LED drivers and Accessories you want, but don’t exist

The PyUPDI thing looks interesting. With a need for only 3 pins instead of 6, it should be significantly easier to design drivers which support acupuncture-style reflashing… particularly since one of the pins is VCC and one is GND, so only one actual programming via should be needed on linear drivers. This also allows using physically smaller chips, so more room on the driver for other stuff.

I’m assuming a usb-serial adapter of some sort would work.

It also looks like gcc got support for attiny1616/1617 about three months ago, with a release target of gcc 8.0, so as soon as that’s widely available it should be relatively straightforward to build binaries for those chips.

A proof of concept could probably be done now, using pre-release tools, but the barrier to entry seems a bit high for common use.

Wow, I really like that MP3431, good find!

That has a LOT of potential.

Now that’s serious stuff!

Quad XHP35 HI under Carclo 10622 with well over 4 klum OTF and little under 100 lm/W while delivering those lumens. 1000 lm for 1.5h from a single 18650 cell. Flooooooody.
C8 or similar with dedomed XHP50.2 doing 3500 lm and (???) 100 kcd, sustaining 2000 lm for long.
C8 or similar with XHP35 HI doing 2500 lm and XXX kcd.

My favourite UT02 with XHP50.2 doing 3500 lm and 180 kcd?

And those numbers are conservative.

Too bad that with one mode it needs a big host to dissipate the heat…

TA, if you want to do some development with the MP3431 with your own boost driver series PM me, I can give you some info on the chip so you can have a board ready for when they go on sale, hopefully soon.

You can already order them on Mouser

Yes, this is the best boost IC I have seen so far for our uses, I am sure that I will start playing with it once the next gen MCU’s ave matured. And once I have some spare time again, been too busy to even read BLF the last few months.

Kind of in a holding pattern until those MCU’s mature anyways. The features they offer are good but the most important thing is the size, we simply can’t progress any further with the large SOIC8 MCU’s, there is no room for anything else.

Now if we could just find a similar speced IC for a buck and buck/boost driver we would be set.

MPQ8633B would be the Buck IC with similar specs, I have info on that one too. No high power Boost-Buck that I am aware of yet, but the MP28164 is a 4.2A Boost-Buck chip.

Not affiliated with MPS in any way, i just found their catalog and thought they make some nice chips for flashlight drivers.

Buck should be possible to find, but buck-boost will be tricky…

Both of those are interesting options. The MPQ8633B looks real good but it is hard to tell from the cover sheet, is the output voltage limited to 5.5v? Or is that just the reference voltage?

The MP28164 looks like it would make a great general use driver for small lights as you don’t need more then 4A anyways in most cases. I also like how few external components it needs.

Vout max on the MPQ8633B is 0.9 of the input, or 5.5V depending on the input voltage.

Yeah, thats how I read it as well. Sad it would not work for higher voltage LED’s, if it supported outputs up to 16V like the input it would be great.

The output voltage is limited to 5.5V, good catch, I didn’t see that. In that case, the MPQ8632-20 Is probably the better candidate, up to 13V output and similar in other specs. Little bit larger footprint though. The MPQ8632-12 has the same size footprint as MPQ8633B but a 12A limit (still probably enough).

Hmm, 13V is not quite enough for a 12V LED at reasonable currents but at least it would handle 6V LED’s.

Who knows maybe they have some wiggle room in the voltage tolerance and could handle ~14-14.5v, kinda like how we push everything else past the datasheets as well lol.

There are certainly plenty of high powered Boost-Buck controllers out there, but many require 4 external MOSFETs which makes space a problem. Using some super small surface mount FETs it might be possible to build something in the 30mm range. LTC3780 comes to mind here.

I’m willing to bet the MP8632 would output beyond 13V by a bit, similar to how the TPS61088 will actually max out around 13.7V instead of 12.5 like the datasheet says. We do tend to operate a lot of things outside of specs here. Haha.

Yeah, those FET’s really eat up the space which would already be tight without them. Thats what makes these chips so attractive. Plus the lack of FET’s reduces the complexity which is always good for small designs.

Maybe I can get an ATTiny13A on the MP3431 board, for biscotti-HD. To control the IC I would use the method used in the H2-C.

How about ATTiny85 for e-switch?

Though either would be lovely…

The 85 isn’t available in a package small enough to fit the underside of the board, I need 3x3mm, even 4x4mm is too big.

Which parts list for this one
I need it fot 1.4A UV LED with 4.1V forward voltage in 17mm

OK, I thought you meant the usual SOIC-8. Anyway, there are many great uses. :+1: :slight_smile:
Though I think there are uses for 20+ mm variants with e-switch too. :slight_smile: