdown) to do the job. Those tend to be quite efficient, up to 95%. Buck will also give you the flattest regulation.

Hmm... I was thinking of rewiring the pack in parallel, but your idea sounds way better, and the variable output would probably be a good idea. Thanks!

The 800mA drivers from DX are not bad at all, they're not the most efficient, but they're pretty good (near 90%), and cheap -- $7 for four. That's not an overdrive as you'd like, but with the money you save over anyone else's drivers, you could always afford to make up for it by adding more emitters. Running 7 emitters lightly will be a lot more efficient and reliable than driving 4 emitters like crazy to get the same amount of light.

But overdriving 7 is even better

The ones over 1000mA are linear regulators, which are limited in voltage so they can only power one emitter at a time -- if you try to feed them with 2xLiIon, they will be extremely inefficient, wasting nearly half the power

You say the ones above 1000ma are inefficient linear regulators... what about exactly 1000ma ones?

 

[stevetexas]

keep us up to date - I'm really interested and working on a 6 emitter light myself... sort of an ongoing project

 

[2xTrinity]

You say the ones above 1000ma are inefficient linear regulators... what about exactly 1000ma ones?

I didn't mean that linear regulators are inefficient because they are over 1000mA. They are inherently inefficient (they waste power by design). I meant that the only drivers that DX happens to sell that are over 1A are linear regulators. However, they are a better alternative to an outright direct-drive with a fixed resistor -- treat them more like a "smart" current limiting resistor than a regulator.

 

[Fallingwater]

I didn't mean that linear regulators are inefficient because they are over 1000mA. They are inherently inefficient (they waste power by design). I meant that the only drivers that DX happens to sell that are over 1A are linear regulators. However, they are a better alternative to an outright direct-drive with a fixed resistor -- treat them more like a "smart" current limiting resistor than a regulator.

I understood what you meant. Sorry for the poor phrasing of the previous answer.
What I meant is: you say DX boards over one amp are linear regulators and 800ma ones are proper buck circuits. Are there 1a boards on DX/Kai/whatever that are NOT linear regulators but proper buck circuits?
Thanks again

 

[LukeA]

I understood what you meant. Sorry for the poor phrasing of the previous answer.
What I meant is: you say DX boards over one amp are linear regulators and 800ma ones are proper buck circuits. Are there 1a boards on DX/Kai/whatever that are NOT linear regulators but proper buck circuits?
Thanks again

This? I know, a little expensive for 4-7, but it does what you want. ($20.00 ea.)

Or this? ($16.50 ea.)

 

[2xTrinity]

I understood what you meant. Sorry for the poor phrasing of the previous answer.
What I meant is: you say DX boards over one amp are linear regulators and 800ma ones are proper buck circuits. Are there 1a boards on DX/Kai/whatever that are NOT linear regulators but proper buck circuits?
Thanks again

As far as I know, the only proper buck cicruits they have are 800mA.

If you're going to spend $20 on a single regulator (such as the ones from Sandwich Shoppe) Instead of getting Downboy/SOB (which is really only built to handle a single emitter from the looks of it) I would recommend going for something more robust, like the Shark Driver. That one is actually a boost, however, the advantage of that is that you can drive 5 emitters in series at 1A. The advantage of hooking all those in series is that using an external 20K-ohm potentiometer, you will have a variable brightness control.

 

[Fallingwater]

One hopefully last question on this matter: I'm thinking of using the 800mah DX buck drivers with two cells in series for normal operation, and direct-drive with the two cells in parallel for super-bright turbo operation, with resistors to limit current to 1200mah for each emitter.
The question is: how much power would be wasted as heat by the resistors in the turbo configuration?
Sorry about asking such probably elementary questions, but as I said when talking about resistance, ohms and watts my brain hurts...