Lithium sucks in cold weather!!

[neptronix]

So.. i had noticed lowered top speeds and worse acceleration out here during the winter with my RC lipo.. many people say that RC lipo performs poorly in the cold.. i had to find out exactly how bad it was..

Took 1 of my funky Turnigy 20C 5AH 5S packs with a weak cell #5 ( bad from the factory ) with about 1 year of life on it and a total of 10 light cycles on 'er..

The pack was charged to the nominal storage voltage ( 3.85v ) before the tests.
The load was 3.85A, just around a 0.75C load.. using two 50W halogen in series for a total of 100w ( which is voltage dependent, so the real power draw was more like 70-73w )

First test.. room temp at 65deg / 18.3C.
Average internal resistance, about 2.5 mOhm, not so bad. This pack was about 2mOhm when i first got it..

Stuck it in the fridge ( 40F ) for 3 hours, and it only dropped down to 51deg / 10.5C..
Average internal resistance, about 7mOhm, wow, that was a big jump..
You can see the sag has became much worse..

Got impatient and gave it 30 mins. in a 5F freezer came out @ 33F / 0.5C..
Average IR is 10mOhm, if you don't consider that cell #5 is starting to act up here.

Another 30 mins. in the freezer, now down to 25F / -5C..
Average IR is ~15mOhm, not considering that cell #5 is really suffering again.

Since IR ( in mOhms ) is a measure of internal resistance, and an indicator of how much voltage sag & heating will occur from a given load, we can sorta think of the battery as losing it's C rate according to temperature..

If 2.5mOhm is a 20C cell..
7 mOhm is more like a 7C cell..
10mOhm is more like a 5C cell..
15mOhm is more like a 3C cell..



Good thing i run a lot of amp hours. If the sag jumped that high on 0.75C, just imagine it at 2C-10C..


If anyone else here has some battery temperature testing for other commonly used batteries, i'd like to see it.

 

[parabellum]

I supose that V jump down is on load connection and jump up is on load disconnection. Have you tried to let them connected for a wile? Bigger sag is quicker heating, I wonder how long would they need to heat up and recover (in freezer conditions on different C rates).

 

[Lessss]

All batteries suck in extreme cold.

 

[neptronix]

I supose that V jump down is on load connection and jump up is on load disconnection. Have you tried to let them connected for a wile? Bigger sag is quicker heating, I wonder how long would they need to heat up and recover (in freezer conditions on different C rates).

Yes, exactly. I connected and disconnected the load to observe how long it would take to spring back to the normal voltage.
The higher the internal resistance, the longer this takes.

You have an interesting idea about the heating increasing the IR, i was thinking about doing that kind of test, but didn't have a spare cell around to try it out right now. I don't have the iCharger temp sensor either.

Higher IR would mean more cell heating though. I'd imagine it would help quite a bit to beat the crap outta your cells in winter..

 

[Doctorbass]

Hey, Neptronix, did you have read my post about my E-trike vs snow upgrade ?

I am using the A123 cells in 4p configuration and i got really good result after i warmed them and protected them with blanket.

I know that the lifepo4 3.3V chemistry work better in cold than any of the 3.7V chemistry.

Yo can warm them without any danger or fear to damage them or reduce their spec life...

The killacycle warm their A123 battery to 70 celsius before each ru and they made that for years with the same pack and theyr score still improved over the time!

There is NO accessible battery that work equal or better in cold condition.. so for me the best option remain the chemistry that you can warm without any issue ... LIFEPO4

A123 post very good spec on their cells for cold conditions and warming option!.. after all... The GM Volt got really close to use have them. ( the 20Ah)

Keep your Lipo for summer time! They are lazy for winter

Doc

 

[parabellum]

i was thinking about doing that kind of test, but didn't have a spare cell around to try it out right now.

I would try it myself, have few spare LiPos, unfortunately my Icharger is died.

 

[Doctorbass]

i was thinking about doing that kind of test, but didn't have a spare cell around to try it out right now.

I would try it myself, have few spare LiPos, unfortunately my Icharger is died.

There is a good post about awsome deal on the 3010B cahrger at 95usd on ebay...

just in case...

Doc

 

[neptronix]

Doctorbass, i just saw your thread a second ago.. .. thank you for your report.

4p, that is about 11AH in A123, and you are pulling around 100A? so 9C-10C?
You said that it delivers full AH only when you heat it. I am surprised that it doesn't heat itself.
How much AH were you getting before the 65C / 150F heat was added?

BTW, lipos can take some heating as well just fine.. 100F-125F.. they're used to that in RC applications when the user is running them at super high C.

Nobody has experimented with heating up RC Lipo, but my tests prove that they could definitely benefit.. :lol:
A controller next to the lipos would be good too.. only in winter of course.. :wink:

 

[Doctorbass]

Doctorbass, i just saw your thread a second ago.. .. thank you for your report.

