Mooch’s Bench Test Results: Tesla Model 3 21700...good performer but they’re salvaged cells

The Citadel - Essential References for Esk8
#1

My apologies for taking so long to post this.

Cells were donated for testing by @Skyart and @BenjaminF. Thank you!

Test Report Summary: This cell is available with a lot of different battery wraps and top contacts, all salvaged from Tesla Model 3 vehicles. While the ones I tested were good performers there is no way to know the condition of any of the cells you buy and you use them at your own risk.

Here’s how BigBattery prepares the cells: https://youtu.be/ckhjYMAE9yg

Cell photos, graphs, ratings graphic: Imgur: The magic of the Internet

This cell has recessed top contact that Tesla welds a cell-level fuse wire onto when they are used in a Tesla vehicle (see photos above). The companies that salvage these cells weld standard top contacts onto them and wrap them in various color wraps with different printing and ratings on them.

There has been concern that this recessed top contact could mean that these cells do not have the Current Interrupt Device (CID) that is in every cell made by the large manufacturers and hopefully every round cell we use. The CID is an important safety feature that shuts down the cell like a one-time circuit breaker if there is too much gas pressure building up inside due to misuse or mishandling. This can help keep the cell from bursting.

I took apart one of the cells I had and there appears to be a structure in the top contact structure of the cell (see the red arrows in the photos) that acts as a CID. But I cannot say for sure and I cannot guarantee that this cell will try to shut down if used too hard.

The cells I tested also have three tiny “stripes” in the bottom center that every Tesla Model 3 cell has and, AFAIK, no other cell has. I believe these marks are from the internal spot-welding of the tab of metal carrying current to the metal can from the spiral roll of materials inside the cell.

These Tesla cells equaled the performance of the Samsung 50E2/50E3 and Vapcell G50 (Samsung 50G rewrap). They performed a little better than the Molicel M50A and a lot better than the Samsung 50E and LG M50T.

The six cells I tested delivered from 4796mAh to 4888mAh and had a DC internal resistance ranging from 20.4mOhm to 22.9mOhm. This is not great consistency but these cells came from three different packs and their prior EV use (if any), storage time and conditions, and the quality of the top contact welding can affect their performance.

I estimate that these cells have a continuous current rating of about 10A to ensure decent cycle life and I rate them at 4800mAh. Like all of the ultra-high capacity 21700’s I recommend staying below about 7A (continuous or pulsed) to improve performance and minimize voltage sag.

Ratings graphic: Imgur: The magic of the Internet

Full Test report: https://bit.ly/3ve5rz8

I want to work for the vaping community full time! If you feel what I do is worth a couple dollars a month and you would like early access to battery availability, battery and charger testing and news, and a say in what I test, then please consider becoming a patron and supporting my testing efforts: Battery Mooch is creating battery and device tests | Patreon.

These tests only note the ESTIMATED ratings for these batteries at the time I tested them. Any battery that is not a genuine Samsung, Sony, Murata, LG, Panasonic, Molicel, or Sanyo can change at any time! This is one of the hazards of using “rewrapped” batteries or batteries from other manufacturers so carefully research any battery you are considering using before purchasing.

Misusing or mishandling lithium-ion batteries can pose a SERIOUS RISK of personal injury or property damage. They are not meant to be used outside of a protected battery pack and you use them AT YOUR OWN RISK. Never exceed the battery’s continuous current rating and keep the plastic wrap and top insulating ring in perfect condition.

Any rating in my ratings tables can change at any time as different grade cells appear on the market, we get swamped with fakes, or new information becomes available to me. Please, never assume that the ratings in the tables are permanent and will never change! Always download the latest version before considering any cell purchase.

To see how other cells have tested check out this link: https://www.e-cigarette-forum.com/forum/blog-entry/list-of-battery-tests.7436/

47 Likes
The battery builders club
Anyone even using 18650s anymore?
Noob question thread! 2020_Summer
Meepo Hurricane Discussion
#2

Excellent work Mooch, thank you so much!

To anyone who isnt giving this man a cup of coffee a month, I urge you to consider how much knowledge, safety, and value he has brought to our community and to so many others!

