found this youtube vid that somewhat fits what you asked?
I am also trying to find an aux keychain fob once discussed here. Something along the lines of AAAA light powering 5mm with joule theif
If you power the aux leds (could be an illuminated tail or behind the optic or wherever) with the output set at trit levels, directly from the battery so no MCU is involved, they could last for 10 years on a battery charge. But of course the temptation is to set them brighter, because you can 
would EL (electroluminescent) make any sense?
it is what they use for ‘indiglo’ backlights
it can;t be all that power hungry or expensive for lights of that size
but i am too lazy to do the design and research
i just say no one mentioned it.
wle
Glo Toob I had one in red, and the battery seemed to last forever.
Another DIY light used a resistor and a 9V battery. I believe you could even skip the resistor, since the battery’s internal voltage is pretty high.
I looked at EL, but it seems they are 100v+, even if the current is low. It’s mentioned in the first post. I think it would be really cool to have an EL ring around the head and tail of the light, and cover it with frosted clear epoxy. It would look like a giant round tritium vial.
Honestly, im content with having a lighted tail switch board, and simply want there to be more compatability of it with drivers. That said i still have yet to switch over from blue to amber, or red on my kronos set, but i will at some point
It would be interesting if such a LIT module could be developed. Not sure 1.5mm x 6mm would be possible. Might need to be larger.
So the output for the LIT would be an extremely small drain relative to LED emitters.
The backing that the LIT is mounted upon would have photovoltaic properties, for charging.
The charge would go to a micro capacitor, built-in behind the photovoltaic layer.
If material specifications would allow, the LIT would output tritium level brightness, or perhaps a slight bit more. The capacitor would be capable of storing up to 48 hours of run time. If the LIT is stored inside the LED chamber, the spill from the LED would be enough to fully charge the LIT in 30 seconds, maybe less. If the LIT is in the tailcap or some other location away from the emitter, all you need to do is hit it with another flashlight LED emitter for 30 seconds or less. Of course, ambient light in a room or sunlight would charge it. The LIT would not emit while it detects sufficient light, thus always in charge mode when available light is above the threshold.
I’m not an engineer, just brainstorming off the top of my head. Does this proposal above sound viable?
i know they use a high voltage, i was thinking maybe some of them come with the booster built in now, or EL drivers are available
wle
Yes yes glo toob
Those are definitely in my wishlists till now but the price just made me buy proper lights instead hahaa
Someone needs to invent coin cell batteries assembled with built-in LEDs or a COB strip. Pull out a little insulating strip, and it lights up.
I wonder if solid state electronics exist small enough to do that along with an off-during-daylight sensor/switch.
Or, you just make a slightly larger coin cell sized light that has a slot for a coin cell battery you just slip in. It can have an incorporated pull tab where you pull it to activate then push in to turn off. Magnet on the back.
Y’mean a throwie?
There is always this, the cost is pretty high, but if you get a high enough content of radium it is pretty bright. of course there might be some health considerations
Here’s the DIY version. About a dollar each or less in quantity.
Here’s an instructable how-to.
A photoresistor or phototransistor allows more current when it’s in the light, so you can’t just put them in series with the LED to shut it off. You’d need a slightly more complex circuit as in” this instructable”:https://www.instructables.com/id/LED-nightlight-that-turns-on-in-darkness/. With SMD devices, you could fit it all in the outline of the coin cell.
@wle - I was mentioned the high voltage to point out that it probably puts it out of the energy budget. Maybe not, though.
Adafruit: Each meter of high brightness EL draws about 10-15mA at the high voltage, which means about 1.5 Watt/meter (at 100VAC).
The boost / invertor circuit would probably not be all that efficient, but could be tied to the main battery. If you could drive a reasonable (1cm?) length at a 1-10mA budget, you’d be looking at only 30 or so hours per battery.
You guys have my wheels turning on this, despite me having no actual use for it :person_facepalming:
I had considered trying to integrate a 10180 with openings for charging, a small PCB, some SMD LEDs, and a 3D-printed housing. But now I’m wondering if it would be better to just use a CR1632. With a small draw (0.05 mA, perhaps) it would last approx 4 months. You could go even lower for longer runtimes. It could probably be as small as ~22mm diameter x 6mm tall.
The coin cells (such as CR1632) have an extremely flat voltage curve at ~2.9V until they’re dead which means fairly consistent brightness and very little excess voltage to burn off. And with a coin cell, there’s no need for a protection circuit - just run it until it’s dead. Hmm…
I have lighted tail caps for sale, they can last well over a year with low drain
For sure, I agree Lexel. Where you can, a good lighted tailcap is indispensable.
While we have your attention, have you ever seen the HM1160 ? Sure, it’s not as configurable as individual voltage references or voltage reset monitors, but having a SOT23-6 footprint 4-level battery monitor IC sure is compact and convenient (and cheap, too!).
some coin cells are rechargeable iirc
those for rtc clocks. Even some keychain solar light uses them.
looks similar to cr2032 but rechargeables
Those are the LIR series coin cells, such as LIR2450. Capacity (per volume) is much lower than their non-rechargeable brethren. And they’re many times more expensive. And then a protection circuit would be needed (a simple voltage reset monitor would suffice, but it further complicates things. These things are acceptable depending on what you want the end result to be. That is, would the extra size (or decreased runtime) and expense worth it to make it rechargeable?
Supercap…