Is this thing gonna blow up in my pocket?

I'd want to use high current cells for that. The normal recommendation for safety is that the maximum discharge rate should be no more than twice the capacity in mAh. So for a 16340 claiming 800mAh it would not be wise to pull more than 1.6A. The protection circuit ought to cut in at such draws.

If the cells are good, they will tolerate this treatment for a while, but I personally wouldn't bet on it. The cells will probably fail benignly but I wouldn't bet on that either. You'd be wise to get some lithium manganese (IMR) cells for that purpose. They aren't that expensive here.

http://www.bestinone.net/battery?product_id=311

Not sure, but I know I inquired about this b4. But an lMR 16340 might do this. Do what Don says, I agree.

EDIT: by no means draw that many amps out of the cell u are using

OK that is what I was afraid of. I will just put the stock emitter back in, judging by the weight of this light heat managment will be the next problem. That's good information, thanks Don.

Im amazed the cell can supply the current .... , swap in another driver .... 1.4A 3 mode should do the trick ...

But still , Im just stunned that the cell can supply the current .

Still , if you dont mind killing the cell , I dont think it will flame on you , but certainly you will shorten it's life . [ Interesting light ]

I tried 2 different cells, protection circuits did not kick in.

You know what else is odd Old, I put the same cell in my WF501A with the famous Manafont drop in that draws 3+ amps on a 18650 and it only pulls 1.7 amps.

1.7 , Now that sounds like normal . But 2.7 , I just didnt know RCR123A's could do that .. Highest Ive seen myself is about 2.1A

If the light is multimode , you could just run it Med or Low ... [ You dont have to run it on high ]

Now that Ive had time to think about it , you might want to check for a mild short [ just in case ] ..

All things being OK , Id leave it be , OK so it might eat cells , but Id be prepared to feed the bulldog .

wow I have an a30b and it gets hot with the xr-e after a few minutes..I can only imagine that thing runnign an xm-l..lol

1.7 , Now that sounds like normal . But 2.7 , I just didnt know RCR123A's could do that .. Highest Ive seen myself is about 2.1A

If the light is multimode , you could just run it Med or Low ... [ You dont have to run it on high ]

Now that Ive had time to think about it , you might want to check for a mild short [ just in case ] ..

All things being OK , Id leave it be , OK so it might eat cells , but Id be prepared to feed the bulldog .

----------------------------------------------------------------------------------------------------------

Hi Old, it is a single mode light and I did recheck it for a short. Once I figure out how to post beamshots it will be obvious were all the amperage is going. I am going to try one of these is my MCU-C7 3 mode to see how it works before I change the driver. Thanks for the help.

Those IMR only have 350mah, AW 16340 are 550. 7 bucks or so though.

Almost every 16340 grossly overstates capacity. The IMR ones seem to be honest, at least the 14500 ones I got from bestinone are within a few percent of rated capacity. Now checking the AW IMR 16340s I have for capacity. Results in a while

I'm looking at getting their 18500's, 10440, 18350, and 16340 in IMR-cheap, and hopefully actually IMR. Can you test the 14500 at.... 3 amps?

Not and measure the result - my charger won't let me discharge at more than an amp. What I can do is measure beginning and ending current on my XM-L modded Mr.Lite J4. It pulls just over 2A at switch-on. The cell measured 568mAh at 500mA discharge but it makes no sense to measure IMR cells at such low discharge rates - that really isn't why they are made.

If it holds the initial current and can hold up at that discharge rate it should give 15 or so minutes on full power. Won't know till I try....

4:45 no perceptible drop in light output but the whole light is warm, not hot.

8 minutes, light output still steady - I have a light meter measuring output. Capacity at least 2000mA x 8 minutes = 16000mAminutes. Divide by 60, capacity is at least 267mAh

10 minutes - still steady 333mAh minimum. Light now hot but still holdable. Now using my hand as a heatsink

12 minutes - still steady - at least 400mAh

14 minutes - still steady output - at least 467mAh

14:50 output down a little - current draw now 1680mA. Will have to do some calculating to work out capacity. If we take the lower value, we still have at least 476mAh at 17 minutes. The light is now cooling - the driver isn't wasting as much energy as heat.

At 18:20, voltage down to 3.8V current draw 1670mA - at least 500mAh. In the time taken to measure voltage and current, light output has recovered to initial level (Measuring ceiling bounce)

20 minutes, voltage down to 3.71V, current down to 1500mA - another 40ish mAh - so overall capacity on a 1.5-2A load is around 540mAh.

Apologies for quoting myself - I got 437mAh at 1A from one of my AW IMR 16340's which I made sure was fully charged first.

Second one 402mAh

Third one 415mAh.

So a 100ma difference from AW to generic-for almost double price. For low current, generic wins for value, but at high currents, AW's are almost guaranteed to win.

I will also try to improvise a jig so I can connect one of these 14500s to a XM-L dropin which does pull over 3A at the tailcap.

Now where did I put the duct tape?

Depends a bit on your budget and required runtime. Yes, the AW ones will almost certainly perform better but they also cost quite a bit more.

They also probably charge better: thats the thing with high capacity cells: you can put more charge in faster. Also why IMR are good, if you have a charger that can supply large currents.

Now recharging the bestinone 14500s to try them in a Solarforce L2r with a dummy battery and the Manafont XM-L dropin.

It pulls 3300mA at switch-on. Time to fire up the stopwatch.

Down to 3080mA at 1 minute - 3080mA minutes / 51mAh

3080mA at 2 minutes 102.67mAh - measured one minute at a time as I have to hold the meter leads in place.

Minutes 3-4 3100mA on average 103.3mAh - 206mAh total.

This is with no heatsinking at all as none of my dummy AAs will fit in the L2r body. The dropin gets hot enough to hurt after 2 minutes on high.

In fact hotter than I'm happy with. Time to rethink this approach.

A roll of paper and L2i?

Real problem is holding the meter probes in place. I really need to build a jig to do this so I can use the datalogging meter and then get the area under the curve for a reliable capacity. It doesn't help that my high-current meter leads are very, very stiff (3mm2 solid copper). These leads are overkill but make the results more trustworthy.

These cheap IMR 14500s seem to have no trouble providing 3.3A. But the dropin is getting too hot as is the driver and I'd like to try to eliminate those thermal effects.

I'll get it done, just not tonight. I really could use one of those CBA things - but carriage kills the deal here - maybe when I'm in the US in August.