Any Tesla 18650 data sheets available from years gone by?

brumbrum

100 kW
Joined
Oct 30, 2010
Messages
1,383
Location
Wales U.K
Is there any info out there that correctly states how tesla car cells may have been different to average Panasonic 18650 cells?
And is there any official data stating the AH and discharge rating of any of the tesla cells used over the past 6-7 years in their cars?

I need to know whether if it is worth pursuing a purchase of these types of cells with consideration to their age/shelf life/capacity and C rating according to year of manufacture.

Anyone have any knowledge of information to share?
Cheers :D
 
Thanks, yeah i did a bit of digging on here and elewhere on the web. It seems there have been at least a couple of differing cells used over the years. It would be nice to know how the cells differ from the standard panasonic cells of equal capacity and discharge. I have seen test claims stating from 6,8,10 and 12amp continous draw, but also as being able to maintain much higher and longer peaks than the 'usual' 18650's and having a longer shelf life even when abused. So whats the secret recipe? And is this recipe not used elsewhere purely due to design obsolescence?
 
Tesla's cells are not very high performance cells.
There are many 18650's with much higher discharge ability, some with more capacity,
Tesla's real advantage is their long cycle life with minimal capacity loss together with an aparent ability to tolerate low discharge voltage levels.
 
brumbrum said:
So whats the secret recipe?

Faith.

Tesla cells are Panasonic technology, there is no magic about them and nothing involved that Panasonic is not able to use in their own cells.

Afaik the Tesla cells lack the shrink warp, lack the usualy built in security feature (CID?) and maybe their cell can wall is a bit thinner (?)
 
"Tesla" cells can be a bit missleading.
Whilst they have been using mainly cells made and supplied by Panasonic, made to Tesla's spec, it is known that they have also used cells from LG, Samsung, Sanyo (now part of Panasonic) and likely others also.
The only cells Tesla have actually manufactured themselves are the 21700 size used in the Model 3 and some of the Storage system modules ( though they have also used other suppliers for those 21700s too !)
Testing has shown the "Tesla" 18650 is near identical in performance to the Panasonic BE spec cell .
 
The critical thing for longevity of any 18650 cell is cooling - Tesla's cooling system is rather good, and it strives to maintain the cells at a constant 25'C (+/- 5'C) for as much of the cells life as possible. Pretty much any cell will live a long life if thermally managed.
So yeah, there's nothing particularly special about Tesla's cells - they just treat them nicely.
 
jonescg said:
The critical thing for longevity of any 18650 cell is cooling - Tesla's cooling system is rather good, and it strives to maintain the cells at a constant 25'C (+/- 5'C) for as much of the cells life as possible. Pretty much any cell will live a long life if thermally managed.
So yeah, there's nothing particularly special about Tesla's cells - they just treat them nicely.

Thanks,, that makes it clear. Constant optimum temperature is the special sauce, not something in the cell constuction itself. Thanks guys.
 
Tesla cells in the model S are clones of the the NCR18650BE and NCR18650BM cells. The BM is essentially a slightly improved BE.

These cells are state of the art, not because of their initial capacity (which is average+) but because of their excellent cycle life, tight cell-to-cell dispersion, very progressive degradation of impedance and extra safety (top+bottom vent and manufacturing process control)

I have no idea what are the specifications for the 21700 cells, and no one will have until a model 3 wreckage ends up in an ES member's garage :D
 
youyoung21147 said:
Tesla cells in the model S are clones of the the NCR18650BE and NCR18650BM cells. The BM is essentially a slightly improved BE.
I can confirm that BM cells in my cycle life tests behave almost identical to the Tesla 85kWh battery cells.

youyoung21147 said:
These cells are state of the art, not because of their initial capacity (which is average+) but because of their excellent cycle life, tight cell-to-cell dispersion, very progressive degradation of impedance and extra safety (top+bottom vent and manufacturing process control)
The information about "excellent cycle life" sounds to me as advertising message when I compare my own 100% and 50% DoD cycle life tests of used Tesla 85kWh cells and NCR18650BM cells with other industrial NCA cells. Tesla cells shows almost the same capacity drop after 500cycles as NCR18650B or NCR18650PF when cycled at 0.5C-1C rate. They have significantly lower IR than NCR18650B but higher than NCR18650PF.
 
Pajda said:
youyoung21147 said:
Tesla cells in the model S are clones of the the NCR18650BE and NCR18650BM cells. The BM is essentially a slightly improved BE.
I can confirm that BM cells in my cycle life tests behave almost identical to the Tesla 85kWh battery cells.

youyoung21147 said:
These cells are state of the art, not because of their initial capacity (which is average+) but because of their excellent cycle life, tight cell-to-cell dispersion, very progressive degradation of impedance and extra safety (top+bottom vent and manufacturing process control)
The information about "excellent cycle life" sounds to me as advertising message when I compare my own 100% and 50% DoD cycle life tests of used Tesla 85kWh cells and NCR18650BM cells with other industrial NCA cells. Tesla cells shows almost the same capacity drop after 500cycles as NCR18650B or NCR18650PF when cycled at 0.5C-1C rate. They have significantly lower IR than NCR18650B but higher than NCR18650PF.

At what discharge rate?

Isn't part of the Tesla secret sauce simply over-provisioning. The more cells, the lower the discharge rate for any given power draw; the longer the life.
 
brumbrum said:
I need to know whether if it is worth pursuing a purchase of these types of cells with consideration to their age/shelf life/capacity and C rating according to year of manufacture.

I would never buy a cell used in a EV battery even if it's a Tesla super-ultra-extra-18650 one. I'd better buy unused Boston-Power Swing cells if the cycle life is critical or Sanyo NCR18650GAs if the capacity is more important. I doubt that Tesla's technology is more advanced than Sanyo/Panasonic's.
 