4p, that is about 11AH in A123, and you are pulling around 100A? so 9C-10C?
You said that it delivers full AH only when you heat it. I am surprised that it doesn't heat itself.
How much AH were you getting before the 65C / 150F heat was added?

BTW, lipos can take some heating as well just fine.. 100F-125F.. they're used to that in RC applications when the user is running them at super high C.

Nobody has experimented with heating up RC Lipo, but my tests prove that they could definitely benefit.. :lol:
A controller next to the lipos would be good too.. only in winter of course.. :wink:

These celsl was alot abused in teh past ( i have this pack since 2008-2009) and i used it for my speed record. I have the old V2.5 Gary BMS on it and i just use the LVC function ( i did not repaired the carbonized mosfet section :lol: ) and i bulk cahrge it at 1500W...

btw it's a 9.2Ah not 11 Ah ( 4 x 2.3 = 9.2 )

and in the best summer condition at 25 celsius i get 8.5Ah.. ( i believe that some 4p cells block are low..)

at 65 celsius warmed conditions i got 8.4Ah so it's so it's nearly 1% variatin that ALSO can depend on the other component conditions at low temp.. ex the C-A calibration might be off a little bit at -25C... and also the controller shunt that benefit of coller heatsink might be a bit off ...

So that 8.5 and 8.4Ah can not acurately compared with the E-trike setup... a lab test could give better result with a dymmy load, constant current control and same C-A temperature... ( the op amp and resistor value would not drift)

BTW.. at 8kW.. and ... -25 celsius the 5304 motor is awsome !!! it stay cold!!!! and offer remarkable performances!


Doc

 

[USAEBIKECO]

The pack I use is lithium manganese chemistry. I wrap my battery in a towel and have not noticed any performance difference in the freezing cold. I ride for about an hour and noticed no performance degrade at all.

 

[Ykick]

I've been playing with a loose cell in the freezer but I don't have the logging software installed or any graphs. What I learned is that at 5F RC Lipo sags like breasts on a 90 year old granny. However, at reduced current levels the capacity remains roughly the same and the cell doesn't seem to be damaged dropping into 2-2.5V range. Your LVC may be set too high in order to be used with cold Lipo but the capacity is still there if you don't mind pulling them down below usual bottom voltages.

Pulsing the load does seem to induce enough internal heating so that after some initermittent use the sag isn't quite as severe. Like I said, they will deliver majority of capacity but only at much reduced C rate.

It seems the main danger would be from charging a sub freezing Lipo cell because as it warms up, the voltage may increase well above safe upper voltage for room temperature Lipo.

I simply double my usual capacity when I'll be parking in cold temps. Seems to be working and I only recharge indoors so no chance of getting into trouble there. Based on all that, 10-15F is where I draw the line for RC Lipo on my commuter.

 

[shock]

My first 3 test runs on a 12c3p lipo config, 20*, 45*, and 35* (F) Yeilded interesting results. Top speed was 4 mph slower @ 20* vs 45*. Pack capacity also seemed to diminish the colder the temps were. It's too bad, because you can dump some serious current into a hub and controller without getting them "smoking hot"

 

[Spacey]

I'm noticing a larger voltage sag as well at -3C but end voltage after using 8Ah was about the same. Have set up my Ca for amp control on throttle so I can ease off a bit when it's cold. 16 Cell 17.4Ah 30C Lipo about 300 to 400 cycles charged to 66v and end voltage at rest always around 61.5v

 

[jana]

I've been playing with a loose cell in the freezer but I don't have the logging software installed or any graphs. What I learned is that at 5F RC Lipo sags like breasts on a 90 year old granny. However, at reduced current levels the capacity remains roughly the same and the cell doesn't seem to be damaged dropping into 2-2.5V range. Your LVC may be set too high in order to be used with cold Lipo but the capacity is still there if you don't mind pulling them down below usual bottom voltages.

Pulsing the load does seem to induce enough internal heating so that after some initermittent use the sag isn't quite as severe. Like I said, they will deliver majority of capacity but only at much reduced C rate.

It seems the main danger would be from charging a sub freezing Lipo cell because as it warms up, the voltage may increase well above safe upper voltage for room temperature Lipo.

I simply double my usual capacity when I'll be parking in cold temps. Seems to be working and I only recharge indoors so no chance of getting into trouble there. Based on all that, 10-15F is where I draw the line for RC Lipo on my commuter.

So if a cell was charged to 4.2 volts in the cold and then brought to room temperature - the voltage will increase to dangerous levels?
What if the cell was cold, but fairly good isolated. Is it possible that charging would cause temperature to rise and voltage to stabilize at 4.2 volts?
That would be a dream situation for my "plug and forget" charger - batteries always full and temperate

 

[parabellum]

What if the cell was cold, but fairly good isolated. Is it possible that charging would cause temperature to rise and voltage to stabilize at 4.2 volts?

LiPo do not heat while charging in rated current ranges, they actually cool down while charging.

 

[liveforphysics]

All batteries, including A123, lead, NiCd etc have badly diminished performance in cold weather.

