ROAR of the Pelican (CR123 Explosion during use, firsthand account)

[Lightmeup]

LM, I agree that those MSDSs are mighty lame. They could put a lot more specific information on them about what to do in these circumstances and the properties of the toxins that you could be exposed to. There was way too much vagueness and ignoring situations that they felt "normally are not experienced by the typical user", etc. The "typical" user isn't the one who really needs an MSDS.

[Paladin]

In the meantime, I think I'll just keep checking the Flash Amps of my 123s before installing them.

Seems like the easiest thing most of us can do.

Am I understanding the "flash amps" procedure to involve placing the ampmeter directly across the cell, subjecting it to a VERY low resistance short circuit. I've done that inadvertantly, and it always seems to drop the cell voltage several millivolts.

Why is a fixed resistive load not being used. Say a 10 ohm large wattage resistor was used. The cell voltage under a REASONABLE nominal load would seem to be a better "test" as it would be within the normal operating parameters. How a cell dumps energy into a dead short is NOT related to the design and useage of primary cells.

3 volt / 10 ohm = 300mA, a level several lights run at. And a resistor could be selected that would provide a load representing the nominal current drain of the specific flashlight the batteries would be used in. Having over 30 years playing with electrons, the whole concept of testing into an almost dead short runs counterintuitive to how I understand things. A test that does NOT simulate normal operation is often meaningless. For example, we could drop the cells from a window, and see which ones resist damage better...but that would tell us NOTHING about how recoil from a firearm might batter them.

I fail to see the significance of drawing excessive current from the cells as a test procedure. Perhaps I'm missing something?

Paladin

[batterystation]

Kevin: Thanks for the work you're doing. Is there a way we can send you small donations to help offset the costs of your investigation? We all benefit from your work, so we can all help. --Josey

Absolutely NOT!!! I have a LOT of flashlights here to choose from for attempted destruction. This thing has just begun. Lunar has donated enough already! I just want the safest flashlights and batteries money can buy (and sell). We will be offering SF batteries on the CPF page in about a week too for those that want them. I believe in our product but also realize that some customers as pointed out in earlier threads want other brands. I respect that and will do my best to accomidate.

The note on the Surefire box is a result of someone skewering a pallet of Lithium Thinyl Chloride batteries at LAX (Los Angeles Airport) about three years ago if my memory serves correct. The pallet caught fire when stabbed by the fork truck. A lot of different batteries would have done the same thing but it just happened to be lithiums.

[mdocod]

Am I understanding the "flash amps" procedure to involve placing the ampmeter directly across the cell, subjecting it to a VERY low resistance short circuit. I've done that inadvertantly, and it always seems to drop the cell voltage several millivolts.

Why is a fixed resistive load not being used. Say a 10 ohm large wattage resistor was used. The cell voltage under a REASONABLE nominal load would seem to be a better "test" as it would be within the normal operating parameters. How a cell dumps energy into a dead short is NOT related to the design and useage of primary cells.-Paladin

As I understand it- flash amps may actually be a BETTER representation of cell capacity than amps into a reasonable load.. 2 cells of significantly different life remaining- could generate very similar current accross a reasonable load and be difficult to distinguish- whereas into a dead short- a parcially depleted cell is more likally to show a comparable and significant variation in results...

In the experimentations I have done with lithium primary cells, sets of cells that all deliver a solid 10-12 flash amps, will generally result in LONG runtimes with even discharge in my lights.. I have also found many "new" cells that will only deliver a reading of around 2-4 flashamps... if I use one of those cells in conjunction with "good" cells in a light- that cell dies quickly, leaving me with many parcially drained cells, and one fully drained cell.. When I use sets of those poor performing cells (the ones that deliver less than ideal flash amps) together in a light- they tend to discharge somewhat equally- but the flashlight does not last as long as it would have with "good" cells...

my "gut" tells me that "flash amps" produce a result that has greater resolution than a test with a normal load... The proof of this concept is the ZTS tester... while it may give some indication that something about a particular cell may be "different" it is very obvious from the testing I have read that the numbers it generates for capacity should be taken with a grain of salt... Kevin was just saying a little while ago that a battery that tested low, and a bettery that tested 100%, both run through a single cell light, generated a runtimes that were nearly equal to the minut.. The ZTS tester- aparantly uses "normal" loads to determin it's results... I think there is too much acceptable variation in "full" lithium cells into normal loads to determin capacity....

I think it would be very interesting to compare what the ZTS tester says- to flashamp comparisons..

