[WIP] *15mm* DD+single-7135 driver - single sided Dual-PWM compatible using 3.3x3.3mm FET

I figured you’d like this one; I know your mind’s been on 15mm drivers for a bit now.

Thanks RBD. It’ll be nice to have this stuff in the OSH Park projects thread, but I suspect that it’s also time for me to put a thread together just to organize my drivers, outside of my signature. RE: the bypassed diode thing - good point. That description started off a little lacking and I’ve copied it between my driver threads without improving it.

BTW, I think I’ve asked you before, but where do you like to order Khatod optics? Cutter? Or maybe a source in the US?

Apparently I still haven’t figured out how to find specific parts on mouser. can you link for the correct FET?

I haven’t picked one out (and I will not be doing so immediately).

Please use these instructions to help find one. See what you come up with by focusing on these things, mostly in this order:

  • I’d stick to 20-30v rated stuff
  • high current rating
  • a fairly low gate threshold voltage since it will be powered by no more than whatever the MCU gets as input. Take a look at the graphs in the datasheet for PSMN3R0-30YLD which show Rds(on) vs either Vgs and Rds(on) vs Id. Remember that our Vgs will be lower than battery voltage… it’s battery voltage minus the voltage drop in the diode and in the MCU. I usually eyeball the performance around Vgs = 3.3v.
  • gate charge – don’t let it be too high. Look at the other selections we’ve been discussing, but I think you should be fine with <50nC or <100nC or something.

Cutter Electronics has them but both price and shipping are high.
Both are taller than Carclo quad optics and don’t fit the same peg radius as Carclo 25mm quad even though the led locations are the same. On some mcpcb’s all that’s needed is to open the peg holes to the edge of the board.

Thanks, that’s what I was afraid of (the cost in dollars, not the peg layout).

How about:


I’d call SiS414DN the clear winner for this application except for the fact that it’s in a flat-no-leads package. I expect that it will be very difficult to solder for most people. (Actually reflow only for most people I think…). It’s also “Package Limited” to 20A, but I don’t think that’s very relevant here. PSMN4R2-30MLD has more exposed leads and should be a bit easier to hand solder.

EDIT: Clearly I made a mistake. PSMN4R2-30MLD is not that great.
EDIT 2: Clearly I can’t read a datasheet. SiS414DN isn’t that great. I made an error with units (ohms vs milliohms).

I have not evaluated the following FETs beyond Rds(on). Listed in order of Rds(on) performance at our voltage levels:

Possibly functional, but not characterized well by the datasheet (for our purposes):

Not considered due to high Vgs threshold:

Ok I just added some PSMN2R4-30MLD to my order. hopefully they end up working for our purposes. Just to confirm, other than the FET and the 22k resistor, everything else is stock from a 105c, correct?

You also probably want a 1uF offtime cap such as C2012X7R1E105K125AB

I have some 80-C0805C105K4R in my cart, that will work right?

That’s fine, no problem.

Oh man..... Is life getting good or what ?

I was wondering when/if anyone would incorporate the smaller LFPAK size but resisted suggesting it because of the tight pin spacing(I barely get by on attiny’s and 7135’s). Curious to see how much better it might be than the 2502 or it’s successor(you listening Dale?).

He's asleep in your back yard. Yell a little louder and he may wake up.

I agree, the pin spacing is very tight. Fortunately only one pin actually needs to be separated from it’s neighbors. Still, if a person is a beginner at reflow soldering I’d say that this is not the place to start their journey. The same thing applies doubly to hand soldering.

PSMN2R4-30MLD vs AO3400? Single LED performance is what matters for this driver I think, and that should restrict us to looking at higher Vgs numbers vs triples & etc. At 3v Vgs it seems like we’d have around a >20milliohm difference in favor of the PSMN2R4-30MLD, which translates to maybe a 0.08v difference in voltage dropped over the FET. To me that looks like approximately half an amp. As you know, all this napkin math doesn’t get us very far. We’ll have to wait for them to be built and tested. (of course it should beat the 2502 by a wider margin)



Scratching my head.

You’re trying to make me rebuild the Texas Poker yet again, aren’t you?

I would order a stencil if I were to do this one, I’ve been finding the re-flow is so much neater with a stencil and Kester E256 paste.

Edit: I don’t have many (if any) 15mm driver lights. The TP is 10mm and I’m pretty happy with the way it’s performing. That said, I do have a little stainless steel light on the way that I believe uses a 15mm driver…

With stencils I’ve found it very helpful to have a piece of cardboard or something the same thickness as the board or mcpcb with a hole to drop the board into. This lets the stencil lay in place flat and is easier to keep aligned.

I use the acrylic L shaped guide from OSHStencils.

Picture? I’ll show you mine if you show me yours.

It’s just a plexiglass L the thickness of a board, I have white ones and black ones. 1 1/2” wide for easy overlap of the stencil to hold it by.

Does anyone build these? I don’t have the tools, time, or skill to build one.