123 Primary Lithium cell info/testing/links

[Beamhead]

Slow motion video would be sweet!.

[NewBie]

Okay, how is this?

http://www.molalla.net/~leeper/vent10s.wmv

We have some sparks a bit past the middle, and additional ones at the very end, so be patient. Watching the beginning a few times, you see something new each time.

[cy]

betcha... a lot of folks would like to do more than say thanks, by supporting your efforts. As this valuable work benifits all of us.

sure wish you would post paypal info or other means to send funds!!!!

Thanks. Hopefully I don't send myself to the poor house in the process, lol..

[McGizmo]

Newbie,
What CY said!!!!! I have 3 more boxes as well as a 27LT headed your way as you know but it would be nice if the community could help pick up the tab on this work of yours!!

I don't know about Tom (SilverFox) but I was a bit disappointed that we ended up with as much of the tab on the light tests as we did. (Many thanks to those who did contribute to Tom; I don't know who you all are but that is not the point, really).

Anyway, if you don't want to provide your PP address, perhaps someone else could accept funds on your behalf and forward on to you? (I am not volunteering here, BTW!!!)

Thanks again though and I look forward to page 2 where we won't have such a long load time!!!

[liqht]

newbie, have you seen any reports of incidents involving single cells? is this phenomenon attributable ONLY to configurations where there are multiple cells with mismatched charge? thanks...

[NewBie]

newbie, have you seen any reports of incidents involving single cells? is this phenomenon attributable ONLY to configurations where there are multiple cells with mismatched charge? thanks...

There is the one Air Canada reference above, and a link to the document. I think it was a single cell LED. But there are no specifics, so it is really hard to say. For the scenario we are talking about here at the moment, you need two cells. Thats not to say there isn't another scenario which could cause issues.

I have not seen any reference or mention from a CPF'er on a single cell lithium primary failing in an LED light.

I'd been searching for a failure mode for a months worth of testing and so far, the only thing I have found is the mis-matched cell scenario- which is easy to imagine occuring, due to new cells comming in various states of charge (some folks have had occasional cells so depleted, that they have only worked for five minutes).

All the incidents have been on 2 cell Primary Lithium 123 batteries with Incandescents which we have seen around here, so far. Besides the single cell AA Lithium that failed in some cell phone charger hooyah.

Humm, well, McGizmo, if folks want to pick up a bit of the cell cost, I guess I could accept that. I'd prefer not to get dollars, but cells. Right now, I'd really like to see a few more BS cells (for instrumented testing), and some other brands that folks utilize, which I haven't tested yet. It would be good, if a well know CPF'er could handle it.


Anyhow, I was asked for a small snippet of the main part of the last video, in slow motion. I just got back from buying 1GB of CAS 2 RAM to add to the machine I do the editing on. I kept the size small for low bandwidth users, it is about 2.6MB.


http://www.molalla.net/~leeper/ven10sm.wmv


.

[wquiles]

Thanks NewBie for all of the valuable and relevant data you have collected and shared with us here - much appretiated

Will

[NewBie]

Here is that video of a cell venting down at 100 degrees Fahrenheit, only 2MB in size:

http://www.molalla.net/~leeper/100degv.wmv

[Billson]

I have not seen any reference or mention from a CPF'er on a single cell lithium primary failing in an LED light.

Actually, I've personally experienced a Yuntong cell failure while using my Arc LS about 2 years ago. I guess this was still an unheard of incident back then so no one responded and the thread died without resolution.

Here's the thread

http://candlepowerforums.com/vb/showthread.php?t=52326

[NewBie]

Okay, here is another venting video, this is a very minor fireball type event that can happen:

http://www.molalla.net/~leeper/vent12sn.wmv

[CroMAGnet]

Thanks again Newbie. I guess that super cheap 2C to 3CR123 MAG-mod has the potential to cause a fire. I'm taking those batteries out. I don't even use that light anyway, since I have much nicer ones.

