My first driver builds. The good, bad and ugly.

Flashlight Modding and DIY Parts
1

A couple weeks ago I decided I wanted to build a driver…or maybe a bunch of them.
I read through numerous threads here to learn. I decided on some Fet+1 and Texas Avenger fet+1+x as my platforms.

I started ordering pcbs from oshpark. Resistors, caps, fets etc from digikey.
I added a few pcbs to my most recent order from MTN so I had something to tide me over while I wait for stuff from OSH.

Lastnight, my new microscope showed up, so I decided to do a hasty build on an MTN17dd v1.13 FET only board

Applying the solder paste proved to be much harder than anticipated without a stencil.
I’m using Kester ep256. There was no way it was going straight from the syringe onto the board. I squeezed a little onto a silicone mat and used a .3mm apllicator made for applying modeling glue. The paste was pretty grainy (maybe a little dry) and it wanted to stick to the applicator, but not the pads.
I did my best. It was a mess.

Once I placed all the parts on the board in a grotesque fahsion, I was pretty excited to put it on the hotplate.
In my haste, I forgot to put flux on the board. Oops. Once I realized my mistake, it felt like it was too late. I ended up using my c210 iron to hand solder everything under the microscope. I was skeptical that I didn’t mess something up. At this point, I was out of time and had to call it a night.

Fast forward to after work today, I successfully flashed blf-a6-13-rmm.hex. I was excited it flashed successfully.
I was concerned about bench testing it without the proper means to dissipate heat, and I was anxious to see if it worked. So I threw it into a convoy m8+ host.
I used an LHP531. Chucked in a samsung 30Q.
It actually worked!
All the modes work and it’s bright as hell.

Even though it started pretty ugly, I’m excited to do my next one.
I need to figure out a way to make applying solder paste easier/cleaner, without buying a bunch of stencils…yet.

Maybe I need to thin the paste a little with 99% ipa?
Should I flux the board before putting paste on?
Should I be pre tinning anything?

Even though it works, I’d like my future builds to go smoother.

Here are pics of the carnage.

20260331_183845
20260331_1838451152×1152 498 KB

And proof of life.

20260331_182632
20260331_182632984×1312 251 KB

10 Thanks
2

What’s the date on it?

1 Thank
3

1/2/26

1 Thank
4

Maybe you just need to squeeze some out to get to the fresh stuff.
It should have the consistency of thick toothpaste and shouldn’t look granular like that.

I wouldn’t recommend this.
I would recommend mixing a little flux with it instead.
However you would want to make sure the flux is compatible with the paste.
I try to use the same brand and series with my Solder, Solder Paste and Flux.

I only add flux if I’m reusing a component or pcb and it already has solder on it.

Nah

3 Thanks
5

I use an old computer CPU heat sink.

2 Thanks
6

Good info. Thanks!
I’ll squeeze a bit out into a small container and see if the “good stuff” is further up in the syringe.

I’ll mess around with the stuff I squeeze out and see how that goes as well. You know, for science. I do have compatible kester flux to try out some mixing.

But if the stuff further into the syringe seems good, I’ll use that first and foremost.

1 Thank
7

I definitely have a few of those around after building computers since the early 90’s.

1 Thank
8

I used a stencil a few times and didn’t have much luck.
It was for a ATTiny1616 with is 3mm x 3mm and has 21 solder connections.

1 Thank
9

Yikes. I’ll have to work my way up to those little guys.

1 Thank
10

Had the same issue with a fresh solder paste syringe, it was dry and crumbled at the beginning.

After some squeezing it got a lot more runny and sticky, I can easily stick a little bit on each pad directly from the needle.

3 Thanks
11

Yea. I ended up pushing a bit out and then it started flowing properly.
I’m doing a 20mm fet+1 as we speak and it’s going way more smoothly.

1 Thank
12

Just flowed the 20mm board with paste from further up the syringe. Came out way better.

20260331_225329
20260331_225329864×1536 474 KB

3 Thanks
13

So after playing around with the build for a while, moon mode stopped working.
I did a little digging through old threads and found that the 380ma 7135 chips could be an issue.
So I swapped it out with a 350ma and all the modes are now properly working!

The only issue I’m having is the length of time for medium and long presses.

It’s like 6 seconds to go backward.

Does that mean I need a different OTC Cap?

1 Thank
14

You’ll need to compile your own version of the A6 firmware after getting the OTC values using offtime-cap.hex.

NOTE: when the light is hot the OTC decay rate is faster. The hotter the light, the faster the decay.
So you’ll want a bit of a buffer.

Also if the voltage readings are off, you’ll need to run battcheck.hex as well.

1 Thank
15

Okay. I appreciate the help. It looks like I’ve got quite a bit more work. All of my attempts in microchip Studio have been a failure so far. None of the hex files I’ve created have worked on any of the drivers I’ve practiced flashing. I think I just got a lot more learning I need to do.

16

IMG_9149
IMG_91492318×1074 221 KB

1 Thank
17

When you create a new project are you choosing the MCU you’re working with?

18

Yeah. I’m starting a new project. Executable C. Choosing Attiny13a in the list.

19

“GCC C Executable Project” right?
And it compiles and it successfully flashes to the mcu?

Maybe the lfuse and hfuse settings are wrong?

1 Thank
20

Correct.

It compiles/builds and flashes properly.
I believe the fuse settings are correct.

I feel like I’m missing steps somewhere.

Whether its not selecting/uncommenting specific parameters properly.

Do I need to have other files in the project folder for microchip studio?
Or do I just need to work off of a single C file to build the hex?