16 Likes
#3

Wow, glad to hear these potentially perform better then the almost $9 M50A i was tempted.

7 Likes
#4

Thank you @Battery_Mooch your work is awesome
Thank you Siri when your driving and can’t read :laughing:

image
image1284×2580 570 KB

6 Likes
#5

I have a feeling I know the answer, but just curious what you think about the floatwheel (diy onewheel) crew running the supplied 12s1p 2170 pack and routinely pulling 20-40A out of the thing (most of them are leaving max bat vesc settings at ±99A).

1 Like
#6

What cell does the pack use? Tesla 3?

2 Likes
#7

That’s what’s advertised.

1 Like
#8

Using these Tesla 3 cells at 20A-40A, even if only bursts, is just silly. Their performance will be terrible with huge voltage sag reducing the riding time.

Did they chose this cell because of its capacity? Bad idea. That spec is almost useless when over a few amps.

I’m not too worried about safety though, it takes a lot to heat up a cell if not short circuiting it, but the cycle life could be reduced a lot too.

The P42A would probably give them noticeably more riding time if they’re drawing that much current.

9 Likes
#9

Thanks for this @Battery_Mooch i have been eager to see ur report. I was kinda holding these would have more than 10A discharge but oh well lol.

2 Likes
#10

The floatwheel boys did a lot of good things, but some things were equally bad. With the motor KV they chose I would suggest going at least 16S, adding extra P wont fix anything since the issue is duty cycle on 12S. (over 25 km/h and past 50% charge mind you.)

3 Likes
#11

That 20A curve is really interesting tho… would not have expected the capacity to rise back up like that.

Any theory on why that may have happened? Fluke or something to do with how Tesla heats cells to a relatively high(?) temp for the extra current output in the ludacris modes?

3 Likes
#12

Happens between 5A and 10A too.
It’s common though. A lot of cells show some extra capacity at the lowest voltages when being run harder.

IMO this is due to the cell warming up as the discharge proceeds and that lowers the internal resistance. This lower internal resistance results less voltage sag so it takes a bit longer before the cell drops to the low voltage cutoff point.

12 Likes
#13

Ordered 200 today! Thank you!

2 Likes
#14

Hey @Battery_Mooch I have a general question for you about these cells and possible mating.

Ideally no-one would attempt to use X cells (for example these) and Y cells (for example another manufacturer) together. But in a non-ideal world. Is there a comparable off the shelf sku for cells like these that you’ve tested? Asking because I am a few cells short of a 12s10p and while I can find these on ebay, I’m not sure if that will last much longer and I’m not keen on spending the money right this second unless I really should.

2 Likes
#15

No one ever reads my posts. :weary:
I mention some equivalent cells in my OP. :grin:

I understand your predicament but I strongly urge you to not mix cell types or to use salvaged cells and to only use closely matched cells from the same batch and purchased at the same time from the same vendor along with a very good BMS.

It’s your call though. I must always err on the side of caution and safety if asked for advice.

7 Likes
#16

OOOF, thank you for calling me out on that non-sense. I yell at people almost daily for not reading documentation (so I should have known better). Thank you again for not just putting it in the reply and making me find it :smiley:

I really dont want to mix cells either just wanted to know how you felt about it, so chances are ill pick some up or just “deal” with having a 12s8p.

5 Likes
#17

How many cells short are you? @DerelictRobot has several left over from the 12s11p that I built for him.

5 Likes
#18

20, there’s a guy on ebay selling 25 for 5.76 a pop and another guy selling 50 for 3.40 a pop - I really cant use the extra 30 cells so I dont want to spend the money on the cells but its really the only financially responsible option.

3 Likes
#19

Hey quick question. Does someone know the charge rating? 1C and therefore 4.8a per cell?

1 Like
#20

No one knows.
Ignore C ratings, they are completely useless for anything and so widely abused as to be nonsensical much of the time. They result in high charge rates for ultra-high capacity cells but those cells are the ones that have the lower charge ratings compared to the lower capacity “power” cells.

In my opinion 4.8A is much, MUCH too high a charge rate and will result in reduced cell life. 1A is probably a good rate for decent life and 2A would be about as hard as I would charge them.

2 Likes