[mdocod]

[after] saying that... I thought it would be usefull to try to prove my point..


here's a comparison of 16 cells..
cell number/flash amps/reasonable load
01/08.9/0.76
02/10.2/0.77
03/09.9/0.78
04/10.7/0.77
05/05.4/0.74
06/09.0/0.76
07/08.8/0.75
08/08.0/0.75
09/08.9/0.76
10/11.1/0.76
11/09.7/0.75
12/10.2/0.76
13/05.4/0.74
14/03.9/0.74
15/10.2/0.77
16/10.7/0.77


cells 1-4, and 13-16 are "brand new" fresh from package cells..
cells 5-6 were used in various lights and are in my "unkown" pile.
cells 7-9 are parcially used from a 3 cell flashlight
cells 10-12 are parcially used from a 3 cell flashlight

here's what's most interesting to me..
a cell that produces 10.7 flash amps, produced 0.77 normal load amps
a cell that produces 3.9 flash amps, produced 0.74 normal load amps....
a 275% variation shows up in 1 test, while a 4% variation shows up in the other... which one is more obvious to you?

[OutdoorIdiot]

Brilliant data, thanks mdocod.

Inspired by that, I found this:

http://www.candlepowerforums.com/vb/...ad.php?t=76293
...which, for this discussion, gets interesting at post #30.

I wonder just how much of the cell's overall capacity is used up by a flash-amp test, though, and what damage it may do?

[bwaites]

WOW,

milkyspit was all over this a long time ago!! I found this particularly enlightening in reference to the current thread:

"Anyway, I ALWAYS measure flash amps of the partially used cells that I pull from my lights, then use a Sharpie permanent black magic marker to mark the reading directly on the side of the cell. I store them all upright in a Tupperware container with no lid (so it's more like a tray), with non-conductive spacers (foam, paper, plastic...) filling the empty space so there's no possibility of the cells falling over and perhaps shorting against one another. I keep the cells organized in order of their flash amps readings. Then, when I want to feed the MC2, I select a pair of cells with the same flash amps readings... plus I RETEST the cells just before popping them into the light, because sometimes the readings will have changed while the cells sit in storage. I believe strongly in this, and it's applicable to new cells and other 123 cell lights: NEVER PUT CELLS HAVING SIGNIFICANTLY DIFFERENT FLASH AMPS READINGS INTO THE SAME LIGHT!"

I will see if Scott still has those posts, (he seems to keep everything he ever wrote backed up!) and I can maybe post them in a new thread about cells in the battery section.

Bill

[OutdoorIdiot]

I will see if Scott still has those posts, (he seems to keep everything he ever wrote backed up!) and I can maybe post them in a new thread about cells in the battery section.

Great. I was a bit disappointed that the link to the thread with that quote is no longer valid. I'd certainly like to read the rest of that thread.

It would be nice if a best-practice for matching batteries can be generally agreed upon.

[greenLED]

When my new Surefire M6 arrived yesterday this was printed on the box in big bold red type.

PRIMARY LITHIUM BATTERIES -
FORBIDDEN FOR TRANSPORT
ABOARD PASSENGER AIRCRAFT
...

That only applies to commercial shipments, loaded in the cargo section of the plane. You can take a reasonable amount of primary lithiums in your carry-on.

[yuandrew]

Been following this so far. Hopefully, Lunarmodule, you've recovered by now. Those were some shocking pics in the beginning.

I'm still trying to visualize the venting of the batteries; I have some pictures I saved from Pydpiper's 123A experiments and there was one battery he got to vent with a flame by microwaveing it. Judging by how the flame is comming out, I can see what it might probably do inside a light.

http://img.photobucket.com/albums/v4...F/1deb7ccd.jpg
http://img.photobucket.com/albums/v4...F/2fa146a8.jpg

There was also the "charge with 120volts AC directly" experiment I suggested. Might have to give it a try again in a blast chamber. Pydpiper had it outside and was in his shop when he plugged it in so he didn't see it go off but reported the lights dimming and the yard lighting up in a huge flash.

http://img.photobucket.com/albums/v4...F/845518e4.jpg
http://img.photobucket.com/albums/v4...F/0d4aedc0.jpg
http://img.photobucket.com/albums/v4...F/PICT5963.jpg

[brightnorm]

...I will see if Scott still has those posts, (he seems to keep everything he ever wrote backed up!) and I can maybe post them in a new thread about cells in the battery section.Bill

Bill,

If you do this would you please provide a link in this thread?

Thanks,
Brightnorm

[bwaites]

I've asked him about them, I'm sure he'll get back to me tonite, as he had some family obligations today.

I will either post them here, or if they are long, I'll post in a new thread with a link.

Bill

[brightnorm]

Thanks.

BTW, I've just seen evidence that the ZTS tester needs to have 10% increments or smaller to give an accurate picture of remaining battery capacity (assuming this kind of testing is valid).

