Li-mn cells reliable in winter temperatures?
- Thread starter mpteach
- Start date
Help Support Candle Power Flashlight Forum
Marduke
Flashaholic
Keep an eye on this thread. A similar question was asked, but never answered.
js
Flashlight Enthusiast
deleted. I was posting about CR123A type primary cells. Sorry.
Last edited:
StarHalo
Flashaholic
No rechargeables in extreme hot or cold conditionsHey starhalo genius don't buy a hybrid or a full electric vehicle. They're coming out with lithium batteries but their going to be unreliable Detroit will only make them in spring and fall when it's not to cold or warm for them to work /sarcasm.
Last edited:
StarHalo
Flashaholic
Batteries become more unreliable as you approach their operating temperature threshold. Many users here have reported their alkaline cell lights working intermittently/not working at termpatures as modest as 20F. The only first-person report of a Li-Ion cell in cold temperatures that I've seen here noted that after being exposed to 0F temps for a while, the battery would no longer charge beyond 3.9V, indicating internal/lattice damage.
In short, all of this could be avoided by using a primary lithium battery, which works comfortably in temps down to -40F.
From HybridCars.com's FAQ page:
"Honda's specs indicate that its Integrated Motor Assist system will operate as low as 22 degrees below zero Fahrenheit. We have seen reports of a Prius in Barrow, Alaska suffering from a frozen and damaged battery pack at 56 below zero."
In short, all of this could be avoided by using a primary lithium battery, which works comfortably in temps down to -40F.
From HybridCars.com's FAQ page:
"Honda's specs indicate that its Integrated Motor Assist system will operate as low as 22 degrees below zero Fahrenheit. We have seen reports of a Prius in Barrow, Alaska suffering from a frozen and damaged battery pack at 56 below zero."
Jarl
Flashlight Enthusiast
Is that because they tried to charge it below freezing? (~30F). If you do, then you damage the battery, but they should be ok discharging at lower temperatures.
StarHalo
Flashaholic
Original post (update follows):
Update:
Another, separate instance:
Last edited:
mdocod
Flashaholic
I'm going to throw a pure assumption into the mix. Do not take this as fact. This is a Guess.
Maximum discharge rates and voltage "maintenance" of all cell chemistries are hindered in extreme cold. Some are better off than others, I would venture to guess, that the higher performance the cell is at regular operating temps, the less of a hit you are going to take at lower temperatures if you are operating the cell at less than it's maximum rated discharge. Remember, many of the examples of li-ion failing in cold are small cells in high performance applications. In the same application, a LiMn cell would not even be stretching it's legs at room temp, so it would stand to reason, that there is a good chance that a LiMn cell would work better than a LiCo in the same application at cold temperatures.
I'll try to illustrate an example of why I'm thinking this:
Lets say we have a protected RCR123 cell. We know that on a GOOD day, at room temp, it will deliver ~1.2A reasonably well for ~20-30 minutes depending on the true capacity of the cell in question (varies by brand). We know that it will be driven hard at this load, pretty much at it's maximum recommended safe discharge rate, in some cases, even slightly exceeding it. We know that it we move up with 1.5A, or 2A, the cell really starts to fall on it's face, and fall far below label capacity. So it would stand to reason, that moving towards cold temperatures, would reduce the cells performance when operated at the edge of it's performance capabilities pretty dramatically.
But... assuming a LiMn RCR123 cell, now we are dealing with a cell that can run a 1.2A with it's hands tied behind its back. We could move up to 2 amps, 3 amps, even more, and the cell holds reasonably close to it's label capacity and delivers reasonable voltage into these extreme 2-6C and higher loads. It would stand to reason, that if one were testing this cell at 1.2A, that it would maintain a more impressive performance at reduced temperatures than a protected LiCo cell due to it's inherent lower internal resistance.
If I get the chance.... I'll try to do a test to determine if my hypothesis is correct.