Has the L92 or L91 been discussed regarding these concerns? (sorry if I missed it somewhere.) These threads have been popping up a lot lately and I haven't kept up. Just kept out... of the multi battery CR123 lights.

[cy]

Holy Sh*t!

what a fireball, temp spikes to 263 degrees...

[Grox]

I would also be willing to contribute to your testing expenses NewBie!

[Hans]

Okay, here is another venting video, this is a very minor fireball type event that can happen:

http://www.molalla.net/~leeper/vent12sn.wmv

*That* was a "very minor fireball event"? Whow. It certainly didn't look "minor", and I hate to think what would happen if this sort of thing occured in a sealed flashlight.

I'm not really sure how one can proceed from here. Your experiments have established one set of conditions under which such incidents may occur. But the question is of course whether there are other possible scenarios. Because one thing is for sure - this is not the sort of thing one would like to happen when you hold a light in your hand or carry it on your person.

I'm not sure whether all the people who reported incidents with CR123 cells over the past year or so are still here, but it might be interesting to analyze these incidents again to see whether they occurred in the same sort of conditions.

Hans

[cy]

I had my first almost instance three days ago. fortunately I was using three Surefire cells that checked 100% with ZTS MTB-1 tester.

Was carrying Surefire L6 w/clicky in my backpack as usual. operator error of not locking out tailcap. L6 got accidently clickied on.

fortunately I discovered L6 in on position. picked up L6, it was almost too hot to hold. (would estimate 150 degrees+) quickly unscrewed cap and dumped out cells to cool.

tested cells later with MTB-1 at 80% for all three. Extremely glad I purchased my ZTS tester!

[NewBie]

Holy Sh*t!

what a fireball, temp spikes to 263 degrees...

Yes, so we have seen events on tape with temperatures right before the venting at 100F, 140F on this latest one, 153F,
161F, 161F, 171.2F- with temperatures running up to 714F.

The "fireball" one that was just made: http://www.molalla.net/~leeper/vent12sn.wmv
The 714F one was here: http://www.molalla.net/~leeper/vent6_6.wmv
171.2F one: http://www.molalla.net/~leeper/ven10sm.wmv
Another 161F that goes above 700F: http://www.molalla.net/~leeper/vent5_1.wmv
A 153.6F one: http://www.molalla.net/~leeper/minorv~1.wmv
The 100F one: http://www.molalla.net/~leeper/100degv.wmv

I've seen many more of these, which I do not have on tape. These are still by and far not the worst events I have seen- only what I happened to catch on tape, once someone asked for a video of some of the failures. I've missed more dramatic ones. I've tested a few more of these than I'd originally planned so I could demonstrate what can happen.


I'd really like folks to go back to the first post in this thread, where I talk about the safety factors and some of the things folks can do. Take the time to read it, it is worth it. I'm doing this because I see an actual possible danger, and there are things that can be done which greatly reduce the risk.

There is a label on every CR123A cell that I've seen in the US, and it would be good for folks to take a pause and read it. I believe it is there by law, for a reason. It is not there for sh*ts, grins, and giggles.

Please use these cells wisely, and treat them carefully- These ain't your grandma's carbon-zinc cells.

I have further testing to do, to learn more about the events leading up to these events, such as watching currents and voltages across each cell, while watching the temperatures- to glean anything out of this info, so designers might be able to design in countermeasures.

[Kevin K]

Maybe a little off topic, but, has anyone heard of lithium batteries violently venting in digital cameras?

[milkyspit]

Cy, I wouldn't have imagined the L6 causing an incident. I know the L4 left on can get literally too hot to hold... I've done it once... but the cells had no problems at all. Did your cells show any signs that anything was amiss?

I had my first almost instance three days ago. fortunately I was using three Surefire cells that checked 100% with ZTS MTB-1 tester.

Was carrying Surefire L6 w/clicky in my backpack as usual. operator error of not locking out tailcap. L6 got accidently clickied on.

fortunately I discovered L6 in on position. picked up L6, it was almost too hot to hold. (would estimate 150 degrees+) quickly unscrewed cap and dumped out cells to cool.

tested cells later with MTB-1 at 80% for all three. Extremely glad I purchased my ZTS tester!

