NewBie
*Retired*
2nd run of the 17% discharged cell:
123 Primary Lithium cell info/testing/links
- Thread starter NewBie
- Start date
Help Support Candle Power Flashlight Forum
I've been reading this thread, and the one that started it, for two solid days! You ate my weekend! :-$
I'm impressed by the level of persistance you guys, and especially NewBie, have shown with these issues.
I posted a bit of a theory on the other thread, I'll repeat it here:
---
I have a theory, and it basically boils down to the internal protection in the 123s not tripping due to the heat being conducted away from the protection circuit by the aluminium heatsink that is the body (Aluminium is the 3rd most heat conductive metal, iirc) with the result that the cell can get hot enough to reach venting temperatures in one part of the cell, but not in the other.
Of course, once venting temp is reached in one place, the runaway causes seals to fail, and the whole thing becomes catastropic.
Does the M6 case have a fairly tight fit to the batteries? Does anything else insulate them from the walls (e.g. a battery holder?) [I note that to get these cells to vent it was required to wrap them in tin foil. I also doubt that the positioning of the bulb in this test stand is actually heating anything noticably, as it is remote on a wire.]
Has anyone *ever* had a plastic torch go pop?
I'm impressed by the level of persistance you guys, and especially NewBie, have shown with these issues.
I posted a bit of a theory on the other thread, I'll repeat it here:
---
I have a theory, and it basically boils down to the internal protection in the 123s not tripping due to the heat being conducted away from the protection circuit by the aluminium heatsink that is the body (Aluminium is the 3rd most heat conductive metal, iirc) with the result that the cell can get hot enough to reach venting temperatures in one part of the cell, but not in the other.
Of course, once venting temp is reached in one place, the runaway causes seals to fail, and the whole thing becomes catastropic.
Does the M6 case have a fairly tight fit to the batteries? Does anything else insulate them from the walls (e.g. a battery holder?) [I note that to get these cells to vent it was required to wrap them in tin foil. I also doubt that the positioning of the bulb in this test stand is actually heating anything noticably, as it is remote on a wire.]
Has anyone *ever* had a plastic torch go pop?
NewBie
*Retired*
soapy said:
Yes. I believe it was Paul in Maryland, but I'm not sure if that is the right fella.
If you look, you will note the cells, when the conditions are right, vent when there is a reverse voltage on one. I've also tried an air gap in an aluminum tube, and minimal gap aluminum tube, both cause venting, but I do not see venting when they are in open air. Also remember, that aluminum reflects IR (heat) nicely back to the source. I have not had a chance to try it with a black anodized aluminum tube, inside and out, which would absorb a lot more of the heat radiated from the cell, and re-radiate it into the air.
Anyhow, here is the plot in an LED flashlight with 11% discharged cell, 89% remaining:
Thanks. Being fairly new to the whole high powered flashlight thing, I vaguely know that most of the high end tubes are aluminium. Don't forget that still air is a very good insulator, the exact opposite of aluminium.
Having looked at your breakdowns of the cells, the mechanism for this that I can see would be that heat transmission to the PTC wafer is what counts. In a plastic or insulator torch, the heat would naturally tend to pipe up towards the end cap where the PTC is, letting it get hotter faster than in an aluminium heatsink body, which would allow the cells to lose heat to the walls, rather than slowly conduct up the cell itself, transmitting less heat to the PTC.
Once any part of the cell gets hot enough (and looking at the vented cases where everything is gone, it seems clear to me that violent complete ejections start at the bottom, at the furthest point from the PTC) it goes into thermal runaway, ejecting the contents.
I've a 3 cell AAA Zipka+ that I keep in my trouser pocket 24/7 aside from when it is on my forehead! (I seriously doubt the current drain is nearly enough to be any kind of issue from 4 5mm LEDs though)
Could 2 cells at this current drain be marginally hot, and three don't go pop because they tend to have the same bulb and so less drain per cell? To test 3 or 4 cells for this issue (which is rarely if ever seen) we would need to draw more current by either 50% or 100%. Or it might be that there is some kind of weird mechanism preventing the voltage reversal being an issue in >2 cell devices.
Having looked at your breakdowns of the cells, the mechanism for this that I can see would be that heat transmission to the PTC wafer is what counts. In a plastic or insulator torch, the heat would naturally tend to pipe up towards the end cap where the PTC is, letting it get hotter faster than in an aluminium heatsink body, which would allow the cells to lose heat to the walls, rather than slowly conduct up the cell itself, transmitting less heat to the PTC.
Once any part of the cell gets hot enough (and looking at the vented cases where everything is gone, it seems clear to me that violent complete ejections start at the bottom, at the furthest point from the PTC) it goes into thermal runaway, ejecting the contents.
