I'm searching for a keychain white light running on 2x2032

[Calina]

Do you know of any keychain light running on 2 x 2032 cells ?

A 2 cells 2032 light would still be very small but would have much better runtime.


Thank you.

 

[paulr]

2x2016 lights do run for a very long time. Have you ever actually used one up? If it's really an issue, maybe you should think about a 1AAA light. However, the new Petzl E-light+ headlamp runs on 2x2032. Maybe there's a way to set it up as a keychain light. It has three white leds and one red led.

 

[abvidledUK]

RT graph

Keychain light, RunTime, One LED, 2x 2016



is in
https://www.candlepowerforums.com/threads/133886

 

[Calina]

Altough 2x2016 lights do run for a fairly long time, I find that their intensity drops
very fast. I end up using a new light (fauxton) every couple of weeks or so, I also happen to have easier access to cheap 2032 cells than cheap 2016 cells.

I saw the Petzl E-light and was wondering to, if it could be used as a keychain light.

 

[Calina]

I would say mine drops at least 50% after two hours. Still usable but too dim for my taste.

 

[paulr]

I think coin lights run at really severe overdrive during the first 1/2 hour or so. You can't count that as normal operation. To do a reasonable runtime measurement, I'd say run it for 1/2 hour, measure the brightness (call this B30, the brightness at 30 minutes), then run measure the time until the brightness reaches B30/2 (50% of B30).

More controversially maybe, I think the "runtime to 50%" notion is based on incandescent lights, since their color shifts towards infrared as they get dimmer, so at 50% they're really a lot less useful. LED lights don't shift color much, so their usefulness as they dim is logarithmic, like the response of the eye. So I'd actually go much further than 50%, say to 25%. Since the brightness of a direct drive LED light decays slowly, it takes a long time to get to 25% of B30. In fact I'd say the Fauxton's output is comparable to a CMG Infinity (a popular 1AA light of a few years ago that inspired the Arc AAA) even at 10 hours of operation. The fundamental problem with Fauxtons if you think they're not bright enough after a few hours is simply that they're 5mm led lights which inherently don't emit all that many lumens.

So again if you're really burning through Fauxtons that quickly then maybe you should consider a different type of light, like an Orb Raw or Fenix L0D CE, etc.

 

[abvidledUK]

RT graph added to my earlier post

 

[2xTrinity]

I do agree though with the idea. 2 x 2032 would not be physically much larger than 2x2016, and I would love the extra runtime as I use my photons a lot, as anything much larger really wouldn't work on my keychain (I like to keep my EDC items as minimalist as possible). One problem though is since these batteries are direct-driving LEDs, the 2032s have much lower internal resistnace than 2016s, so they would cook 5mm LEDs almost immediately. I actually hooked up 2 fresh 2032s directly to a Cree star, and they were pushing 300mA through the Cree, which on paper should be close to 100 lumens! Even if that settled down to half of that shortly, that would sure be one hell of a coin cell light at 50 lumens in a package only marginally thicker than a photon!

Of course, fitting an optic on there that could concentrate the light efficiency without killing the form factor would be hard, but at that brightness, I would jsut as soon leave it as a pure flood light. I believe I saw somewhere that there are rechargeable 2032s available as well...

Right now though all my crees are already accounted for (the one I tested is actually soldered to a heatsink, so it won't work to try that in a coincell) Making a crazy bright 2032 light with a Cree LED, ideally out of some sort of makeshift metal case, is an idea I've toyed around with.

 

[AndyTiedye]

AW doesn't sell the li ion button cells. He sent me here.

I haven't tried them yet. I was kind of hoping someone would do a group buy
so we could get them for $1 each instead of the $5 "sample" price.

If we use li ions, 1 can provide enough voltage, and won't fry the LEDs.

 

[TORCH_BOY]

All the ones I know of run from 2 x 2016 cells

 

[ViReN]

Mek Turtle Light is 2 CR2032

 

[Calina]

Mek Turtle Light is 2 CR2032

Do you have a picture or an address?

 

[Calina]

OK, found it at MEC (Mountain Equipment COOP) for 3.50$ CDN; nice.

http://www.mec.ca/Products/product_detail.jsp;jsessionid=FSD0zY8x5CnG1Spc9J2yb6zS9m241TvQ93hFTL9gbnC8QfKrHQ7V!-1836568381?PRODUCT%3C%3Eprd_id=845524442588127&FOLDER%3C%3Efolder_id=2534374302692895&bmUID=1171440680214

Thanks Viren

 

[paulr]

300 mA from 2032's?! Such a light would definitely need some kind of voltage limiter or regulator then. Chanik posted that he found 2032's (at least from Energizer, I guess) to have around the same internal resistance as 2016's and therefore around the same 50 mA limit. I'm surprised that low voltage leds (e.g. red leds have 2.2 volts or so Vf) in 2032-powered coin lights don't burn out. Maybe those were unusual 2032's that ran that Cree? Anyway interesting.

 

[2xTrinity]

300 mA from 2032's?! Such a light would definitely need some kind of voltage limiter or regulator then. Chanik posted that he found 2032's (at least from Energizer, I guess) to have around the same internal resistance as 2016's and therefore around the same 50 mA limit. I'm surprised that low voltage leds (e.g. red leds have 2.2 volts or so Vf) in 2032-powered coin lights don't burn out. Maybe those were unusual 2032's that ran that Cree? Anyway interesting.

Another point worthy of note is that the resistance on the Cree, as a higher-powered LED, is going to be lower than on a 5mm. I actually tested one of my old crappy LEDs with the same two batteries, and it was pulling 120mA, a severe overdrive compared to the 90mA draw from 2025s, and a 60mA draw from 2016s, but not 300mA.

My understanding of internal resistance is that the higher the capacity, the lower the resistance. This is for the same reason that a larger diameter wire will have lower resistance than a smaller diameter wire of the same material. I believe this is also why the resistance increases as batteries are discharged -- the discharged chemicals are poorer conductors, so the remaining capacity is lower, thus resistance increases. This appears to agree with my results with the 2032s, 2025s, and 2016s using the same LED.