Eric
Maximum discharge rates and voltage "maintenance" of all cell chemistries are hindered in extreme cold. Some are better off than others, I would venture to guess, that the higher performance the cell is at regular operating temps, the less of a hit you are going to take at lower temperatures if you are operating the cell at less than it's maximum rated discharge. Remember, many of the examples of li-ion failing in cold are small cells in high performance applications. In the same application, a LiMn cell would not even be stretching it's legs at room temp, so it would stand to reason, that there is a good chance that a LiMn cell would work better than a LiCo in the same application at cold temperatures.
I'll try to illustrate an example of why I'm thinking this:
Lets say we have a protected RCR123 cell. We know that on a GOOD day, at room temp, it will deliver ~1.2A reasonably well for ~20-30 minutes depending on the true capacity of the cell in question (varies by brand). We know that it will be driven hard at this load, pretty much at it's maximum recommended safe discharge rate, in some cases, even slightly exceeding it. We know that it we move up with 1.5A, or 2A, the cell really starts to fall on it's face, and fall far below label capacity. So it would stand to reason, that moving towards cold temperatures, would reduce the cells performance when operated at the edge of it's performance capabilities pretty dramatically.
But... assuming a LiMn RCR123 cell, now we are dealing with a cell that can run a 1.2A with it's hands tied behind its back. We could move up to 2 amps, 3 amps, even more, and the cell holds reasonably close to it's label capacity and delivers reasonable voltage into these extreme 2-6C and higher loads. It would stand to reason, that if one were testing this cell at 1.2A, that it would maintain a more impressive performance at reduced temperatures than a protected LiCo cell due to it's inherent lower internal resistance.
If I get the chance.... I'll try to do a test to determine if my hypothesis is correct.
Eric
Looking forward to hearing any info on this!
Mr Happy
Flashlight Enthusiast
From a chemical point of view there are some things we can observe about the effect of temperature.
Batteries are devices that turn chemical reactions into electricity, and chemical reactions always slow down when the temperature decreases. Now the maximum rate you can draw electricity from a cell may not be limited by a chemical reaction rate at room temperature -- there may be physical processes like diffusion that become limiting. But what we can say is that if the chemical reaction rate inside a cell slows down too much, then it will start to limit how well the cell works.
Now the effect of temperature on a chemical reaction varies with different reactions. Some reactions are much more sensitive to temperature than others. So if you have a cell whose reactions are insensitive to temperature, then when you cool it down it will work better than one whose reactions are more sensitive to temperature.
So in essence what mdocod proposes is correct. If you had a cell where the chemical reactions go so fast at room temperature that they effectively provide no limit to the output current, then cooling that cell down may give lots of room before the chemical reactions slow down enough to matter.
It all depends on specific cell chemistries and cell constructions, and every kind of battery is different.
Batteries are devices that turn chemical reactions into electricity, and chemical reactions always slow down when the temperature decreases. Now the maximum rate you can draw electricity from a cell may not be limited by a chemical reaction rate at room temperature -- there may be physical processes like diffusion that become limiting. But what we can say is that if the chemical reaction rate inside a cell slows down too much, then it will start to limit how well the cell works.
Now the effect of temperature on a chemical reaction varies with different reactions. Some reactions are much more sensitive to temperature than others. So if you have a cell whose reactions are insensitive to temperature, then when you cool it down it will work better than one whose reactions are more sensitive to temperature.
So in essence what mdocod proposes is correct. If you had a cell where the chemical reactions go so fast at room temperature that they effectively provide no limit to the output current, then cooling that cell down may give lots of room before the chemical reactions slow down enough to matter.
It all depends on specific cell chemistries and cell constructions, and every kind of battery is different.
mdocod
Flashaholic
never-mind, need to sleep...
Last edited:
baterija
Flashlight Enthusiast
- Joined
- Feb 7, 2008
- Messages
- 1,053
Some interesting discharge graphs at -20C with a 18650 IMR here (link to pdf file). They don't have the same "Discharge Temperature Characteristics" graph as the ICR cell (link to pdf) specs though.
It doesn't look like the IMR does too badly at -20C though. Maybe 10% lower voltage at 2.5A draw with the same Ah capacity. The ICR graph in comparison shows about 65% of capacity at 1.2A draw and can't handle 2A draw at -20C.