[milkyspit]

Newbie, this is a GREAT thread and yes, someone please sticky it!


One question, and please don't be offended by it... I'm just performing the requisite due diligence... you seem to speak kindly of Duracell relative to the other brands... do you have any sort of affiliation with the Duracell folks? I mean no disrespect but do have to ask.

As I said, the above notwithstanding, the info you present here is much appreciated!

[NewBie]

Newbie, this is a GREAT thread and yes, someone please sticky it!


One question, and please don't be offended by it... I'm just performing the requisite due diligence... you seem to speak kindly of Duracell relative to the other brands... do you have any sort of affiliation with the Duracell folks? I mean no disrespect but do have to ask.

As I said, the above notwithstanding, the info you present here is much appreciated!

Zero connections with Duracell, Energizer, Streamlight, Pelican, MagLite, Browning, Panasonic, Energizer, or anything flashlight/battery that I know of. I do know a few folks a tiny bit, like georges80, dat2zip, McGizmo, Beamhead, and such, by virtue of being an involved CPFer. I've only met one CPFer once that I know of, and that was to purchase an ARC AA from him. I don't even work in the consumer products industry, nor do I own stock in any of those companies.

I can however appreciate a few extra things Duracell did in their construction, a while back, when I was opening cells and looking at their internals and crimping/welding marks to confirm x was made by x. An example is a separate PTC crimped assembly, where one can do a much better job of crimping the PTC to just the right amount (materials stackup). All the other cells I looked at are crimped as a whole stack of things, where it is alot harder to control the precise crimp on the PTC. The PTC pays a very vital role, and is quite sensitive to crimp pressures.

BTW, SureFire 123 cells are in fact most definitely NOT made the same as the Duracell. Completely different critters. But the SF cells are absolutely mechanically identical between Streamlight, Browning, Panasonic, Energizer- including internal marks from the machines.

Sanyo cells from China are also different, like the Duracells, than the rest of the cells that I've seen so far.

[bwaites]

Newbie,

Do I understand correctly then that from what you have seen, Duracell, Surefire, Streamlight, Browning, Panasonic, and Energizers marked "Made in the USA" are all made on the same machines, and therefore in the same factory?

Or are you saying that Duracell and Surefires are the same, with the same type crimp, and that the rest are identical to each other?

Bill

[NewBie]

Newbie,

Do I understand correctly then that from what you have seen, Duracell, Surefire, Streamlight, Browning, Panasonic, and Energizers marked "Made in the USA" are all made on the same machines, and therefore in the same factory?

Or are you saying that Duracell and Surefires are the same, with the same type crimp, and that the rest are identical to each other?

Bill

Absolutely not!


Duracell is one type.


Sanyo is another.


Surefire, Streamlight, Browning, Panasonic, and Energizers are the same.


There is nothing similar between the Duracell and Surefire cells in their construction.


Surefire, Streamlight, Browning, Panasonic, and Energizers all have identical construction.


.

[Knight Lights]

This is what confused me, from post #50:

"BTW, SureFire 123 cells are in fact most definitely made the same as the Duracell. Completely different critters. But the SF cells are absolutely mechanically identical between Streamlight, Browning, Panasonic, Energizer- including internal marks from the machines."

OK, so I understand it:

US made Surefires, Streamlights, Brownings, Panasonics, and Energizers are all identical mechanically.

Duracells are different, as are Sanyo's.

Bill

[NewBie]

OK, so I understand it:

US made Surefires, Streamlights, Brownings, Panasonics, and Energizers are all identical mechanically.

Duracells are different, as are Sanyo's.

Bill

Yup. Internally, including machine marks, welds, crimps, wrap, plastic, plastic markings, etc... Surefires, Streamlights, Brownings, Panasonics, and Energizers are all identical to each other.

[cy]

tested all three cell later with ZTS MTB-1. all three cells tested at 80%. I'm not convinced there is 80% energy remaining in cells, but I am convinced all three cells have same remaining capacity.

still using the same three set of matched 80% reading Surefire cells in L6. No way I would use Battery Station primary lithium cells of recent production dates in series.

