Tesla Model S 18650 Cell Test Data

[Longshot]

it appears the cell has not been damaged but we dont know for sure. it is common for a cell to recover but it is still recommended the cell be scrapped. you cant tell just by looking at capacity.

I dont want that one...

 

[katou]

What are you using to test these cells? It certainly produces good results!

Katou

 

[okashira]

What are you using to test these cells? It certainly produces good results!

Katou

I was using this: https://www.tindie.com/products/Kaktus/mightywatt-kit-70-watt-electronic-load-for-arduino/
which was an impulse buy.
It has it's quirks but it works well, and I calibrated it with my multimeter so it's nice and accurate.
It can datalog up to ~10 S/s on current and voltage (4-wire sense)

Now I have the Chinese tester which is much easier to use since it can charge and discharge and cycle all in one.
I setup a supercharge last night using 9 steps:

Unfortunately I forgot to save the data... I setup a thermocouple and peak cell temperature was only 99°F IIRC. No airflow at all and tape on the cell.
^^^ this one's for you flathill

Kind of cool getting 0%-60% charge in 28 minutes.
And I was pretty conservative with my 9 step supercharge. I made sure current never exceeded what Tesla does. Normally takes hours since I use 700mA for safety. See the video and a shot of the procedure I programmed is in there somewhere.
Video here: https://www.dropbox.com/s/01pkdqcoofpa916/2015-04-08%2002.14.20.mp4?dl=0

I'll do it again to capture temp with the actual data.

 

[okashira]

Scratch that, did a supercharge again, this time started at a slightly lower SOC, and max temp was only 94°F. Same starting temp.
I guess the cell hadn't been fully "broken in," I may have been wrong about break in procedure, you might need higher current and to heat the cell up a bit before capacity picks up and DCIR goes down

Data and Video:
This time video starts before I start charge.

https://www.youtube.com/watch?v=vFrTUlYIQww (turn off audio, nothing there)

 

[999zip999]

Wasn't there a lifpo4 cell that ask for a high initial charge as a wake up, that I wouldn't praint because I can't remember ( 4.2 ). Because it was so high for that chemistry. Maybe it had a different electrolyte or anode. Meaning don't all cells have a different favor.
Maybe different cells enjoy different charge profile's.

 

[katou]

Which Chinese tester is that? I checked out the Arduino, that's pretty slick.

Katou

 

[burningwings]

Wow ! What fun , some REAL TEST data . Overall made me feel good about my newly made 18s 6p 18650pf battery. The differences I see between the Tesla cells and pf's are mostly explained by the 3.4 ah and 2.9 ah ratings , of the cells . Panasonic cells are good .

 

[riba2233]

Wow ! What fun , some REAL TEST data . Overall made me feel good about my newly made 18s 6p 18650pf battery. The differences I see between the Tesla cells and pf's are mostly explained by the 3.4 ah and 2.9 ah ratings , of the cells . Panasonic cells are good .

It's actually 3.2 Ah, not 3.4 Ah.

 

[okashira]

Ive been thinking about supercharging and the poor cycle performance I got out of the model s cell with what I thought was a pretty conservative "super" charge rate.

I bet they use pulse charging at low SOC.

Instead of charging at 5A at low SOC, they do 10A for 200ms, off for 200ms, 10A for 200ms, etc. The reasoning:
1.) There are a bunch of journal articles on using pulse charging and how it's better for fast charging li-ion
2.) My theory is that pulse charging will heat up the cell more. This is needed to ensure the cell can take a continuing fast charge. Higher temperature = faster Li diffusion through the anode

Model S cells, esp the B pack cells have always shown an extraordinary response to higher temperatures. Their DCIR goes down by like 40% or more.
Perhaps Panasonic has a better seperater material such that the micro pores don't close up at higher temps like other mfg cells do.
Samsung cells - not nearly as much.

 

[neptronix]

Thanks for the data. Yeah, this cell is about as saggy as i thought it would be under high load. It's no surprise for a high energy density cell.

I'm sure that the car cruises at a nice 0.33C-0.66C rate, which is pretty gentle on the batteries. It's acceleration and hill climbs where they start shedding decent amounts of their internal energy as heat. Explains the comprehensive liquid cooling in the pack.

 

[okashira]

Thanks for the data. Yeah, this cell is about as saggy as i thought it would be under high load. It's no surprise for a high energy density cell.

I'm sure that the car cruises at a nice 0.33C-0.66C rate, which is pretty gentle on the batteries. It's acceleration and hill climbs where they start shedding decent amounts of their internal energy as heat. Explains the comprehensive liquid cooling in the pack.

Actually, the liquid system in the car is not really used for cooling of the batteries. It's for environmental temperature control of them. Used for heating more then cooling.
Cooling would be for when the car is left on hot asphalt in the sun with a high SOC.
The cells are designed for high temp durability, but don't perform well in the cold. Heating is more preferable to cooling since it only needs to be done with the car is in use, can be done for cheap, i.e., waste heat from the motor.

 

[neptronix]

Yes, but surrounding the cells with a large thermal mass is ultimately a form of thermally controlling the battery temps. It's added mass and volume. Kind of a bandaid in lieu of a cell that would consistently run cool under those kind of loads, and not have such an issue with cold temps.

Sorry, i'm an efficiency nut. These things bug me :lol:

 

[okashira]

Yes, but surrounding the cells with a large thermal mass is ultimately a form of thermally controlling the battery temps. It's added mass and volume. Kind of a bandaid in lieu of a cell that would consistently run cool under those kind of loads, and not have such an issue with cold temps.