After doing a SPY005 run test the light still maintained a regulated "level 3" but the ZTS tester registered ZERO. I put the batteries back in the light and was again able to activate and sustain level three. I'm no longer confident in the ZTS' ability to give an accurate reading. What is our next step?

Brightnorm

[snakebite]

just for the hell of it today i took 2 titanium cr123 that were mostly discharged out on the dock at the store.
set my variac up to run a hamburger cooker at 150 f
figured thats about where most folks go ow thats hot.
heated the cells 10 minutes then hit the crimp area with a punch and hammer.
in about 4 minutes i got a venting with flame result.will repeat later with the minidv rolling and the ir thermometer handy.
was kinda neat to watch the cells spew flaming chunks of lithium that left white marks on the concrete where they landed.

[bwaites]

I don't (and haven't) thought the ZTS is THE answer. It's a piece, but my real belief is that this is a combination of multiple issues. SPY005 level 3 isn't that much draw. I can believe a cell reading zero would still run it.

I think the ZTS is capable of giving us a reasonably cost effective way to get BASIC ideas about cells, but probably isn't as accurate as flash amping. However, it is safer, and a general guide to status. After a cell has been used, it shouldn't be used with another new cell, that's pretty basic.

It also allows us a basic guideline for out of the box cells. It at least tells us they are in some way different, and shouldn't be used together.

If all the cells you test are new, it gives a basic idea of their status. If they don't test the same, don't use them together.

My current belief is this:

You start with a damaged/partially discharged cell AND a good cell AND a draw sufficient to slowly heat the damaged cell up. (If there is too much draw, it simply kills the cell and/or trips the safety) AND you use the light for long enough to get it all hot, AND the light has to be sealed AND the environmental conditions have to be right. (There may be other factors too, but I'm pretty tired right now!)

I am believing more and more that it is a multitude of factors, not a single issue.

It may just be that the PM6 has enough of those factors, especially the current needs of the lamp, perhaps compounded by the tight springs, that it IS more likely than most to create those conditions.

Bill

[MikeF]

just for the hell of it today i took 2 titanium cr123 that were mostly discharged out on the dock at the store.
set my variac up to run a hamburger cooker at 150 f
figured thats about where most folks go ow thats hot.
heated the cells 10 minutes then hit the crimp area with a punch and hammer.
in about 4 minutes i got a venting with flame result.will repeat later with the minidv rolling and the ir thermometer handy.
was kinda neat to watch the cells spew flaming chunks of lithium that left white marks on the concrete where they landed.

Man, I am really concerned that someone else is going to get hurt by all of these "tests". The delayed onset of LM's symptoms show how serious exposure to the components of CR123A batteries can be.

PLEASE BE CAREFUL!!!!!!

[batterystation]

Just for giggles I ran a set dead in the Hellfire that was a 10% and 100% mix. Nothing special happened. After the light went out and cooled, I dumped out the batteries. An hour later, I took the front one and ZTS tested it and it read but did not light anything. It ran the test LEDs back and forth but lit nothing on settling. I put it in my Surefire L1 and had cats eye green glow on low and decent light on high. The darn thing still had juice in it AFTER being "reverse charged"? I am very puzzled by this.

[slinco]

Holy $hit-I did'nt even know that could happen! Violent....

[OutdoorIdiot]

{Edit: I was responding to Brightnorms post when I wrote this - hadn't realised so much had happened in the meantime!}


I don't feel qualified to answer a question like "what is our next step" in terms of trying to match or assess cells, but I'll happily post my recent findings.

I've just been doing some experimenting which, among other things, has involved making flash-amp measurements at various stages as batteries are depleted.

I don't think I'll have completed all the things I want to do until tomorrow or maybe even until after the weekend, but here are some interim results that relate to flash-amp readings, taken from batteries running a P60 lamp:

Fresh: Approx 6.6A
After 20 mins: Approx 5.6A
After 40 mins: Approx 5.2A
After 60 mins: Approx 4.4A

And for a P61 lamp (which I think induced thermal shutdown in the cells, but anyway):
Fresh: 6.6A
After 12 mins, and cooling: 5.7A

Comments:
(1) Nice spread of results, i.e. there is a good distinction between one reading and the next as the battery depletes, which, as mdocod has pointed out so well, is a good thing.

(2) Taking a flash amp measurement does not seem to appreciably shorten battery life or damage batteries (this was a concern of mine). The P60 in G2 was still blindingly bright after three 20-minute runs and 2 flash amp measurements on both batteries. Since the runtime is given as about 1 hour, and it was still going strong after this, I'm happy with that aspect. Also, the needle on my ammeter is slow to move, so I would think that if anyone is going to damage their batteries, I would. It probably took about 3 seconds to make a reading.