I've a 3 cell AAA Zipka+ that I keep in my trouser pocket 24/7 aside from when it is on my forehead! (I seriously doubt the current drain is nearly enough to be any kind of issue from 4 5mm LEDs though)
Could 2 cells at this current drain be marginally hot, and three don't go pop because they tend to have the same bulb and so less drain per cell? To test 3 or 4 cells for this issue (which is rarely if ever seen) we would need to draw more current by either 50% or 100%. Or it might be that there is some kind of weird mechanism preventing the voltage reversal being an issue in >2 cell devices.
NewBie
*Retired*
Three cell lithium primary lights like the Surefire 9P have been known to actually rip open the side of the light before. This is a different chemistry than most AAA light batteries, even with AAA lithiums they are different.
Another 11% plot :
Another 11% plot :
Last edited:
NewBie
*Retired*
The same 27LT with a Battery Station cell @ 11% discharge:
What happened here?
http://mjaaskel.unlink.org/IMG_3552.jpg
Few unused cells (Batterystation cr123a 1205) inside Pelican 1010 case...
One has wet looking top with sticky liquid. Storage temperature has been between 10-35 C / 50-100 F.
http://mjaaskel.unlink.org/IMG_3552.jpg
Few unused cells (Batterystation cr123a 1205) inside Pelican 1010 case...
One has wet looking top with sticky liquid. Storage temperature has been between 10-35 C / 50-100 F.
SilverFox
Flashaholic
Hello Rigor,
It looks like the seal did not perform well.
Tom
It looks like the seal did not perform well.
Tom
Damn...
Didn't have balls to lick and taste like old bio-chemists do with liquids...
Might have been only condensated water... Still only 1 wet cell...
Voltage same as others, top of wrapper darker (color transfer from case?)
Didn't have balls to lick and taste like old bio-chemists do with liquids...
Might have been only condensated water... Still only 1 wet cell...
Voltage same as others, top of wrapper darker (color transfer from case?)
NewBie
*Retired*
Thank you goes out to our own infamous Dat2zip from the Sandwich Shoppe (http://theledguy.chainreactionweb.com/) for donating a HD45 for testing purposes. This is quite a valuable test, as it draws current in the range of a Pelican M6 or Streamlight Tasklight 2 or the Streamlight Night Hunter 2. But it's converter cuts back the power draw when the battery voltage drops low.
A warm thank you to cpf member Olephart for purchasing a batch of Lithium Primary 123 cells to continue tests with that brand.
CPF legend, Skunkworks McGizmo, has also provided some additional items, which we will get into at a later date in more detail.
Please give these fellas a warm round of applause for their contributions!
Okay, so here is the first run with the HD45 (btw, I was **very** impressed with this light, well done fellas!).
Things to note, I have rescaled the current in Amps, back down to 1x. You will note that it basically draws the same amount of current as the incandescents that were blowing up the batteries. Also, the wide band of "noise" you see, is where the light has gone into moon mode, and it pulses. The circuit turns off, as the battery voltage falls below a threshold, then the battery voltage recovers, and the circuit turns back on, then repeats, this creates the wide band of "noise" you see.
A warm thank you to cpf member Olephart for purchasing a batch of Lithium Primary 123 cells to continue tests with that brand.
CPF legend, Skunkworks McGizmo, has also provided some additional items, which we will get into at a later date in more detail.
Please give these fellas a warm round of applause for their contributions!
Okay, so here is the first run with the HD45 (btw, I was **very** impressed with this light, well done fellas!).
Things to note, I have rescaled the current in Amps, back down to 1x. You will note that it basically draws the same amount of current as the incandescents that were blowing up the batteries. Also, the wide band of "noise" you see, is where the light has gone into moon mode, and it pulses. The circuit turns off, as the battery voltage falls below a threshold, then the battery voltage recovers, and the circuit turns back on, then repeats, this creates the wide band of "noise" you see.
NewBie
*Retired*
Another run:
NewBie
*Retired*
Next run, at 17% discharge mismatched with a 100% cell, on the HD45 flashlight:
Last edited:
NewBie
*Retired*
Note, the error in the HD45 chart is fixed, the current indicated is the actual current, not the current x2.
Next run results:
Next run results:
Ok, You asked if anyone might see something in the numbers that has not been mentioned. At the risk of exposing my lack of electrical knowledge, here goes.
All of the graphs are starting to look alike for a particular light. The amount and duration of the negative charging that the particular combo can apply to the weak battery appears reasonably predictable.
We also have some data as to the amount and duration of reverse charging required for venting in certain batteries.
Seems simple (to a simpleton). Define the venting parameters and then see what general classes of lights can produce those conditions. The problem:
I think the excellent information provided to date indicates that we need a better definition of the danger zone if we hope to avoid individual testing of every light ever made.
Well, this is where I'm on shakey ground. It looks like an average of 700ma for 5-10 minutes is a point within the danger zone. Where does it go from there?
To find out, could you take partially depleted cells and hook them up to a charger backwards, then, work on the variables until the danger zone is mapped.