My theory posted earlier in this thread is: it takes a combination of mismatched primary lithium cells, defective PTC and light accidently being left on to cause failure.

this L6 getting HOT instance was 100% due to operator error of not unscrewing clickie tailcap out.

since then I've changed to LOTC tailcap, instead of clickie. failue mode due to tailcap being accidently left on is all but eliminated.

clickie tailcap lights have increased odds of accidently being switched. could this be why Surefire M6 with LOTC still has a very low reported failure rate? inspite of increased potential for mismatched cells.

I'm also getting ready to change to 2x 17500 protected li-ion cells for L6. As there have been no reported instances of failures yet for li-ion cells in flashlights.

Cy, I wouldn't have imagined the L6 causing an incident. I know the L4 left on can get literally too hot to hold... I've done it once... but the cells had no problems at all. Did your cells show any signs that anything was amiss?

[dragoman]

Wow....lot of information here.....

guess the ZTS tester is a necessity if you are going to use 123s.......off I go to buy one!

Thanks Newbie for all the time and effort....it is appreciated!

dragoman

[McGizmo]

Hi guys,

It may be worth consideration that numerous run time tests have been been made by various CPF members of CR123 lights and in many cases, the light was left in still air with no induced air flow for thermal relief. I have done many myself and observed that the flashlight would get hot, in many cases, above any comfortable level of contact. I have not done Incan run time tests but many many have. I am not aware of any reported events or venting in these tests. These tests will often deplete the batteries well beyond useful output levels.

A constant on runtime test can provide a viable and useful data point for relative comparison to another light or in qualified absolute terms, IMHO. I do not think it mirrors real world operation times when a light is started and stopped often. A constant on runtime test can set the stage for venting or an event if the batteries are not up to the demand placed upon them and in light of recent work and experiments, this should be taken into consideration. However, I have not been convinced by the reported findings that a light left accidentially on or intentionally on through the full course of a battery's discharge is inherently unsafe or cause for undue alarm. If the light gets too hot to hold or dims significantly in use, it would obviously be prudent to cease using it until it cools down and/or the batteries are replaced.

Is a hot flashlight or batteries or a dim light an almost event? I suppose it depends on your definition. With a broad enough view, any light turned on could be an almost event.


********

The other day I accidentally turned on my light. I was able to keep it from exploding by immeditely turning it of; event avoided. I now have adopted a safer approach by never installing batteries in the light.

********

Events and venting can and has happened, don't get me wrong here. However they are no where near a certainty and many lights have warnings as to caution that the light will get hot during operation. Somewhere, somehow, in some isolated instances, some CR123 batteries are being taken past safe use and in some documentated cases not fail safe; or venting or event safe, as the case may be.

What are the precursors to a CR123 failing to function properly? Included in this proper function is the onboard PTC and other protective mechanisisms in place for the battery to shut down under unsafe conditions. Well one precursor is actually using the battery and another precursor is elevated temperature. Are these two precursors to be avoided? I have not seen any evidence yet to support this. There have been some conditions and applications identified though which do appear to be significant precursors to venting and these should be respected and avoided if possible.

Are CR123 batteries inherently safe? No! They have way too much stored energy to qualify. Are CR123 batteries inherently unsafe? I think the answer here is a qualified, no. It is the qualification that needs to be identified and better understood, IMHO.

Are all CR123 batteries manufactured to the same standard and equal in performance and onboard safety and protective devices? Are all CR123 batteries manufactured equaly suited for a particular application?

[NewBie]

Yes, so we have seen events on tape with temperatures right before the venting at 100F, 140F on this latest one, 153F,
161F, 161F, 171.2F- with temperatures running up to 714F.