(3) As Milkyspit points out in the thread that I linked to a few posts ago, it would seem that one ammeter will give quite different results to another. I was concerned that my ammeter only goes up to 10A DC, and mdocod's results occasionally reached around 11A. Fortunately, for me, my top reading was no more than 7A for fresh cells - so it does seem important to stick to one ammeter if you want to compare cells.

I'd avoided using the flash amp method in the past, but after a few hours of experimenting and getting confident with it while doing some tests, it's "grown on me", to say the least.

I have just had an unusual experience which has produced an awful smell (I'm still referring to flashlights and batteries, btw), but I have no reason to suspect it was as a result of damage caused when shorting the cells to read the current. Rather, I think it was just because something got too hot - the batteries seem to have survived well, both physically and in terms of output.

[OutdoorIdiot]

just for the hell of it today i took 2 titanium cr123 that were mostly discharged out on the dock at the store.
set my variac up to run a hamburger cooker at 150 f
figured thats about where most folks go ow thats hot.
heated the cells 10 minutes then hit the crimp area with a punch and hammer.
in about 4 minutes i got a venting with flame result.will repeat later with the minidv rolling and the ir thermometer handy.
was kinda neat to watch the cells spew flaming chunks of lithium that left white marks on the concrete where they landed.

Thank you. I wish I was brave enough to do something like that myself.

If Keving gets similar results on Saturday, by preheating batteries and then crushing and/or mis-matching them, then we could well be on to something. The next step would be to see if there is indeed something about the original PM6 that allows batteries to get hotter than the average flashlight, but that would be very much jumping the gun at the moment.

[Lightmeup]

Just curious, does anyone know how LM is doing? He hasn't posted here for awhile.

[OutdoorIdiot]

Just curious, does anyone know how LM is doing? He hasn't posted here for awhile.

Is it just me, or has everyone else been thinking that but too afraid to mention it?

Anyway, in an earlier post, InFlux said that he had his phone number - possibly worth PMing InFlux if LM doesn't respond to PM.

[PlayboyJoeShmoe]

Like I said, reasonable pre-cautions SEEM to be safe. I'd be WAY WAY careful if my early PM6 was stock. But with a cut down LM spring and a kinder softer Kroll spring, I ain't skeered!

I learned early on about flash amping and I ain't skeered of that either.

Still

[Archangel]

When you say it registered zero, do you mean it didn't even "run" the LEDs? Technically that means that it's less than 1%, but not necessarily "dead".

BTW, I've just seen evidence that the ZTS tester needs to have 10% increments or smaller to give an accurate picture of remaining battery capacity (assuming this kind of testing is valid). After doing a SPY005 run test the light still maintained a regulated "level 3" but the ZTS tester registered ZERO. I put the batteries back in the light and was again able to activate and sustain level three.

[batterystation]

I just spent an hour+ on the phone with Lunar and he is doing great. Back to normal he says. He will catch up later.

[brightnorm]

When you say it registered zero, do you mean it didn't even "run" the LEDs? Technically that means that it's less than 1%, but not necessarily "dead".

As best I remember it didn't run the LEDS, but even if it did it ended up with all LEDs dark.

BN

[chiphead]

Oh crap, I've got over 50 of those batteries here!

chiphead

[Presidio]

Glad your ok LunerModule.

I had the same thing happen to my pm6 (minus the blood) but with new surefire batteries, and after repeated unanswered emails to pelican I said the hell with them and I'll never purchase anything made by pelican again.

[bullinchinashop]

O H M Y G O D....
I've got about 15 of the Battery Station 123's sitting in a drawer right now. The reason this freaks me out is this : About a month ago I had these batteries in a sealed plastic container. I opened the container to take out a couple and there was a V E R Y strong metallic/acidic smell coming from the batteries! I left the top of after that and the smell is still there but much fainter. I'm VERY tempted to throw these batteries in the garbage after reading that. What if both of those batteries had exploded at once while that light was still in his hand ?!

[bullinchinashop]

Glad your ok LunerModule.

I had the same thing happen to my pm6 (minus the blood) but with new surefire batteries, and after repeated unanswered emails to pelican I said the hell with them and I'll never purchase anything made by pelican again.

So maybe this is just a Pelican problem? If I remember correctly the last post where this happened (inside a wooden cabinet) involved a pelican light. hmmm...
And the fact that they are not even answering your emails make me even more paranoid about them. They could at least shoot of a quick "@!$# you! Our lights are fine the problem is elsewhere! Now leave us ALONE!"