[WIP] 15mm PAM2803 w/ ATtiny13A rough layout / possibility

Honestly, it’s nice to know there is still interest… keeps me going on.
If it were only for myself, I would have wrapped it up with what I already have months ago.

Could you state what is “good” in your eyes?

Im keeping a close eye on this too.

I would be happy nanjg 102 with custom mode group selections a la guppydrv or blf a6 driver.

Still interested here too, watching this thread. :slight_smile:

I’m all for those small single AA lights that all need a 15mm boost driver. I’m looking for efficiency and good output on an eneloop/alkaline if possible, but even more for a more invisible PWM and a less medieval UI. My best example of a nice and good output AA boost driver is the Olight S15 driver (but I would like clicky operation instead of e-switch), but that is probably too good to be true.

For me. I’m looking for 1.5 volts usage, 14500 is not a big deal to me.
3 or more levels with a great moonlight mode (lower the better. 0.1 lumens? Less?). If super low moonlight need 4 levels to keep the spacing reasonable.

Option for no memory or hybrid memory. Off time, short latch time.

Lvp not a big deal for me since 14500 is optional. Max mode, 300mA or more? I’m not looking for a barn burner.

Programmable is nice. Invisible pwm.

15x4mm size? I want to put it in a Tain Aura titanium AA.

Led- same signal?
Wouldn’t that just bypass the FET entirely?
It looks like on this one the FET drain is going to led- and the FET source is going to boost driver led~~. Part of my confusion stems from the use of 3 different boost sample(Fastech, nanjg110, and the 15mm 2803 driver I’m messing with). Not a knock since it forces me to actually see what I’m looking at but sometime feels like a Gordian knot on 3 pages. The example above looks like it has the Fet drain to mcpcb led~~ and Fet source to boost led- so not same signal?

I think an updated op would help including schematic, latest oshpark board(s), and maybe an example for those that might be just stacking a controller board(simple pin/pad connection list).

Is this ready to use? The driver I bought from KD has next mode memory and the one fro Ebay looks identical.

HQ has a non LVP version ready and listed in post 84. Not much testing on it yet though.

Can I get the controller already programmed?

I have a couple of copper A1 on their way to me, it would be great to use this driver in them to get rid of the strobe in the original driver. Hope this will be available soon, thank you HarleyQuin.

I’m ordering some boards now but have no idea how easily this can be part swapped from a donor board. Probably a bit harder than a nanjg 105C. If you want to try it both the parts list and oshpark links are done. I believe it is usable with a donor mcu but not sure and to have off time switching you’d need a donor with that programmed in.

Ordered 1/2 dozen 2oz, .8 mm boards. We’ll see if they can be done by hand.

Looking on with interesr RBD. If anyone can it wiil be your good self.

Hmm, perhaps we were talking past each other.
You asked about the drawn schematic, this is the schematic of the Fasttech boost driver.
L- represents LED- of this driver, which is where the black cable to the LED is connected (the drain pin of the FET). So yes, on this driver it represents the electrical connection of driver LED- and mcpcb LED-.

L- (on this driver) is not the direct output of the boost circuit. That is pin3 of the PAM2803, which would be LED- on a driver without FET, like the Nanjg110.

On the FT driver the FET is sitting between pin3 of the PAM2803 and the LED.

Glad to have you on board. :laughing:

[EDIT:]
I opened a 15mm/17mm boost driver thread and will continue to fill it. There are still some parts missing, but it’s a start.

I figured it out but was temporarily confused(happens all too often). I was comparing the drawn circuit to the two boards wired together rather than the single modded Fastech board. Now to wait 3 weeks for the oshpark boards, plenty of time to re-read the thread a few more times.

Are hand wound toroids used to preserve board space at the expense of volume since they only need small wire pads instead of SMD inductor footprints?

RBD, just in case you ordered HQB17C v1, I just saw that I swapped the labeling of R2 and R3 (the FT-driver resistors of unevaluated purpose). Sorry… just double check before you build.

Inductors: I have not more than a small clue. But for efficiency (e.g. lower resistance R) they seem to be better when

- large or

  • capsuled
    Best would be large and capsuled. I believe these large open inductors are cheaper to manufacture and they increase size to compensate for the lack of ferrite (or whatever material around them). These fully capsuled super expensive Coilcraft inductors have the best “less resistance per given size”-ratio.

Pretty sure I ordered V2. The question on inductors was unrelated. Just curiosity. I thought that the toroid shape closed the ends of the magnetic field left open on a cylindrical coil enhancing the fields properties but even there it was more about the board space savings of a toroid over the SMD inductors. The smaller the board gets the more critical this becomes. 13mm AA boost drivers utilize both sides and the only smaller one than that had stacked boards around a toroid.

I’ll be checking the smaller part component placement several times while waiting on the boards and additional caps/coils as well as following your new thread for any updates.

Definitely V2 and I’m on the May6 panel. Hopefully by the end of the following week I’ll have them. Just for kicks I ordered some of the Ax2002 buck drivers from Fastech as well as some of the Pam2803 boards. Best price for these though I hope they don’t all have the broken inductors as in the picture.

a small q (sorry if OT), what IC is similar enough to adapt here that would boost 3.7 to 7.5V? want 6-7.5Vout at the same current as these ~1A

Looked but didn’t locate any 23-6 boost IC’s with 3V input. Some 8 pin IC’s maybe but don’t really know which of the various types to spec. My V2 boards went to fab today. :slight_smile:

8 pins is fine… just trying to get between a yugo and a viper. >80% driver efficiency would be nice 1A arena