Thanks Dale - I've been buzzing out everything, just can't find a problem. According to checking against the Nanjg, it all looks perfect. Hhmm - gotta double check the resistor values.... Maybe I screwed up, or ordered the wrong size... I'll check now. thanks!! Sometimes I look for the hard things, and it's the obvious things I miss 
.
I haven’t built a BLF17DD. You should be able to measure those resistors in-circuit to determine whether they are out of spec due to damage or being faulty. That seems like the most likely culprit to me. I guess too much vdrop over the diode could also be an issue. Diode test on your DMM should show ~0.25 on that. I know you can measure that in circuit on a Nanjg 105c, not sure about the BLF17DD.
I just woke up so my math is broken, but I’m almost 100% positive that the ATTiny’s VREF tolerance (+/-10%) cannot account for this problem.
To build the 20mm ‘Zener Mod’ Linear LED Driver as a regular single cell 7135 based driver, all I need to do is jumper R3 and leave D1 unpopulated… right?
Think so wight, but you’ll probably want better confirmation than mine before you proceed.
Are you planning on using a regular nanjg protection diode on this? If not then yes just jumper R3 but you could use a regular diode over R3 for polarity protection if you wanted.
Oh yeah, thanks Cereal_killer. I forgot about reverse polarity protection.
Is it okay to run them without the protection diode in a regular 4.2v single cell setup if we dont mind not having polarity protection?
I got the Diodes (BLF17DD part D1) from DigiKey, as in the shipping cart list. I don't see any white or line mark on these at all. The Nanjg ones are labeled S4 with the white mark on the 'S' side, and the DigiKey ones are labeled 41. Any idea how to place these?
Tom the correct values for the voltage ladder are R1-19100 ohm and R2-4700 ohm. The other values are for the gate on the FET and only the 130 is usually needed.
You might have I check the data sheet from the part documentation. It should orient the diode according to the printing somehow.
Ok - I'm asking about the Diode? The 10V 0.57A Diode? The ones I got from DigiKey are unmarked - cant' tell which end is the cathode. I'll try swapping it around I guess. Looks like I'm getting no power thru it to the MCU. Voltage read near nothing.
I don’t see why not. 4.2-4.35v is perfectly safe for the ATtiny13A, so you don’t need the 0.25v voltage drop from the diode. It seems that the only thing it’s protecting is the MCU. If you have physical reverse voltage protection then you don’t even have to think about it.
Check them with the diode test (beep) mode on your DMM. They only work one way, so it’s easy to figure out by comparing the two diodes. I’d also assume that the the text is rightside up on both diodes when they are facing the same direction, but I could be wrong about that! As I said, easy to find out using the DMM.
Edited, figured out how to do it without piggy backing, this is the bottom of a BLF15DD V3 I’ve modded to work in a piston drive Nitecore D10, the ground ring has the traces to it cut so it can be the contact for the piston, a wire runs threw the drill hole from that ring to MCU pin 3, the top ground ring is still ground as is the 3 sections with via’s on the bottom.
Ohh, missed the datasheet link on DigiKey. Been looking around at generic sources. No idea how to do a "diode test" on a DMM. Looks like my best guess on the cathode following the lettering was wrong - I have to flip it around. This is my last effort to get this board working - resistors I measured on the board were dead on - the diode was the only thing left I could try.
Anode left and cathode right of the lettering. I had to look it up myself just yesterday. But my memory, you know…the 1 went towards the ground ring, the 4 towards the battery positive. That’s what I remember. Anode, cathode, geode, whatever.
Cathode is the “line side” so print right side up, cathode(line) is on the right. The only + trace for the tiny 10 goes from the through via to D1 to Vcc. The via is a direct path from B+ To led+.
Those diodes are marked silly. The only way I can remember it is that the "1" in "41" is on the side where the line usually goes ;)
Ok - the replanted diode from a Nanjg was the culprit. Replaced it (properly) and low voltage monitoring is working. I must have had the "41" in backwards... Ooops! I think it is consistent with the text (cathode to the right).
Wondering - used a Samsung 15M cell to test low voltage. It first triggered at about 3.0v, but if you measure the cell when not under load, it's about 3.2v. So under load, the voltage is lower... Is it fair to use the under load voltage then, or should the voltage level with the LED off be used? Of course it's impossible to test with the LED off, but maybe a fudge factor should be subtracted to account for the loaded drop?
Always measure cell voltage under load. I would elaborate on this but I can't quite think of a good example.