Oshpark Projects

I can try that Scott, will do today, but it won’t work in my TP. Pretty much have to have a contact only on the battery side of the board. The body is machined with near zero tolerance, if the contact to the battery get’s any longer from the board mount position, that will keep the battery tube from screwing in all the way and possibly cause a breach around the water resistance the o-ring provides.

I’ll build the Tiny 10 this way and check it out. Then if I can duplicate it, I’ll see if I can rebuild the PICcolo board for the TP. (fingers crossed, which is gonna make it difficult to do this board. :stuck_out_tongue: )

I defer to your experience here but it seems the lowest reliable value is in order. 100-130 is a place to start and it can be changed if needed. 10K on the other, correct?

Yes, 10k on the pulldown. That's low enough to provide a drain, but too high to affect the signal delivered from the MCU.

Instead of having to buy separate 10K resistors, you can use the 4.7K value used in the battery monitor circuit.

The picture I’m getting is to put that 100 ohm resister under the FET, jumping between the outer ground pad and the smaller inner ground pad. ?

Would the 19.1 work too or is that likely too high to be effective? I haven’t tested anything over the 10k I used, other than to see that it makes a difference when it’s not there.

Any ideas about what's different between the big PCB & little PCB that would require one to have the resistors and the other not?

Close, between the control pad and the center pad for the 100, then leave 2mm of ground trace for the FET and cut the ground trace. The 10k goes between the FET control and ground pins so it could go under the FET but doesn’t need to. The drain will need a jumper. As small as the FET is, there’s real estate for all 3 parts without stacking. Well just need to ask Matt to set it up. Hey Matt! :bigsmile:

Surely that tiny little FET takes less to drive than the big ones, does it need the gate & pulldown resistors?

No middle modes without the 100ohm on any cell. I don’t have the pull down 10k so can’t tell you. With 2 fets it seems fluky with higher drain cells(low modes jump up) and with 1 behavior is erratic(sometimes fine then a minute later not). Somewhere north of 1.5A is where the erratic behavior begins as it was consistently well mannered below that point. It seems to handle more than 2A well enough so maybe the 10k will fix the middle modes at the higher current? I would give much credence to the numbers posted in an absolute sense as there were a number of extra clips, cheesy meter wires, and solder connections involved. Reducing wire lengths and connections could easily change lots of values and operational details. This was just a run through to see what to try next using ready to hand parts. I’ll also need to play with the different stars as #4 was all off. I think a different chip might be in order for the blf15 as some of the cells I used are easily capable of topping 3A and none did more than 2.1A but I should try it again with better shorter wires before that choice is made. Some good testing done but plenty more to do.

It’s quite a bit brighter without the meter in the circuit so when Dale gets his hooked up maybe he can get some decent numbers. Anyone know what’s different about the firmware for star 4?

I used a new ATiny13A, flashed it with my FET hex file, all new components except for the diode, which I took off a Qlite board since it’s so small and the new ones I have are TRex’s by comparison.

I used a solder mask to re-flow the MCU side of the board, it’s gorgeous, very neat and everything appears to be sitting just exactly where it fits best. Those 2 resistors are VERY close to the MCU pins but there just ain’t no room on this thin BLF Tiny board.

I did some of the best soldering I’ve ever done on the backside, did not put a resistor in the circuit as it’s such a small FET and the current will be so low I didn’t think it’d need it. Beautiful piece of work, I’m so proud! :slight_smile:

But it doesn’t work.


Time for something you are good at. Pictures please. As you say, it’s tight on that board, especially on the mcu side. If you have the BLF 15 you could use the stencil for just the mcu and hand solder the rest. I wasn’t going to downsize until I had an upsize version working.

Edit- I’m not good at soldering, I’m just barely good enough.

I might have the diode inverted, would that do it?

Sure looked like I was seeing the line to the left, but it appears here (blown up quite a bit) to be on the right. I will go pull that and reverse it, see how that goes.

I’m pretty good at getting things wrong these days! lol

Yep, diode is backwards. On the S4-marked parts the line goes to the output, the other side without the line goes to B+.

With that set up you should get something other than a buzz from smoke inhalation. It looks good but well have to take your word on the led connections. Can’t see anything there.

Sssh, we could gave kept him at it for hours.

Bad news is, I had the diode reversed. Good news?

I have 4 levels, .02, .29, 1.47 and 3.02 on an Efest IMR10440 sitting rested at 4.12V. :slight_smile:

Oh, and I can barely see at the moment…

Awesome relfow job…very smooth!

Is that the current pull (.02, .29, 1.47 and 3.02) thru that Tiny10?

Is 3A safe for a 10440? Whoah…

Awesome. Well done. Care to do a failure analysis working your way up the current ladder?