I have buyed v24 of the driver layout at Oshpark. Now its time to order the comonents. I’am very new in flashlights so I will make sure that I order the right parts.
Are the size of the resistors proof? Is D1 a Schottky-Diode? I’am not sure?
Next Question: For Zener-Mod I have to place C1 at zener-spot and place the zener diode on top. Thats clear to me. But what about C1? I have to leave C1 empty? I have to change some resistors or other components for zener-mod, too?
If the version of the board you ordered is one of the early desings then yes. The old boards had a defect in the layout so in order to make them work the 10uf cap has to be placed on the zener diode pad with the zerner diode over the top. Make sure you get the diode placed on it in the correct direction. The end with the little stripe on it goes toward the inside of the board.
The other components that need to be changed for the zener mod are R1 and D1.
R1 will be replaced with a 36K ohm resistor and D1 will be replaced with a 200 ohm resistor. Notice that all resistors accept the D1 are in K ohm, and the D1 is just 200 ohm.
BTW, anyone know why it uses a 22k resistor instead of the 1912 we had been using before?
I’ve tried both, and I found that 1912 makes the OTC work quite a bit better. The 22k resistor puts the useful OTC values at the far edge of the attiny’s measurement range and makes the OTC timings particularly temperature-sensitive.
wight states in the OP: "My 22k R1 replaces the 19.1k R1 due to the change in how the divider hooks up to B+. 22k is not guaranteed to be perfect, but it’s close enough for me and for now".
It was supposed to compensate for the design change in order to keep the LVP values the same (130 and 120), but I usually use lower values anyway. wight's and C_K's board all had the design change - changed from the original Nanjg, and the original BLF DD driver (https://budgetlightforum.com/t/-/27210). On the Nanjg and original BLF DD, the Batt+ came after the diode, while on the new designs, it's direct from Batt+. The diode has a voltage drop across it, so maybe the theory was it would be more reliable without the diode? This goes back a while. I thought at the time, there was a good reason for the change.
No idea it effected OTC though -- if the 19.1K is better on this exact board design for OTC, I'm thinking it can be used, just that we would have to come up with different values to use for LVP, as long as it stays in range of course.
Ahhh - wight or C_K would know this stuff cold (both MIA), RMM would also probably know. There are a few others around knowledgeable on issues like this.
The only purpose for the 4.7K and 19.1/22K resistors is LVP, so these can be changed to anything as long as LVP can be made to work.
I need to get some 1912 resistors to swap out on my current drivers so I can test it, but from what I’ve seen so far the 22K resistor is a downgrade. Both allow LVP to fall into the useful measuring range, but 22K seems to mess up the OTC.
With a 1912 resistor, useful OTC values (0.5s and 1.5s, room temperature) are about 200 and 100, and after being left in a freezer the medium press takes 2s instead of 1s.
With a 22K resistor, useful OTC values (~0.75s and 1.5s, room temperature) are about 250 and 190, and after being left in a freezer the medium press takes 10s instead of 1s. In both cases, the value stays at 255 for a while before dropping at all, so shorter lengths of time cannot be measured.
I tried using the memory decay trick to measure shorter times, but it doesn’t work when the OTC is enabled. If the OTC has power, it will keep the attiny powered up enough to keep memory from decaying, so the shortest amount of time I could measure was like 4s with that approach.
I dont think there is a download for guppydrv at nerd camp. As far as I know the only open source codes DrJones has are the ones found in ToyKeepers repository.
The OTC shouldn’t power the MCU because it’s not on the VCC. As far as I know, the MCU cannot be powered by any other pin than the VCC. The input capacitor would keep the MCU alive, but I can’t see the OTC doing it. On the other hand I’ve had to deal with some OTC weirdness myself, so I wouldn’t bet my savings on it…
In any case, if you want to shorten the measurable time of an OTC, put a so called bleed resistor on top of it. I’ve used values of about 360K for mine. I’ve had to deal with bleed resistors enough to incorporate a dedicated resistor spot for them on my driver boards. You just put them in parallel to the OTC, which basically means on top of it if there isn’t a dedicated spot.
It shouldn’t power the MCU, but it appears to do so anyway. The mem decay time goes from ~0.5s to ~4.0s depending on whether the OTC was charged when the light shut off.
Weirdness… And even more weirdness as that I have a 13A based driver where the memory decay time is about 10 seconds, and it doesn’t even have an OTC on it. It’s all this weirdness that has me back on OTCs but with bleed resistors. I use the memory decay trick for storage between short clicks, but no longer base on/off time checks on it.