Need help with Mateminco driver mod.

Flashlight Modding and DIY Parts
1

As the title says, I’ve run into what seemed like a straightforward mod, but things don’t always work out. Have this Astrolux FT03s Mateminco driver which had an issue being dimly lit, always ON. BG sent me a replacement so I set out to correct the problem one (replaced the de-bouncing cap – leakage). This worked out fine.
Want to use this original driver in a new build and this venture got deeper than my expectations.

I’m taking out the charging board and traced back the leads. I was hoping to have this running with the stock Anduril software and the indicator switch LEDs.

A picture for reference: (click on pic for Imgur)

Getting ahead of myself, a few pics of driver/charger assembly:

On my 1st attempt, I erroneously connected things and my load LED went flickering from low to high speed. The switch was not responsive. I concluded I blew the switch cap so replaced it with a spare eSwitch. Good; MCU responsive to the clicking.
I then did some better investigative analysis and connected the switch V+ to the Bat+, the green LED– to the pad as so marked. The MCU pin #7 is thru the 20kΩ to the adjacent pad, which would be the 2nd channel. The .03nF cap seems to be a filter to Vcc (pin #8).
Again the load LED went into a frenzy. I disconnected the switch, but the output remains in stasis (continuously ON, flickering). The Green switch LED gave a short burst and now has died.
Recreating my experiments in my head, I think the 20 kΩ resistor is too low. Would a 0.17mA current burst knock the LED out? (testing with 3.4 V+). Perhaps I inverted some switch leads – fuzzy to recollect.

––> My question; did I kill the second channel? If so, would this also send the primary channel into chaos?

I cross-checked all the peripheral resistors and caps, short of one (#3) I can’t get a reliable reading on my UT-160CT SMD tester.

I don’t think a factory reset (emulated) would work being the switch isn’t responsive. Can an MCU Reset revive the AtTiny, and if so, how to go about it. I just so downloaded the specifications PDF and it’s overwhelming for a newcomer. I have no experience in these programmable chips, let alone have any hands-on other than the direct mapping of pinouts.

2

Charging chip hollow?

3

Huh?

4

The first thing you did was connecting things up the wrong way? And you’re not sure how it was originally? And now you wonder that it doesn’t work? Seriously?

I think you made it much worse than at the beginning. Before you start changing anything, ask yourself what would cause the fault and what modifications would change it.

My guess would be a leaking 7135, maybe even the FET. Now it’s in an unpredictable state and everything can be faulty. Good luck.

5

You didn’t get the mod. The first driver was screwed up and I fixed it. I removed the charging board as it doesn’t fit in my new build. The replacement driver from BG is working fine and in the FT03s.

I went at it methodically and had the driver work with the eSwitch. I also had the switch led working, independently from the main LED (I’m using some dummy load before committing to reflowing a quad). It’s when I put them together that the driver went into a rapid flashing stance.

Addendum; I’m not trying to fix this driver but to have it work without the charging board. If I’ve taken the time to explicitly map out the pins and pads on the driver board, I had given it more thought than randomly plugging wires. I always assume the error be mine, but this driver board was defective. I presumed I may have erroneously connected the wires on my first attempt. Changing the switch got it working. When I then connected the switch LEDs it went crazy. The 2nd channel has a 20k resistor. Shouldn’t it be enough to limit the current to the MCU?

Nevertheless, your insight may be helpful.

6

Maybe you’ve created a sentient lifeform and it is trying to communicate with you.

1 Thank
7

Nah. As Sammy pointed out, faulty FET/7135. I triggered a fault.

1 Thank