What did you mod today?

More boost driver shenanigans today. I ordered one of the OG convoy T2 drivers a while ago specifically to see if I could put a attiny in place of the stock mcu. The stock mcu runs something like 700hz pwm with mode memory. I’m not particularly sensitive to pwm but I refuse to own a light with mode memory.

This is a link to the driver.

There is a good thread with a diagram of the driver here. Inside Convoy T2

Either my driver was different or there was an error in the diagram in that thread. The diagram shows pins 1 and 3 as being common. On my driver pins 1 and 4 were common. 3 was unused.

So on the factory mcu pin 1 was VCC and was common with pin 4. Pin 8 was ground. Pin 5 was PWM.

If you rotate an attiny13a 180 degrees from the factory mcu orientation then VCC and ground match. When I removed the factory mcu I discovered a via hidden under the mcu that leads to the fet on the other side of the board. I scraped off the solder mask and ran a jumper from the pin 6 pad to the via. I flashed a 13a with the modified star firmware I used on the harleyquin boost driver I built recently. The driver is currently running in a kaidomain aa host(convoy t2) with 3 modes and no memory and it’s powering a nichia 319.

Compared to the harleyquin driver the low mode seems to be lower and high seems to be significantly higher. This boost controller shuts off at 0.9 volts just like the controller used on the HQ drivers. It also seems to play nicely with 14500 cells.

I attempted to modify bistro to work with this driver but after hacking in several different offtime capacitor values I gave up. I couldn’t get short/long presses to work. Short presses always registered as long.

I have a picture of the via location with jumper, but if you have a weak stomach don’t look. I managed to rip the trace off the top of the board on my first jumper attempt and had to dig in to the board to reconnect to it. In my testing I also managed to rip unused mcu pads off the board. I also lost a leg on pin 1 on the attiny. I was on struggle street today.

Nice mod Bourbon Guy. :slight_smile:

Good recovery . I recently damaged one but don’t yet possess the skill / knowledge you have to fix it . Bravo !

Nice mod! An attiny412 would fit perfectly with that pin layout (which is what the PIC12 chips also use). But attiny412 is a new 1-Series chip with a different programming style.

Bourbon Guy… Thanks!! :+1:

I saw that but I don’t have the skills necessary to configure old firmware for it, and I didn’t see any examples for clicky firmware.

It won’t help at this point, but I’ll try to get some posted this week. I just flashed a clicky driver with a customized Star Offtime.

Alright last one. I don’t want to clutter up the thread with basic stuff like this.

Yesterday’s successful result got me thinking about mcu swapping the stock Tool AA 2.0 driver so I could have a lower low mode and no mode memory while still keeping 14500 compatibility. That ticks all of my boxes and I prefer the Tool AA to the convoy T2.

I really don’t understand how this driver works but I identified the VCC, ground, and PWM pins just like the convoy driver. Pin 1, the question mark on the top left, is coming from the small fet to the left of it. No idea what it does and I didn’t use it. The question mark pin on the bottom right runs through the 100k ohm resistors below the mcu and connects to the leg of the main fet. Also no idea what it’s used for.

The mcu pad spacing is tight and I didn’t really have wire small enough so I only connected to the pads for VCC. There are 2 common pads next to each other so it’s not as difficult. For ground there is a convenient spot marked G on the other end of the board. For PWM I connected directly to the fet. I used a attiny25 with minidrv 3 modes no memory for this one, and I swapped the stocki xpl for a nichia 219B. I measured a NIMH immediately after the light shut off on high and the cell was just over 1 volt.

I really enjoyed reading about your mods, please don’t stop posting. :+1:

Not clutter at all. AA Boost drivers are really lacking so it’s great to see what people can come up with.

Here’s a couple things I threw together for using the attiny412 with a clicky driver. Honestly, I feel like the 1-Series is a lot easier to code for - the register names are more like English names instead of 4-5 character abbreviations.

  • Very basic testing firmware that shows setting up PWM and using the ADC to check an off-time capacitor (OTC). Link
  • Cobbled-together (but tested and working) firmware with bits off Star (JohnnyC), RampingIOS (Tom E), and some custom stuff (myself) that can be setup with or without memory. It has voltage monitoring and a basic temperature control algorithm. Link

Note… I had to use an OTC instead of our neat “noinit” trick because the 1-Series memory doesn’t seem to decay for upwards of 45 minutes. That’s way too long to be useful for “short press” detection.

Awesome. I’ll be playing around with this soon. Thank you very much.

Anduril2 on a Sofirn SP10S? Sure, why not! Made a PIC12 to AttinyX16 adapter board.

Note: for the voltage divider, I accidentally picked a pin on ADC1 instead of ADC0. Rather than hacking Anduril to switch between the two ADC’s (temperature sensor is on ADC0), I just ran an airwire. Not great, but it works.

I currently have a klarus mi7 ti pulled apart and I’ve been pondering how I’m going to proceed. I’d love some details.

Did these lights have a voltage dividing network for LVP stock and you just utilized that?

Does the boost circuit run completely independently from the tiny? If it’s always on have you measured quiescent current?

Any idea what boost controller is used?

Seriously epic stuff. Well done.

gchart, That awesome!

Thanks guys!

@Bourbon Guy: I can get you more details later, but it already had a voltage divider network. I think it’s a 2-stage boost. A small one is dedicated to powering the MCU and a second larger one for driving the LED. I haven’t measured the quiescent current yet but I will when I get the final config settings dialed in.

Great work gchart. I know you have spent a lot of time on this and thanks for sharing.
So is this the first modified diver to work with a AA and Anduril?

To my knowledge, yes. I did have to add in extra code to handle the dual voltages.

.
You guys are the next level of flashlight modders. :open_mouth: Amazing accomplishment. :+1:
There are guys here who can follow in your footsteps, so you are raising the level of flashlight options and fun.
I love watching your modifications ‘gchart’ , Thank You for sharing your hard work :+1:
.

gchart - this is remarkable! If you have time, I think we need to go over this in detail in the BLF SP10S thread.