Sorry, i'm an efficiency nut. These things bug me :lol:

It's great that you are an efficiency nut, however your concerns are misplaced. The Model S is the most efficient EV out there, pound for pound at least.
In addition, the batteries are a negligible source of heat in the Model S.
The pack would run just fine without any thermal system whatsoever for the battery (except it would suck in the cold.)
On the other hand, if you would suggest they were to run HK lipo's, the pack would probably not last a week in the Arizona sun on sitting on black asphalt.

 

[neptronix]

The Model S is the most efficient car for sure.
The car is very aerodynamic and the motor is very efficient, but the battery negates those positives. It could be better.

This situation with heat and energy wastage just gets worse and worse as the battery ages and gains internal resistance, so it is important to have a lot of headroom here. The liquid cooling will be used more and more as the pack ages and the voltage sag and energy wastage will become even worse than it started. This pack will not age as gracefully as it could.

HK Lipos will operate fine in major heat as long as you are not making major heat with them in the first place. It is easier to not make heat on a higher discharge cell that is overrated for it's application. Cold temperature performance will also be better, removing the requirement for heating the battery pack in the winter.
I'll give you an example: i have had a 20C 20AH 36V RC Lipo pack for 5 years. Under the typical 2C load i'd give it, the whole pack's voltage would drop by 0.5v. At 4 years old, saw double the internal resistance, and the drop was 1V. I am still not making any heat in the pack.
At below freezing, the pack would sag down 3 volts. I now have the performance of a standard commercial ebike battery and get around 90% of capacity or more. Not bad!
A standard commercial ebike pack would be smacking the low voltage cutoff in such temperatures very quickly.

Nissan had this discharge efficiency problem in Arizona. If ambient temps are 120f and that battery chemistry reached 160f or higher, you've got rapid degradation because air-cooling is basically doing nothing for you and the pack was generating enough heat to take it over the edge. Their pack had no headroom in it even when brand new in that situation. They had to change the chemistry plus add liquid cooling in order to meet the warranty requirement.

Basically, i like overrated cells because they perform well into old age and have the best discharge efficiency and performance you could ask of a battery. This applies to any battery.
The Model S flogs it's pack pretty damn hard under acceleration and hill climbing when new, and this situation will get worse as it ages.

 

[riba2233]

DCIR on these doesn't increase much with aging. And even if it did, these cells see really low C rate most of the time, like below 0.5C, so it's not concern at all.

 

[neptronix]

DCIR on these doesn't increase much with aging. And even if it did, these cells see really low C rate most of the time, like below 0.5C, so it's not concern at all.

How is the IR increase over cycles and calendar life any different than other cells? got any info? i'd be happy to read it.

Yeah, they do see a low C rate most of the time, which means that range will still be good on long highway trips.
It's during acceleration and hill climbing where the cells see 2C and beyond that IR really matters. How does the car perform in stop and go during the winter after 7 years of use and 100,000+ miles on the clock? time for a new battery already, is car just showing diminished power but climbing the big hill fine, or what?

Admittedly, the Tesla battery pack is built for longer term reliability than many past EVs have been. I'm just imagining that the battery may be pretty damn saggy by the time the warranty period is up.
I think that EVs can last longer than your typical ICE vehicle without major repairs and hope that Tesla engineered the car to do that, but 99% of car companies that produce evs and hybrids usually do poorly on speccing the battery. Honda's guilty of doing this and so was/is Nissan. Leaves a bad taste in people's mouths.

I hope you are right and i am wrong.

 

[riba2233]

Even if it is saggy, you wont feel it in car's performance, controller is buck converter, and has field weakening, so it will always get power it needs at whatever rpm, it will just demand more current (power) from battery. They don't have simple system like we do on our bikes, we feel loss of rpm and saggy performance. One guy did testing on cgr18650cg panasonic cells on vectrix, ir was not changing almost at all after 300 cycles, and they are 1C laptop cells (although tesla roadster used them), and vectrix pulls lot of current (his battery was around 90 Ah, and maw draw was over 200 A). And these cells are inferior, old chemistry, NMC.

 

[devo1223]

I just bought a bunch of temperature probes to measure my ebike Tesla cell pack. What is an acceptable temperature for these while discharging?

 

[riba2233]

45-50? Even more for short periods.

 

[okashira]

0V test now at .....

I should have set the multimeter to log to see just how much capacity comes from the cells all the way down to 0V. :-/

It's been nearly a year since I put these cells in the 0V storage test.
I disconnected the resistor about a month ago and voltage still read about 0.002V.
Over the past month, the cells increased in voltage to about 0.4V.

I went ahead and hooked up one to my tester and charged it with 0.1A to 2V, then 0.3A to 2.8V, then 0.5A to 4V. Then I let the cell rest to see if voltage dropped. Nope, held 3.99V like a champ.
Went ahead and charged to 4.2V with 0.1A cut. And ran a 3A discharge.
The cell showed like 94% full capacity, maybe DCIR picked up a little bit. However, it still holds a charge and works great. I'll run a few more cycles soon.

 

[randyc1]

Hi Okashira , I bought your cells the other day for a 10S4P pack for an eskate , I'm going to buy a simple 42V lithium ion charger,.. what Charging Current would you recommend, 2A, 3A,.ect ??

Thank you?

 

[Pajda]

Great work, thx okashira for sharing this test result

 

[okashira]

Hi Okashira , I bought your cells the other day for a 10S4P pack for an eskate , I'm going to buy a simple 42V lithium ion charger,.. what Charging Current would you recommend, 2A, 3A,.ect ??

Thank you?

4A or less would be fine
Don't charge when pack is cold 45F or less. let it warm up first.
Or use very low charge current

 

[randyc1]

Ok thank you .

 

[LockH]

Awesome to see this data public. Thank you for taking the time.

<insert Thumbs Up here.>