Is that a FET on the non-inductor side? Can't he just bridge the 4.xx volts to its gate directly?
This is a boost driver… so no. If it’s not the output voltage then hooking it up to Gate will stop the boost circuit. If it is the output voltage then… well that’s even more Doctor Seuss.
I knew that. I was just testing you. You passed. I think. 
Was hoping that if it was a bad idea (like it apparently is), it might trigger an idea on a different approach. Please resume your regular programming.
Ok, sorry. I had made two mistakes again due to rushing the readings first time, not a great day for me in terms of reliability… Firstly I don’t think I actually gave off ‘clockwise’ readings, but read it from top left then bottom left, below are genuine clockwise readings. Secondly, the unmarked component needed a bit more pressure on some of the legs to show readings, so I have a couple of extra non-zeros. Below are the ‘real’ results (assuming I haven’t made anymore mistakes).
Battery contact side
1.4, 1.4, 4.9, 0, 0, 1.4
Inductor side (inductor to right)
0.1, 0, 0, 0, 4.9, 4.9
Led voltage reads as 1.24v, which surely cannot be right for an XP-G. But my DMM does give reliable readings otherwise, maybe the PWM is throwing it off a bit.
Some more interesting readings from the inductor side component bridged to the LED negative output from the driver are shown below. I’m not sure if they’ll help, but they seemed interesting to me. The 0.1V deficit of the first pin compared to the next 3 matches the 0.1v reading it gives bridged to negative driver input, and the 1.25s match the readings from the LED, I don’t know what that means though.
2.6, 2.7, 2.7, 2.7, 1.25, 1.25
Frankly I have absolutely no idea what you are describing in your last paragraph. I’m chewing on the rest.
I’m really frustrated. It’s important to not mix stuff like that up. My model of how the driver worked and where invisible connections existed is just blown out of the water now.
If you get bored, please do the same readings on medium.
I’m definitely out of steam on this one for now.
If necessary maybe you can drop the LED into a glass of distilled water or something for cooling while you do the tests.
The last paragraph was measurements taken from the chip on the inductor side with the negative lead on the driver end of the wire connecting to the LED negative, I’m not sure if they were relevant though.
I’m sorry to mess you about with mistakes, and do appreciate your help. I’ve just got readings on medium, but having done a few tests three seperate times I got slightly different results on some pins.
So, clockwise for the chip on the battery positive side I got all three times:
0.8, 2.8, 0.8, 0.05, 0.05, 0.65
On the chip on the inductor side (clockwise starting left to right with inductor on right) I got on two of the three tests:
2.75, 0.05, 0.05, 0.05, 2.77, 2.77
On the one anomalous test I got these for the inductor side chip:
1.4, 0.05, 0.05, 0.05, 3.78, 3.78
I’m not sure why the 2nd test was different, it’s possible the light bumped up to high mode as I flipped the assmbly over, but I didn’t notice at the time.
It was certainly in medium the third set of tests though and they matched the first results so I would go with those.
Just to clarify for absolute certainty, by clockwise I mean: 1. top left, 2, top middle, 3. top right, 4. bottom right, 5. bottom middle, 6. bottom left
Are there any more thoughts from anyone?
wight, I don’t want to rush you as it’s your own time you’ve been spending trying to work this out for me (which is greatly appreciated). But could you let me know if the new results on medium are more useful than the last ones?
The 0.8v pins match your description of a voltage between that of the battery and 0, both of those were on the B+ side chip.
Haven’t looked or thought about it since I stopped last night. Started looking again just now, but I’m still confused. Let’s just take a look at one chip, the one on the battery contact side.
Low: 1.4, 1.4, 4.9, 0, 0, 1.4
Med: 0.8, 2.8, 0.8, 0.05, 0.05, 0.65
Is that right? Pay close attention to the second and third pins. Third especially. From ~5v to <1v eh?
I think/worry/suspect that these pulsed signals have got your DMM so confused as to be less useful :-(.
No, I’m looking at it and it seems that you have 2 and 3 transposed. Is that the case? I can’t stress how critical it is not to make that sort of mistake…
EDIT: to be clear, I have no idea which results you transposed the numbers on. It just seems that one set must be transposed.
Ok, the cell is a little more run down than last time, but these are the measurements clockwise for the B+ side chip.
Low: 1.3, 1.3, 4.8, 0, 0, 1.3
Med: 0.9, 3.8, 0.9, 0.03, 0.02, 0.8
They look very similar to what I measured before to me, there is some slight variance each time.
Having just read your last comment, I have just checked again and the results are certainly correct. This is for starting top left the moving left to right across the top row, before right to left along the bottom, on the B+ side chip with the markings facing the right way up.
Weird. Well, that’s that. I’ll keep thinking about it for now, in the meantime maybe someone who is more familiar with single-cell boost circuit implementations will chime in.
email dqg and ask. you may get lucky.
Ok, here are measurements just retaken for the inductor side, clockwise again starting top left with the inductor to the right hand side of the chip.
1.4, 0.03, 0.03, 0.03, 3.8, 3.8
Oddly, this matches my ‘anomalous’ set of readings from yesterday for this chip…
… and after a break and re-measure I get exactly the same again.
Switching gears. Unfortunately I’m not sure what these measurements indicate. No slam dunk 
So new strategy, more annoying for everyone I’m sure. Let’s use the ohms mode on your meter and check a few things…
- Is LED- electrically the same as GND?
- Is LED+ electrically the same as the small gold pad between 2R2 and the cap?
Sorry, but you’ll have to give me another walkthrough.
Should the LED be on, and if so in what mode? What do I need to touch both wires to, and what range should I set the DMM to?
You don’t need the LED in the circuit. Set your meter to 2000 Ohms. Stick the probes on the points in question and insure that it reads zero. EDIT: or that it reads one. If it reads something else then that means something too but is probably less useful.
I started drawing a circuit diagram but it’s still got more stuff missing than present 
Should it still work with the LED in the circuit? I’ve just testing with it all still intact but with no cell.
LED- (attachment point of wire to circuit which goes to LED- input) to ground (driver negative contact) gave a reading of 091.
LED+ (attachment point of wire to circuit which goes to LED+ input) to the small gold pad between the 2R2 labelled inductor coil and the adjacent little brown box with silver ends (assume capacitor) gave a reading of 007. (To be honest I wasn’t sure where exactly you meant for this one so I made a best guess, please let me know if incorrect).
Having the LED in circuit is not a problem.
The FET / transistor is marked “2300”. Please use whatever methods are necessary to determine what the the thing is hooked up to. We can clearly see the connection to the pin next to the zeros. I would like to know where the middle pin and the pin next to the “2” connect.
- The pin next to the “2” appears to what we have referred to as pin1 of the unmarked SOT23-6
- The middle pin… I have no idea? Does it connect to LED-? Who knows…