Some cells have internal resistance go up by 10x at 0degC.


It's a very noticeable performance hit to have a freezing cold pack. All the EV mfg's have to deal with this issue, and none of the solutions are cheap or easy.


Your options are either to heat it before using, or start with the best you can get and deal with it in it's reduced performance state. For example, my nano-tech pack at 0degC will still outperform a room temperature A123 pack of similar size by a healthy margin. So, while the freezing does hurt my packs performance, it's still good enough to get the job done.

 

[neptronix]

^-- another reason why overspeccing your pack is a good thing!

Anyhow, it's that time again to discuss how poorly our batteries will perform soon.

http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=19740

Found this on hobbyking; could be helpful if you have a suitable DC-DC or seperate auxillary pack.

 

[migueralliart]

So let me get this straight;

You have demonstrated that with lower temperature the IR goes up and therefore more voltage sag etc.

I've read posts about diminished capacity. I wonder if the diminished capacity is due to the increased IR or some other chemistry related issue.

 

[Ykick]

So let me get this straight;

You have demonstrated that with lower temperature the IR goes up and therefore more voltage sag etc.

I've read posts about diminished capacity. I wonder if the diminished capacity is due to the increased IR or some other chemistry related issue.

My Freezer method of testing a loose 20C Turnigy last year indicated it was actually able to deliver near normal temp capacity AS LONG AS current draw (C discharge rate) was reduced significantly.

LVC kicking in is what will limit cell capacity more so than the actual capacity of the cell if asked to deliver lower than normal current while at cold 5-15F temps.

In other words, back off your controller amps or go extremely easy on the throttle. You will likely find close to normal warm temp capacity if you do.

All that aside, 20F starting temp is about my personal limit for using RC Lipo. Keeping your pack indoors before/after use will allow for small amount internal warming and much improved cold weather performance.

 

[neptronix]

Voltage sag can imitate diminished capacity, IE the 'cliff' of discharge dives even harder, and under load, a battery sagging down to 3.5v/cell might otherwise sag down to 3.6v/cell during more regular temps, giving you the impression that the battery is spent.

 

[migueralliart]

So let me get this straight;

You have demonstrated that with lower temperature the IR goes up and therefore more voltage sag etc.

I've read posts about diminished capacity. I wonder if the diminished capacity is due to the increased IR or some other chemistry related issue.

My Freezer method of testing a loose 20C Turnigy last year indicated it was actually able to deliver near normal temp capacity AS LONG AS current draw (C discharge rate) was reduced significantly.

LVC kicking in is what will limit cell capacity more so than the actual capacity of the cell if asked to deliver lower than normal current while at cold 5-15F temps.

In other words, back off your controller amps or go extremely easy on the throttle. You will likely find close to normal warm temp capacity if you do.

All that aside, 20F starting temp is about my personal limit for using RC Lipo. Keeping your pack indoors before/after use will allow for small amount internal warming and much improved cold weather performance.

Wow I was just thinking of that. Reducing current in the controller to accommodate for temperature changes.

What neptronix says makes sense too. I believe a cell that has a higher IR takes longer to recuperate to unloaded voltage so a lot of people hit LVC and it takes a while to normalize.

 

[neptronix]

What neptronix says makes sense too. I believe a cell that has a higher IR takes longer to recuperate to unloaded voltage so a lot of people hit LVC and it takes a while to normalize.

Yeah - there is that recuperation period after a heavy load.
The higher the internal resistance, and the higher the load, the longer the recuperation period usually is.

I've demonstrated an extreme case of this with an extremely saggy old battery ( putting a ~0.33C load on a 1C battery );

http://neptronix.org/posts/dont_buy_unknown_chinese_batteries.html

Our RC Lipos won't be so dramatic of course, unless you've got like 600-1000 cycles on them somehow :lol:

 

[Ykick]

Neptronix defintely provides the proper text and graphs to the behavior I'm trying to relate.

I was quite surprised to see my test cell plunge down to around 2.5V or so delivering 1C (5A) while at 5F. Interestingly, the cell actually delivered near full capacity at that ridiculously sagged voltage. Didn't seem to suffer effects of damaging heat that would likely happen at room temps when pressed to that extreme.

The wife kept asking what I was doing with the fridge and fire extinguisher...

 

[ZOMGVTEK]

I attempted to take my bike out for a spin earlier today. The bike was sitting in the shed, just over 30ºF and 45% SOC. I couldn't believe how slow it was. I could only manage 80A out of the 24S 20C 10AH Turnigy pack, stalling the motor. It wouldn't do a power wheelie or give any hint of tire spin so I put it back. Really disappointing stuff.

 

[EdwardNY]

I noticed that with the temps in the 40-50deg range that my bike does not seem as fast. Normally I can hit 29MPh, now it seems I am only able to hit 27MPH.

Would 70deg vs 50deg really make that much of a difference? It is not that cold out or freezing. I am also running a 12s4p pack, so it should have plenty of amps available.