Polyboost driver

I was having a hard time searching high Vf boost drivers in a compact circular size that would fit in a flashlight body.

At some point I was convinced that I would not find it ready to buy at the size I wanted, and after some tests with the TPS61500 boost controller I finally decided to learn a new skill and design my first real PCB.

This is how Polyboost very slowly came to life over the last months!


Why Polyboost?
- it’s a 7135-less “Nanjg 105c” driver
- that instead of linear, it is a boost driver
- for lots of Vf (OVP at 25v, tested up to ~20v)
- fed by multiple LiPo cells
- with multiple analog dimming modes/levels (Open source FW’s compatible)
- built for multiple projects
- all of them with multiple laser chips in series
- and multipurpose: it’s also a keychain!
- and last but not least: contains a laser shooting bunny, that is also a duck

Very poly characteristics!


What it is not:
- not cool, this thing gets hot!
- not flawless, this is the first PCB design I sent for manufacturing and of course I made mistakes
- not protected against reverse polarity, I messed it up on the design. (Using a schottky for all the current was a simplistic dumb ideia)
- not my original circuit schematics, this project was heavily based on Giannis open source boost driver, TI reference for TPS61500 boost controller and Nanjg 105c schematics
- not a comercial product, I’m just a hobbyist filling all my free time with an obsession


Main features:
- 22mm of diameter
- up to 4 Lipo cells (18v input)
- 2s slow start
- OVP set at 25v
- Multiple modes: 1% to 100% analog dimming
- 1-mode jumper for no thinking, only full blast (MCU and LDO delete)
- Voltage monitoring and over discharge protection (Nanjg 105c FWs, needs appropriate voltage divider for the number of LiPo’s)
- Used FW: Biscuit.c from trebisky/Convoy repo on GitHub


- Tested configs:
- Runs fine at 1A with ~7.4v input and ~17v output (~80% efficiency)
- Tested briefly at ~2A with 16.8V input and ~18V output


It works!

I’m very satisfied with the results, analog dimming works great up to ~1%!
It can shine the Qualas very gently

The extra planning on decoupling the “smart” and “power” sides was totally worth as I was able to test both things separately, always having the option of removing the MCU completely with a single jumper if it didn’t work.

Now it’s time to finish the first build using it!

Great job! Can't wait to see V2! You attaching a heatsink to the chip as well?

Ricker said:

Great job! Can't wait to see V2! You attaching a heatsink to the chip as well?

Click to expand...

No heatsink! Why I’m not doing it??

I’m not confident that this will make it handle the full 3A @20v of the Qualas, but it can help!

I’ll be very happy if I get 2.5A at least for a couple of seconds with the extra thermal mass!

For QuaLas arrays specially the dual stack ones I recommend physically splitting the arrays and powering each with their own driver at lower fV but is not an easy task to split the arrays.

Your first driver? Congratulations, the mini array is also nice. How powerful is it? There are different models, these replace the old brass banks with, say, 8 or 10 diodes with Gball as a collimating lens! @Paulostool