Both data sheets are showing -20C (-4F) as the minimum rated discharging temperature.
It doesn't look like the IMR does too badly at -20C though. Maybe 10% lower voltage at 2.5A draw with the same Ah capacity. The ICR graph in comparison shows about 65% of capacity at 1.2A draw and can't handle 2A draw at -20C.
Both data sheets are showing -20C (-4F) as the minimum rated discharging temperature.
modamag
Flashlight Enthusiast
-20C is typically the limit of most cells.
When you below -30C, just leaving them overnight in such condition will permanently damage the cells, regardless of chemistry LiMn / LiCo / LiFePO4.
When you below -30C, just leaving them overnight in such condition will permanently damage the cells, regardless of chemistry LiMn / LiCo / LiFePO4.
redsfairlane
Newly Enlightened
When any product is being shipped, by truck, train, or just mail carrier, they may be exposed to 30 below (or more) for a day or more easily I think. Batteries I have ordered this time of year may sit in my mail box waiting all day just for me to get home.
So I have to ask, are you sure that just being exposed to such a temperature would damage the cell, or am I misunderstanding the comment?
I have had batteries delivered this time of year, sitting in the cold, I can't say whether I have noticed any performance problems, should I be looking?
mdocod
Flashaholic
The effects of "permanent" damage from cold are more likely to take place on a fully charged cell, from my understanding, since most li-ion cells are shipped ~50% charge, most of this type of damage is avoided in most scenarios.
mdocod
Flashaholic
I did a little test for yall:
--------------------------------------
Cold Tests were performed as follows:
1. Cells charged to ~4.1-4.2V range.
2. 1 of each cell type placed in freezer for ~1 hour or more.
3. SureFire 6P, loaded with a 3.7V LF LED module, and 1 dummy "CR123" cell is placed in a bowl of ice for at least 10 minutes prior to test beginning to chill the flashlight body. (It's aluminum, it immediately starts to melt the ice when placed on it, and immediately gets VERY cold, hehe) CR123 "dummy" cell is loaded at the front, cells to be tested are at the rear, to isolate as much of the heat generated by the LED from the cells being tested.
4. Cell is removed from freezer, open circuit voltage measurement taken, then installed in flashlight, where a tailcap current reading is taken to gage fully charged performance. (this is the "before")
5. Flashlight is turned on, and ran for 15 minutes, flashlight stays on ice throughout the duration of the "run" test". (this is the "after")
6. Cell is retested (open circuit voltage and tailcap current reading).
Warm Testing is basically the same, except without any of the cold
----
Originally, I did a warm test, followed by a cold test, and then decided to repeat the cold test to confirm repeatability and accuracy of the test method. On the first cold tests, the cells were only in the freezer about an hour, on the second testing that I did today, they were in the freezer for about 6 hours prior to testing. Results were about the same.
----
UF cells are relatively new protected type, they were sent to me for free with a purchase of some AW protected cells in the marketplace a couple months ago, and I haven't really used them, and the person who had them before me didn't use em much either. One of them has a finicky protection circuit, but it did not act up during these tests.
AW RCR123s are his black label protected type, I'v had them for years, they are quite old but still work reasonably well.
IMR16340s are only a few weeks old, probably less than a dozen cycles, I've only "deep" discharged them to ~3V for testing once or twice.
--------------------------------------
UF RCR123 LiCo warm:
Before: 4.13V, 0.95A
After 3.83V, 0.72A
UF RCR123 LiCo cold:
Before: 4.13V, 0.57A
After: 3.86V, 0.45A
UF: RCR123 LiCo cold
Before: 4.15V 0.57A
After: 3.87V, 0.46A
-------------------------
AW RCR123 LiCo warm:
Before: 4.12V, 0.9A
After: 3.80V, 0.66A
AW RCR123 LiCo cold:
Before: 4.13V, 0.44A
After: 3.86V, 0.40A
AW RCR123 LiCo cold:
Before: 4.17V, 0.42A
After: 3.87V, 0.41A
------------------------
AW IMR16340 warm:
Before: 4.15V 1.07A
After: 3.78V 0.7A
AW IMR16340 cold:
Before: 4.15V, 0.60A
After: 3.81V, 0.49A
AW IMR16340 cold:
Before: 4.12V 0.54A
After: 3.79V, 0.44A
-------------------------
Semi-Conclusion:
The LiMn cells appear to be much better warm, but seem to take a hit in the cold that knocks em right down to about the performance of a LiCo into this setup. This particular test has shown my hypothesis to be incorrect.