I'm thinking a simple xy graph of average ma vs time would do it. I know there are a lot of variables, but I think the worst case for temperature, % depletion and battery type would expose the most risk to users.
With a map of the danger zone, entire classes of lights (i.e. 5 w direct drive, 3w with downboy, etc.) could be profiled to see if they are capable of operating in the danger area. I believe this will result in the most useful information with the least amount of testing.
This isn't a proposal or a criticism of current testing (most excellent). I'm just fishing for comments to see if this is feasable - I have no idea. If it is, maybe we could get some resources together and do it. If not, I'm going to take a nap and dream up something else.
All of the graphs are starting to look alike for a particular light. The amount and duration of the negative charging that the particular combo can apply to the weak battery appears reasonably predictable.
We also have some data as to the amount and duration of reverse charging required for venting in certain batteries.
Seems simple (to a simpleton). Define the venting parameters and then see what general classes of lights can produce those conditions. The problem:
I think the excellent information provided to date indicates that we need a better definition of the danger zone if we hope to avoid individual testing of every light ever made.
Well, this is where I'm on shakey ground. It looks like an average of 700ma for 5-10 minutes is a point within the danger zone. Where does it go from there?
To find out, could you take partially depleted cells and hook them up to a charger backwards, then, work on the variables until the danger zone is mapped.
I'm thinking a simple xy graph of average ma vs time would do it. I know there are a lot of variables, but I think the worst case for temperature, % depletion and battery type would expose the most risk to users.
With a map of the danger zone, entire classes of lights (i.e. 5 w direct drive, 3w with downboy, etc.) could be profiled to see if they are capable of operating in the danger area. I believe this will result in the most useful information with the least amount of testing.
This isn't a proposal or a criticism of current testing (most excellent). I'm just fishing for comments to see if this is feasable - I have no idea. If it is, maybe we could get some resources together and do it. If not, I'm going to take a nap and dream up something else.
NewBie
*Retired*
I've thought about doing just that, I just want to first get through tests with a number of lights, to see if things keep holding true.- "actual flashlights", not some contraption.
NewBie
*Retired*
Well, I got delayed a bit, sorry about the gap in testing, CREE released some new LEDs into volume production that really shake things up, and I wanted to look at them a little. My goodness!
Anyhow, back to the grindstone.
Bumping up the mis-match in the HD45 LED, to a 25% discharged cell (75%) with a 100% cell:
Anyhow, back to the grindstone.
Bumping up the mis-match in the HD45 LED, to a 25% discharged cell (75%) with a 100% cell:
I think we need to know the temperatures at various points in the cell.
I also think the most important factor here is the PTC not getting hot enough to activate (or activate fully). We know the Battery Station batteries always pop under a certain set of conditions:
Aluminium wrapped, one cell at 60%, second at 100%, high current draw. Pop!
What I would like to see is a way to stop these going pop. I'm thinking that a simple test of my theory would be to do a "pop" run with a few wraps of paper around it as an insulator. First try it on the outside, as if this pops it means that the IR reflectance might have something to do with it. It not, we are on to something.
Next, repeat the test with a wrap of paper inside, to slow heat transfer from the batteries to the aluminium heatsink of the body. I am pretty sure this will never pop. It would give anyone who was worried a simple way to cut the risk to almost nothing, for a stock torch, and let you stop killing batteries every day.
I also think the most important factor here is the PTC not getting hot enough to activate (or activate fully). We know the Battery Station batteries always pop under a certain set of conditions:
Aluminium wrapped, one cell at 60%, second at 100%, high current draw. Pop!
What I would like to see is a way to stop these going pop. I'm thinking that a simple test of my theory would be to do a "pop" run with a few wraps of paper around it as an insulator. First try it on the outside, as if this pops it means that the IR reflectance might have something to do with it. It not, we are on to something.
Next, repeat the test with a wrap of paper inside, to slow heat transfer from the batteries to the aluminium heatsink of the body. I am pretty sure this will never pop. It would give anyone who was worried a simple way to cut the risk to almost nothing, for a stock torch, and let you stop killing batteries every day.
NewBie
*Retired*
soapy said:
They pop in a regular PM6 flashlight also.
Air gapped Al foil of 0.1" gap still pops.
Tightly wrapped AL foil still pops.
You can see the temps in the videos, measured at the side of the cell, with thermal paste present to increase thermal transfer from battery case to 40 ga or 42ga K-type thermocouple (chromel (nickel chromium alloy) and alumel (aluminum nickel alloy)).
I've watched the PTC kick off, and still seen pops.
AmondoTech Titanium cells seem to pop easier than the Battery Station cells.
If ambient temps are too high, PTC kicks in early and no pop.
If ambient temps are below 70F, *very* hard to make pop.
Cell mismatch is present in every pop I've monitored, and reverse charging one cell while "warm" seems to be a common thread.
More results:
Last edited:
Similar threads
