Wow. Just caught up with the thread and really like your renderings. The sled is an interesting idea too especially if you were thinking something like what jib posted. Something has enough room on it for two or even three flex drives and associated wiring would be awesome. Then again I'm not picky so as long as there's a flat spot or two somewhere to mount a couple drivers I'll be happy.
Oh for diode removal you could probably still use FP's extractor if it can fit over the top of the can. You'd just have to gently tap it out instead of screwing the plunger down like in an aixiz module. Actually if you remove the lens depending how deep the focus adapter is you still might be able to screw it into it and use it but if the diode is truly dead I'd probably knock it out with a center punch or something similar. If it's dead who cares if it gets broken in the process.
Edit: Wow you added another rendering while I was typing up my post. I like it more and more with every update. Keep up the great work!
The chip can be set for different voltages, R1 and R2 supposedly adjust the threshold. You would want something between 3 and 2.8V for an 18650. The type of battery should have no effect at all, so long as it's in the supply range of the chip.
I've tested many different things in Linear's simulator, but nothing I try works. Everything I set in the simulator should work in my head, but there's a great room for error there and it's not in the simulator :crackup:
LtLernelPanic - I could make a very simple diode extracting tool specific to this host, all it would be is an aixiz sized lens but it would have a lip that fit between the can and host. You would then thread the removal tool into the focus ring and tighten the focus ring to press the diode out. Something that simple would be cheap to make...
Hmm that LTC1440 looks interesting. It even has a TTL compatible output that can be tied to a driver's enable pin to automatically shut it down when the voltage reaches the programmed threshold. Gotta do some reading on how to program the threshold voltage ... looks like it might just be a voltage divider.
Edit: we could just use the example on the datasheet, 2.9V is a good cutoff voltage anyways. Ill start work on a board.
In theory, there is less energy in 2x16340 than 1x18650. There's less volume, and more space taken up by having twice as many contacts, and potentially twice as many protection circuits.
The only real question would be whether the necessity of using a boost driver for 1x18650 results in an efficiency loss that would more than offset the battery's higher capacity (relative to 2x16340). Unfortunately, I don't know the answer to this question.
A nice alternative, that often works, is the slightly more rare (but still readily attainable) 18350 battery. Two of those will often fit in the same place that a single 18650 would usually go.