Buk___ said:
At what discharge rate?

Isn't part of the Tesla secret sauce simply over-provisioning. The more cells, the lower the discharge rate for any given power draw; the longer the life.

My test methods are 0,5C CC-CV charge rate @ 100mA cut-off (standard charge rate by most manufacturers - I use this value even if manufacturer recomended 0,3C or 0,7C for the same conditions or lower cut-off current), 10minut rest time, 1C discharge rate, 10minut rest time and this is cycled 49x. Then I add one nominal (manufacturer specified) capacity test cycle 0,5C-0,2C. So you need ca 3months at 50% DoD and 6months at 100% DoD for 1000cycles.

I use "low-cost" ZKEtech testers with 4wire measurement, but I have checked them with calibrated laboratory 6 1/2digit multimeter which confirms the measurement errors values and current/voltage regulation indicated by the manufacturer 5mA and 5mV.
 
Buk___ said:
The information about "excellent cycle life" sounds to me as advertising message when I compare my own 100% and 50% DoD cycle life tests of used Tesla 85kWh cells and NCR18650BM cells with other industrial NCA cells. Tesla cells shows almost the same capacity drop after 500cycles as NCR18650B or NCR18650PF when cycled at 0.5C-1C rate. They have significantly lower IR than NCR18650B but higher than NCR18650PF.


At what discharge rate?

Isn't part of the Tesla secret sauce simply over-provisioning. The more cells, the lower the discharge rate for any given power draw; the longer the life.

In my testing the cells are charged at 1A (0.3C) CC-CV to 4.19V and CC discharged at 9A down to 2.8V, then CV discharge down to 3A/2.8V. The tester is a fully calibrated iCharger 4010.

The test method is not conventional, because I tested these cells for a very specific use case.

But what I can assure you is that the BE/BM cells outperform the GA/MJ1/35E holy trinity in ~100 cycles. Can't comment about 18650B or PF in the same conditions.

Tesla strategy to enhance the cell life is multi-aspect :
- over-provisioning (500km drives are rare)
- artificially limited use envelope (charged to 4.1/4.15V max)
- temperature regulated
- advanced charging algorithm to reduce the detrimental polarization voltage during critical charging stages
- flawless balancing
- cell matching

With all these conservative measures, the cycle count of the Tesla battery is only 500 cycles after 200000km which is nothing special especially compared to satellite or stationary applications seeing 3000 to 30000 cycles
 
youyoung21147 said:
Buk___ said:
At what discharge rate?

Isn't part of the Tesla secret sauce simply over-provisioning. The more cells, the lower the discharge rate for any given power draw; the longer the life.
CC discharged at 9A down to 2.8V, then CV discharge down to 3A/2.8V.

My point was that because of over provisioning, the draw on individual cells is (well) less than 0.5C and that goes a long way to extending their longevity.

Testing at such high draw rates is probably okay for comparison purposes, but is a poor indication of their potential under Tesla's scheme.

(Also, you might want to watch Professor Jeff Dahn describing a testing schema that successfully categorises cells in minutes rather than hours days or weeks. I don't know how easy it is to reproduce his test equipement, but if you can it would be a huge time saver if you are doing this a lot.)
 
Thx youyoung21147 for clarification of your test cycle. I agree that your test is very specific and so it is deadly for cells with high internal resistance. The NCR18650B will be ruined soon but I believe that PF can show good results under your test. And I can highly recomend you to try a LG HG2 cell in this particular application.

But in my opinion this test cycle is totally misleading for an EV apllication. Here you have average load up to 0.5C charge and 0.3C discharge rate with only about 50% average DoD, so the high energy (HE) cells with relativelly high IR can perform even better than high power (HP) cells in the term of cycle life. But of course HE cells will have significant problems with frequet fastcharging (1C or more)
 
Buk___ said:
(Also, you might want to watch Professor Jeff Dahn describing a testing schema that successfully categorises cells in minutes rather than hours days or weeks. I don't know how easy it is to reproduce his test equipement, but if you can it would be a huge time saver if you are doing this a lot.)

Nice link ! Will watch the video in full when I have time. I've seen a few papers about EIS testing for various Liion cells evaluation purposes, including online SOC estimation. It's beyond the effort I am willing to put in my personal testing. These tools must be really valuable to battery chemistry scientist when developing new formulations. We have an EIS setup at work used for cathodic disbonding experiments, it's very expensive (30K$) and not always helpful in drawing conclusions ^^

The testing I have performed is in no way similar to the Tesla use case. I was just pointing out that their cell is more about the right balance of all EV performance indicators than pure, 0-cycle energy retention figures. Even under quite severe testing, the cell shows a more consistent behavior over its life and that's what makes it great.
 
Pajda said:
Thx youyoung21147 for clarification of your test cycle. I agree that your test is very specific and so it is deadly for cells with high internal resistance. The NCR18650B will be ruined soon but I believe that PF can show good results under your test. And I can highly recomend you to try a LG HG2 cell in this particular application.

Do you have cycle life test data for HG2 cell ? I've performed some tests of Samsung 30Q cell but not HG2. Not great cycle life but no degradation of internal resistance over time.

I tested 30Q in priority based on this test (quite extreme ! cells reached 80C)
https://www.reddit.com/r/electronic..._test_results_which_lasts_longer_vtc6_30q_or/
 
youyoung21147 said:
Do you have cycle life test data for HG2 cell ? I've performed some tests of Samsung 30Q cell but not HG2. Not great cycle life but no degradation of internal resistance over time.

Sadly not.

But the datasheet looks very promissing: https://www.nkon.nl/sk/k/hg2.pdf
 
Back
Top