SOLVED (see post #31) - Help me diagnose a dead, stripped NANJG driver?

Flashlight Modding and DIY Parts
1

Hi,

I have what was originally 8x7135 driver with an ATTINY13A MCU on it.

I wanted to make it into a FET driver, so I stripped all the 7135s off of it, so then there were very few components left on it, basically the MCU, the 2 resistors next to it and the SS14(?) diode. And the small cap on the battery side.

I burned minidrv firmware with about 6 modes to it (this is working on another FET-converted driver I have).

So then I added an AOD510 FET soldered “upside down” with the FET Gate going to the Vdd pin on one of the
7135 pad locations, and the other leg of the FET going to the tab pad location of the 7135.

Finally, I soldered the emitter - wire to the tab of the FET and the emitter + wire to the location next to that SS14 diode.

However, when I test now, it looks like it only has high mode.

I’ve ohmed around and I don’t see any shorts. Also, visually expected everything and I don’t see any unexpected shorts.

What else could be the problem?

Also, remember that I burned the firmware to the MCU. Is it possible that the burn goes ok, but while the MCU is not functioning correctly? I had kind of assumed that since the burn worked without error that that implied that the MCU was good.

Any ideas would be appreciated!

Jim

2

He’s dead Jim.

Couldn’t resist.

3

“Remember”.

4

Any chance of a clear pic? It sounds like you’re doing it right but what we do vs what we think we do sometimes don’t jive.

5

I’ll try to get some. It’s a kind of mess though, because I think this was a board that I used awhile ago to practice stacking 7135s.

EDIT: Some pics (sorry for bad pics):

Spring side:

Emitter side:

EDIT 2: Better pics (I hope):

6

Output pin on Q1 is tied to gnd. This bypasses the FET and is direct drive.

7

See if you can solder wick that bit off. It’s a good idea to wick excess solder from the empty pads before adding new parts as well.

8

RBD,

I was too slow - I was just marking up the pic - I just noticed that when I reviewing the larger pic!

I’ll try that in a bit:

THANKS!!

Man, you all got good eyes :)!!

9

RBD et al,

That was not it. I cleaned off that bridge, but still get only high mode as soon as I power on the driver :(…

10

Does anyone know if it matters if the red emitter wire is shorted to one end of that small diode (S4)?

I don’t think it is shorted, but it looks awfully close, but I wanted to fine out if it mattered if it was shorted or now?

Thanks,
Jim

11

I see you’ve installed a larger than stock spring with the small end down, make sure there are no bridges on that side between B+ and any of the gate pin pads. That would pull the gate high and leave it stuck in turbo. Also check the center FET pin and make sure it’s not bridging to the gate pad. I use a bit of rubbing alchohol and a toothbrush to clean old flux off making it easier to inspect my work and get good pics(I just use a cell phone too).

12

No bridges between the spring and any of the 7135 pins. I just cleaned the flux off and checked. The center FET pin is not touching anything.

Also, I checked some other NANJG and that place where the red emitter wire is always shorted to the one end of the S4 diode, so I don’t think that’s the problem.

13

No, the led+ pad is connected to that diode. Worse case is you bypass the diode and lose reverse polarity protection but only if the short was to the other end of the diode. Barring any other shorts it would appear McCoy was right “he’s dead, Jim”. Ohmmeter should read in the megohms across between gate and B+, and fet tab and ground. If the short was gate-gnd the FET would be tied low and no light at all.

14

Maybe this is not it but, I have to erase mcu first and only then I can burn new firmware or in most cases it will not work properly.

15

Try the erase reburn thing. It’s beyond my knowledge.

16

Hi,

I tried to do the erase and then re-burn but still it’s stuck on high.

I put a scope on the FET Gate and it’s truly stuck at about 2.x volts as soon as I apply power to the driver.

So, I ohmed from MCU pin 6 to FET Gate ==> good

Then I ohmed from spring (Batt+) to MCU pin 6 and it was open.

So, my conclusion is that, for whatever reason, the MCU pin 6 must be sticking at +2.x volts when power is applied to the MCU. In other words it looks like the MCU pin 6 is stuck high (possibly the MCU chip got damaged?).

17

Just to be certain that pin 6 is really damaged you could remove the chip, put it in a soic chip then power it and test pin 6.

18

Clip #6, reburn using pin 5 instead and jumper to that pin. Don’no, throwing darts in the dark.

19

Maybe you already addressed this. It appears that the Ground pad that is below the "Q1" mark is shorted to the PWM pad.

20

Hi,

I can’t see images where I am now, but is it the same problem that was pointed out in post #5? If so, I’ve already eliminated that short.