The "fireball" one that was just made: http://www.molalla.net/~leeper/vent12sn.wmv
The 714F one was here: http://www.molalla.net/~leeper/vent6_6.wmv
171.2F one: http://www.molalla.net/~leeper/ven10sm.wmv
Another 161F that goes above 700F: http://www.molalla.net/~leeper/vent5_1.wmv
A 153.6F one: http://www.molalla.net/~leeper/minorv~1.wmv
The 100F one: http://www.molalla.net/~leeper/100degv.wmv

I've seen many more of these, which I do not have on tape. These are still by and far not the worst events I have seen- only what I happened to catch on tape, once someone asked for a video of some of the failures. I've missed more dramatic ones. I've tested a few more of these than I'd originally planned so I could demonstrate what can happen.


I'd really like folks to go back to the first post in this thread, where I talk about the safety factors and some of the things folks can do. Take the time to read it, it is worth it. I'm doing this because I see an actual possible danger, and there are things that can be done which greatly reduce the risk.

There is a label on every CR123A cell that I've seen in the US, and it would be good for folks to take a pause and read it. I believe it is there by law, for a reason. It is not there for sh*ts, grins, and giggles.

Please use these cells wisely, and treat them carefully- These ain't your grandma's carbon-zinc cells.

I have further testing to do, to learn more about the events leading up to these events, such as watching currents and voltages across each cell, while watching the temperatures- to glean anything out of this info, so designers might be able to design in countermeasures.

Just to let folks know, I am currently building up a data aquisition system to simultaneously measure the voltage across each cell, each cells temperature, and the current present, so that I can get a much better idea of what is going on, and see what we can do to further understand and prevent these sorts of failures. It will help us see what leads up to, and immediately prior to the failures shown in the videos above.

It may take a little time, as I still have parts comming in to build up the K-type thermistor conversion circuitry. That, and I'm working on programming the microcontroller now.

So bear with me, I haven't forgotten, I'm just building/designing/programming new equipment from the ground up from raw parts.

[McGizmo]

Great Newbie!!! Invaluable service you are providing here!! If I can be of any service in machining a test host, let me know. I won't have access to my lathe for a couple weeks but am open to assist in any manner I can.

I believe that statistically, the only way one can be 100% certain of no danger from a CR123 or any battery for that matter is not to use it. What each of us needs to do is identify the level of certainty of safety that we are willing to accept, if less than 100%. We rely on battery manufactures to provide safe batteries and as designers and users, we hope to be aware of the do-s and don't-s involved in use of these batteries.

At least in the US where rampant litigation is supported and encouraged, there may be vital information and knowledge kept from us for understandable reasons. There is also the potential of a knee jerk response that could bring us to a 100% certainty of safety. In the mean time, wonderful folks like Newbie are working towards a better understanding of the nature of these batteries so that we may use them wisely and with less risk and concern.

Thanks Newbie and you others out there!

[NewBie]

A thought just occurred to me, while I'm working on this test rig.

I've been spending time re-familiarizing myself on programming Atmel uC in assembler, and re-learning the complier and how it works.

Some folks have previously offered to help out.

This is one area that could be done by someone else that is currently up to speed on these puppies.

That would allow me to concentrate on the input signal conditioning circuitry, such as building up the K-type thermocouple compensation and scaling circuits, current monitoring, etc.

I have sitting here, the Atmel STK500, populated with the ATMEL ATMEGA16-16PI.

If any programmers would like to help out, and are familiar with these chips, I'm looking for something to read all of the 8 A/D channels as single ended in the 10-bit mode, convert to decimal, and spit out a string where each channel reading is separated by a space, and then each string of data from all the channels are separated by a comma or carriage return/LF. Use of the internal 2.56V refrence is fine, I can scale accordingly with external circuitry. 9600 baud rate would be fine. It doesn't matter if it is in ASM or C, just as long as AVRStudio can handle the code.


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


On a side note, Chronos just had a *LED* Lithium Primary cell not just vent, but actually eject it's internals:

Link here:
http://candlepowerforums.com/vb/show...9&page=2&pp=40


CPF member nakahoshi also had one of his cells get quite hot in his LED light, found here:
http://www.candlepowerforums.com/vb/...24#post1536124