Close: STAR_on_time and STAR_off_time both use a threshold value to switch between using both PWM outputs and only the ALT PWM output. You may have been thinking of STAR_momentary, which is not usable here.
Makes sense. Like I said, I still have a lot of work to do.
…
Yeah I was just going to try to dial in the pwm for high. I want it to be usably bright without draining my 10440 in 30 seconds or getting too hot
Thanks wight as always, and double thanks for the 13mm expansion and the point about the wires.
About the low voltage protection, that i meant from the driver, my sister wanted an olight i3S with a efest 10440 after she saw mine, but i can’t in good conscience give hear something like that when it can go so low before it dims so she could damage the cells and not even know it before it is to late.
She is a 30 year old women who is still afraid of the dark 
so i want to make her something special so she can feel safer :heart_eyes:
When you say cover the mcu with copper tape do you mean to use as the strap/battery + ? could i just use some kapton over the mcu and use copper tape as the batt+?
And yes now that we have a 13mm board it will fit the driver lip it will be sturdier, i hope it is possible to shorten the bottom spring enough to accommodate the extra height down towards the battery so it still mechanically functions with the proper twisty action. otherwise we have to drop in the driver deeper in the driver compartment and it is then it could be hard to get it really sturdy, as djozz found out in his mod of this host, it is that what i hope to avoid.
the height of the mcu added in the battery tube won’t change where the tube meets the driver, so the twisty action shouldn’t be changed. It might make protected cells not fit, however.
Cajampa, is the “driver area” aluminum or something else? Is it a removable pill or one piece with the body of the flashlight?
pilotdog68
i hope you are right about the twisty not being affected by a shorter travel of the spring, i have never done surgery on a twisty before so can’t know how it will affect, but people complained of iffy mode changes on this light in stock so i have just assumed that changing it to much could make that worse, but it was maybe just a unreliable stock driver on some of the lights.
I just measured the maximum an eneloop aaa can travel down in the tube, and it is 2,6mm…………this could be a problem i wonder how high an attiny13a is from the driver? less i hope, or else i see some shaved attinys in our futures 
Then whole head is one piece, everything is anodized aluminum.
have you seen these previous threads on this light
djozz has already done some mods on this one.
and some more discussions here
Unfortunately we are having a communication breakdown of some kind. Maybe you are not aware of how LVP works in these drivers. I’ll use STAR_off_time as the example. We set a voltage at which the MCU will cause the light to blink 3 times quickly and then reduce the brightness. In a set period of time it will do the same check and if necessary blink 3 times and reduce the brightness again. Once voltage reaches another point (which we also chose) the MCU will turn off the light. At that point the light is only asleep, but for single-cell lights the draw is incredibly low. (I assume that it’s not significantly worse than a protection circuit but I could be pretty wrong about that.) All you have to do is tell the user to take the cell out and charge it if they see the light blink several times.
RE: copper tape. Yes, I’m talking about creating the BAT+ connection. I don’t think you need to use Kapton in between, just put a strip of copper tape across the MCU.
The ATtiny13A is approximately 1.6mm tall by my measurements. The datasheet lists a MAX figure of 1.75mm. Both of those figures are without crushing the legs. The bottom of the plastic package does not touch the PCB unless the legs are crushed, maybe gaining 0.1mm.
Would the trace from a protection circuit work fine as well? I have a few I stripped before I got rid of my ultrafire batteries. I assume yes, but I might as well ask
I suppose so, but it’s awfully tiny isn’t it? I’d want something larger and more sturdy I think.
its about as wide as the longest dimension of the attiny. If it was any wider i would be afraid of shorting on the side. But you’re right when it comes to thickness, its pretty flimsy
The thickness isn’t a big deal I think. I was thinking of the ones which are around 1.5mm wide: too narrow for this purpose IMO. If it covers the ATtiny13A’s plastic package I think you’ll be fine.
EDIT: do you have calipers?
yeah i have calipers at home, but im at work for another hour.
When you get a chance you can take the width and thickness dimensions of that strip if you don’t mind. Then we can see what AWG equiv it is.
the more I think about it, the more I think I’m overstating how big it is. It’s probably tiny and my brain is screwing with me. I’ll measure it when I get home.
Yes i know we set if from the driver, i don’t really know how yet but i will find out when it is time for fine tuning, i thought you thought i was talking about protected batteries before when you said “Now, as far as needing protected cells… why?” as a response to when i said “and i hope to be able to build these good enough to be able to gift to people and with this drivers possible low volt protection on the 10440” if you didn’t then i misunderstood you.
As far as i know the available protected 10440 is crap anyway, and the not officially but 10440 capable small popular lights with workable mode changes, like the olight and thrunite don’t have any protection at all and will run down batteries very easily to dangerous levels.
But i didn’t know exactly how it worked because i haven’t played with them yet, the attiny on these very capable driver i mean, so thanks again for filling in the blanks.
Very nice that the ATtiny13A is 1,75mm this means one less potential stumbling block avoided this feels more and more possible by the minute.
Djozz made it work stacking the big FET on a tiny12, we should definitely be able to do it with the tiny driver wight put together
Yeah djozz used that huge Vishay FET, we have it possibly much easier.
But it still looks to come down to less then 1mm so it is tight very tight
The trace is 2mm wide by 0.09mm thick. Definitely smaller than I thought. (and for what it’s worth the pcb is 0.86mm thick)
That’s nearly 24AWG equiv (it’s better than 25AWG). Who knows what the material is, but I assume it’s not that bad…