5W Driver Recommendations

I'm about to make a 5W light that I'd like to power off 5 rechargeable Sub C/D NiMH cells. My limited understanding is that this will give me an initial 7 volts, rapidly decreasing to and stabilizing at 6 volts until near depletion.

I do not want an initial brightness that quickly decreases. I want truly regulated brightness/current.

I would prefer it if the regulator "shuts down" after the batteries are nearly depleted to protect them. Outputs for a small (red) LED that indicates falling out of regulation would be highly desirable.

I'd like to slightly overdrive the 5W. I've considered some available possibilities, but nothing seems a perfect match to my requirements.

Please suggest some options. Work on custom electronics could be compensated. Thank you.

The MR circuit from the MR Hotlips sounds like just the ticket! Check it out in the Hotlips MR X thread.

The MR circuit contains the basic features that I'd like, but the 6 to 7 volts that I'd like to use don't seem enough to achieve regulation. Sounds like I need more voltage than the 5W emitter Vf to ever get beyond direct drive.

I think you've grasped the problem. Battery has to be greater in voltage than load or you need an up converter. The closer it is to load voltage (but still above my a necessary minimum), the lower the losses and higher the efficiency.

I'd suggest considering dropping to 3 cells and driving several 1 Watt Luxeons in parallel. If you can get a sixth cell into the battery, you're good to go with the five watters (within reason).

Doug Owen

Either 4 cells with a step-up (BadBoy) or 6+ with a step-down (DownBoy) would work.

With Vin close to Vf, you would need to design a custom regulator, most likely a flyback design. National Semiconductor has a nice flyback regulator controller, if you combined it with an opamp it would make a good constant-current source.

Specifically, the only real difference between a controller chip designed for C.V. and one designed for C.C. is the feedback voltage - Most C.V. controllers want 1.2v or so, which is WAY too much drop for a current sense resistor if you want any efficiency. If you use an opamp resistored for 10x gain, you can obtain 1.2v at the feedback input with only 0.12v across the sense resistor.

My question is sorta the same, so I'll add it here. ( sorry, dont mean to highjack the thread)

I want to direct drive a 5W in my 2D mag.

I have made a couple batt holders, that can hold up to 8AA's, or any number of AA's under that, with dummy cells.

I was thinking that 6AA's is good to run the 5Wer. Am I correct? I would like to be able to use rechargable nimh's, is that a problem? I heard something about rechargables having less internal resistance, and blowing the LS.

What do you recommend I do? Thanks.

Otokoyama
Sorry, in my haste to reply I forgot the MR is basically a step-down circuit. D'oh!!! /ubbthreads/images/graemlins/ohgeez.gif

Since you'll be near the Vf of the LED and about 6-7 volts, a wizard *might* work, but I'm not sure how well with a 5w.

Well, it seems I'll have to go with the original plan, a BB750, and maybe go from 5 cells to 4.

Thanks for the information.

If your Vf is near the full charge or withing .3V of it, the BB will operate fine. It will still run up to Vf + 0.3V or so.

So, if your 5W is 6.8V Vf you are safe to run a 5 cell 7V initial voltage with a BB and be fully regulated.

Wayne

Wayne: So a 5W emitter is likely safe with 5 NiMH cells--and maybe even 6--but I really should check the Vf to be safe. Thanks.