Overheat a driver to the point it erases the firmware?

Flashlight Modding and DIY Parts
1

I was using my home-brew MT-G2 build at home, and had it on max for a few minutes (forgot to set the turbo timer, DUH!). It pulls about 10.6A from two 25Rs, and gets HOT in a hurry.

I turned it off as soon as it started to flicker, and let it cool. It was hot enough that it was starting to smell hot, and I’d have to guess it was over 100C on the head. It drained the batteries from full to 3.79V in that time. One it was cooled, I put the batteries back in, and it appears to be a single-mode driver now…. Now it pulls around 2.5A with the same batteries

I can’t see anything on the driver that looks like it got cooked, and the leads didn’t de-solder. Is it possible for heat to wipe the firmware?

Here’s the driver (fullsize link: http://i.imgur.com/s2X2IoN.jpg):

2

This surface is altered, probably too much heat on the junctions

3

I thought that was caused by my soldering iron when I initially swapped the wires for 18awg wires (it worked at that point), but it may be the culprit. I’ll see if I can swap out that chip for one I have on another driver in the next couple of days.

4

It should at least survive soldering temperatures, around 230°C. People (and manufacturers) sometimes program chips before reflow soldering.

5

LED chip damaged?

6

I’m not familiar with this driver or it’s firmware. Did it have a mode with about 2.5A and had mode memory? If so, possibly fried OTC cap, or bad connection with OTC cap. If the OTC doesn’t get charged or for some other reason can’t be measured, the light will always think it was turned on from being off for a longer time, even if you just tapped it for mode change.

Check all leads and connections where the OTC cap should be connected, including the MCU pin it’s connected to.

7

I just replaced the driver yesterday, and unfortunately my MT-G2 is still only pulling ~1.9A (on max) from two 25Rs measured with my shunt resistor. Did I cook the emitter as well?

At least I have modes now! :laughing:

8

Try measuring amps with direct drive (no driver). If it’s still low then it looks like you cooked it.

9

When it looks like this you know it got to hot.

10

That’s a great idea, I should have thought of that! Would it potentially hurt the driver if I hooked leads across the MCPCB while the driver is still soldered on? (total electricity noob here)

Yep, that looks like it got hot!

I take it your username is in reference to the Mazda rotaries? I’m a pretty big Mazda guy myself, but I haven’t had any rotaries yet.

11

A quick way to do a comparison, turn it on while measuring through your shunt, then jumper from LED- to ground (the edge of the pill, or a longer wire down to battery negative, depends on how far it's disassembled and how you're testing). The jumper just bypasses all the driver parts. Watch the amp readout, apply jumper, remove jumper, apply jumper, you can see how much it changes and how much the driver is reducing the current.

12

Okay, thanks, that makes sense. So many tricks of the trade to learn! :slight_smile:

13

Nope, I’m more of a Chevy, Ford, Mopar muscle car guy. My username is just something I had been working on before I setup this account. A rotorhead on a 64.