I'm also working on improving the accuracy of battery voltage reporting - so far it seems dead-on to rounding on a Q8. For example, 4.06V to 4.15V reports as 4.1V.
It is threaded. But your OCD will hate it to scratch the ano on that retaining ring, which is very easily done. So the fix starts with filing to shape a plexiglass device that fits the notches in the retaining ring, or print it.
1) buy a leaning tower of Pisa print
2) Have someone else hang it for you
3) don’t straighten it, because if you do, it will need straightening.
4) I said DON’T, refer to 3
Obviously I was discussing electrical solutions, and I never said the diode on the mcu eats LED voltage. A diode on each battery would. The mechanical solution sounds good.
There’s a no-cost way that everyone can do to avoid shorting out the cells that way. I know it’s a tough one, but at the no cost price point it should be something literally everyone can do.
Pay attention when inserting cells.
It’s THE easiest way to ensure your light keeps working, time after time….
(Yes, I know, it should go without saying. Apparently it needs to be said louder and more often.) Lights like these are not toys, the danger is real. What if, and I’ve seen it before so I say this in truth, what if the spring welded itself to the cell in that sudden burst of extreme current, instead of dropping the cell out of contact? Milliseconds come into play, things get ugly really fast when the incident takes a slightly different turn…
I have used a red Sharpie marker to make a red ring around the + end of all my 30Q’s, both flat tops and button tops. I used red tape on some at first but one light had a clearance problem with the extra thickness. So I went to the marker.
I’m not sure that would make sense. When the cells need to be changed, the light source required to charge the GitD material wouldn’t be functional. So they probably wouldn’t ever glow when glowing would be useful.