IMO, further testing is required: I have another test in mind that may be more revealing I'd like to do a similar test, but with Emoli 18650s vs AW protected LiCo into an incan setup.
More to come later
--------------------------------------
Cold Tests were performed as follows:
1. Cells charged to ~4.1-4.2V range.
2. 1 of each cell type placed in freezer for ~1 hour or more.
3. SureFire 6P, loaded with a 3.7V LF LED module, and 1 dummy "CR123" cell is placed in a bowl of ice for at least 10 minutes prior to test beginning to chill the flashlight body. (It's aluminum, it immediately starts to melt the ice when placed on it, and immediately gets VERY cold, hehe) CR123 "dummy" cell is loaded at the front, cells to be tested are at the rear, to isolate as much of the heat generated by the LED from the cells being tested.
4. Cell is removed from freezer, open circuit voltage measurement taken, then installed in flashlight, where a tailcap current reading is taken to gage fully charged performance. (this is the "before")
5. Flashlight is turned on, and ran for 15 minutes, flashlight stays on ice throughout the duration of the "run" test". (this is the "after")
6. Cell is retested (open circuit voltage and tailcap current reading).
Warm Testing is basically the same, except without any of the cold
----
Originally, I did a warm test, followed by a cold test, and then decided to repeat the cold test to confirm repeatability and accuracy of the test method. On the first cold tests, the cells were only in the freezer about an hour, on the second testing that I did today, they were in the freezer for about 6 hours prior to testing. Results were about the same.
----
UF cells are relatively new protected type, they were sent to me for free with a purchase of some AW protected cells in the marketplace a couple months ago, and I haven't really used them, and the person who had them before me didn't use em much either. One of them has a finicky protection circuit, but it did not act up during these tests.
AW RCR123s are his black label protected type, I'v had them for years, they are quite old but still work reasonably well.
IMR16340s are only a few weeks old, probably less than a dozen cycles, I've only "deep" discharged them to ~3V for testing once or twice.
--------------------------------------
UF RCR123 LiCo warm:
Before: 4.13V, 0.95A
After 3.83V, 0.72A
UF RCR123 LiCo cold:
Before: 4.13V, 0.57A
After: 3.86V, 0.45A
UF: RCR123 LiCo cold
Before: 4.15V 0.57A
After: 3.87V, 0.46A
-------------------------
AW RCR123 LiCo warm:
Before: 4.12V, 0.9A
After: 3.80V, 0.66A
AW RCR123 LiCo cold:
Before: 4.13V, 0.44A
After: 3.86V, 0.40A
AW RCR123 LiCo cold:
Before: 4.17V, 0.42A
After: 3.87V, 0.41A
------------------------
AW IMR16340 warm:
Before: 4.15V 1.07A
After: 3.78V 0.7A
AW IMR16340 cold:
Before: 4.15V, 0.60A
After: 3.81V, 0.49A
AW IMR16340 cold:
Before: 4.12V 0.54A
After: 3.79V, 0.44A
-------------------------
Semi-Conclusion:
The LiMn cells appear to be much better warm, but seem to take a hit in the cold that knocks em right down to about the performance of a LiCo into this setup. This particular test has shown my hypothesis to be incorrect.
IMO, further testing is required: I have another test in mind that may be more revealing
I'd like to do a similar test, but with Emoli 18650s vs AW protected LiCo into an incan setup. More to come later
redsfairlane
Newly Enlightened
Iv'e got a 17670 in an old lux-1 in the car, it's been there all winter, I'll try to give it a test this weekend and get back to you.
(one morning in Dec. was 36 below, another 34, cell was charged last fall.)
(one morning in Dec. was 36 below, another 34, cell was charged last fall.)
