In terms of drivers we can easily program I think it’s really just this one (DD) and the Nanjg AK47 (3*7135).
And all this time I’ve been using the rotary tool to cut out a recess in the retaining ring and making it work! 
Keep on like this and my new Proxxon is going to get lonely…
the past 2 days I’ve been shopping around for parts to make BU_CK smaller now that I’ve got it up and running well, I looked up this FET cause I likes the sounds of it (the spec’s mentions) thinking it would work as a replacement for the D2PAK I’m using but the RDS (on) is pretty high on it, quite a bit higher than the AOD510.
I dont think it’ll work for my application but I wonder why that doesnt seem to be causing issues in this single cell / DD application, I had assumed the resistance was one of the huge things to watch in this application. I wonder what a better spec’d SO-8 FET would do on these things.
Note I’m not complaining, just thinking out loud. When we were first picking out FET’s when this whole DD method came about one of the largest factors (right behind the current rating) was the RDS (on), why back then was there so much fuss trying to find FET’s with super low resistance if in the end it apparently doesnt matter? Just seems to me 9A to an XM-L is pretty damn good. Would a FET with lower resistance do even more than that?
edit: Alex would you send me the .lbr part for that FET?
Order 12 boards and parts will post when I get them
The Rds on has to be compared at the operating voltage we use and there’s a wide range being produced in this package size so it’s important to look for both a 4.5v spec and the graph that shows just where it goes north. I’ve seen values as low as 1-2 mohms which is plenty low enough but the gate voltage where that happens is crucial.
From what I saw comparing the graphs of the two they are virtually identical, both between 2-3 mohms at 10V, ~3 at 4.5V, and ~4 at 3.5V so I don’t see where it’s “much higher”
I’ll post the mouser cart in the op but keep us up dated on results in case there’s an alteration needed.
I think you must be misinterpreting some part of the datasheet. The RDS (on) for this is in the same class as other FETs we use. Just as a sanity check (one can never be too sane, after all…) I did check the respective datasheets just now. Here are the datasheets I referred to: AOD510 & PSMN3R0-30YLD. PSMN3R0-30YLD clearly edges out AOD510 in respect to RDS (on). As we approach 3v the difference becomes more pronounced, again in favor of PSMN3R0-30YLD.
I’ll post the part I use over in Mattaus’s Eagle Libraries thread.
I couldn’t access that AOD510 link but the one I did find wasn’t any better than the one given for the mouser part.
Sounds good to me.
That’s because AOD’s website is down ;).
I’-ve added both boards to the Oshpark op along with links to the mouser cart and back to this thread and placed them under a new “hybrid” heading. Edit -this is the DD only version, not the hybrid. Sorry
Looks pretty awesome. This looks a whole lot easier to fit into a Cypreus EDC than the BLF17DD I used. I especially like that it’s compatible with retaining rings, since I hate soldering in something I might want to reflash later. And it’s ready for zener and off-time cap and everything!
Any chance of adding a single 7135 chip somewhere, attached to a PWM-capable pin? Giving it “moonlight special” hardware would be icing on the cake. No more need to use tricky fast PWM=0 hacks to make a high-powered light work at low levels, not to mention longer runtime on modes under ~350mA.
Edit: For my purposes, additional 7135 chips wouldn’t matter. I just want the lowest modes to work better without sacrificing the ZOMG WTF BBQ blinding output offered by a FET. One 7135 could handle modes from 0.1 lm up to about 50 lm, and the FET could handle the rest.
Oops! Headed over to edit the Oshpark thread, this is the dd only version not the hybrid.
Thanks ToyKeeper.
I’ve just checked for you: bad news. Even once the Zener diode is removed a single 7135 just does not fit on top along with all the other components in a layout where an SOIC clip can be used. [The keepouts required for a clip actually use up a fair amount of space.] I didn’t try, but I think even with a switch to 0603 there would be trouble.
The good news is that I don’t really think it’s the end of the world. To me it looks like a 7135 could be easily stacked on top of the FET. I can easily bring the second PWM line out to a pad next to the OTC label. Even though you’d have 2-3 short/small airwires you’d maintain enough clearance to use an SOIC clip.
I suppose you could drop ground straight down to one of the three gnd pins of the FET. Which mcu output would you use for the 7135?
Pin 5. There are only 2 PWM output pins, that’s the one which is not already in use. It routes cleanly under the ATtiny and over to the area I mentioned. We could also potentially place a pad for it between the Zener and R2, or just do both.
It might help to place the 7135 upside down.
Haha, don’t kill yourself trying to cram another bulky chip onto such a streamlined and elegant driver.
I can wait for the 2-sided version. 
Just so we’re on the same page, and I’m sure we are… the 2-sided version already exists, we’ve talked about it: 17mm DD+7135 — linear regulated driver w/ FET turbo. If you find your need a little more pressing now having thought about the driver a bit more then just send me a PM.
I have no pressing needs, just imagining a bright future.
(and a dim future, gotta get those low modes too)
For reflow work I have been using cheap Chinese “Mechanic’s” solder paste, a propane torch and a little frying pan. I’m hoping Christmas brings a reflow station.
I’m still using a $20-$40 1500W heat gun with good results. (Milwaukee MHT3300)
Do you have a nice soldering iron?