Model 3 battery cell specs

okashira

10 kW
Joined
Nov 17, 2014
Messages
963
Location
Houston, TX
I thought I'd share what I have learned so far of the Model 3 cells.


  • Made by Panasonic in the USA (duh)
    21700 form factor. 21mm diameter and 70mm length.
    4.8Ah nominal.
    3.65V nominal.
    Reduced cobalt and increased manganese in the cathode compared to previous generation
    Silicone in the anode like the NCR18650GA / 90kwh/100kwh Model S cells.
    Recessed positive terminal. This creates a new challenge for spot welding.

Ps my model 3 due for delivery this week. :D
 
Based on some more speculative estimates on my part, the Model 3 pack is probably capable of more power then even the P100D pack. Perhaps as much as 650kW in short bursts.
 
is the internal resistance already known?
love to see discharge curves with temperatures. i am really interested how cool the cell can stay without active cooling.

and more importantly: what are they going to cost if they hit the market...
 
flippy said:
is the internal resistance already known?
love to see discharge curves with temperatures. i am really interested how cool the cell can stay without active cooling.

and more importantly: what are they going to cost if they hit the market...

I suspect ~.00045 ohm per group or 0.043 ohm for the whole pack. Lower then the model s.
That's room temperature burst dcir.
I have no sources just speculation ;)
 
I would love for you to be right. Awesome hotroding potential.

okashira said:
Based on some more speculative estimates on my part, the Model 3 pack is probably capable of more power then even the P100D pack. Perhaps as much as 650kW in short bursts.
 
I read a report that indicated they were somewhere between 300 and 333whrs/kg.
If that's true, i wonder what we're all doing with these lower density cells laying around.
 
neptronix said:
I read a report that indicated they were somewhere between 300 and 333whrs/kg.
If that's true, i wonder what we're all doing with these lower density cells laying around.



Building battery packs for bikes of course! Besides, we don't necessarily need the highest energy density for a standard bike. People will be plenty happy with average cell density because they can still get 10-30 miles of range on a bike that's more than enough for their commute. I forget what company they're called but they have 18650 cells that get around 400wh/kg density right now and are planning higher than that. I think around 500-600 upcoming. So I think it's possible. Just a matter of time and technology. Nothing is better than all the ev competition coming for different auto manufacturers. Someone will eventually Crack the key to high density, fast charge. Amazing range..
 
So I will be first one to be sceptical about this new Tesla 21700 cells. I think they will have the same cycle and power performance as Tesla actual 18650 cells at best. I believe that their greatest benefit is that they will be significantly cheaper in mass production on both cell and battery pack level.
 
Philaphlous said:
Besides, we don't necessarily need the highest energy density for a standard bike. People will be plenty happy with average cell density because they can still get 10-30 miles of range on a bike that's more than enough for their commute.

When you start building fast bikes, 10-30 miles becomes 3-10 miles pretty quick, lol.
I gave up on my 45mph capable bike and started driving my car to work again because i just couldn't jam enough battery in to travel at the speed limits in places where there's no bike lanes. That sucks!

Better energy density means that i can ditch gas without forking out for an EV car.
 
neptronix said:
I read a report that indicated they were somewhere between 300 and 333whrs/kg.
If that's true, i wonder what we're all doing with these lower density cells laying around.
Unless they have magicly produced a cell much lighter than the 69gms recorded for other 21700 cells, then its still at the 250 - 260 Wh/kg...(same as many existing 18650cells.)
Infact they would have to be down at around 55gms just to crack the 300W/kg !
 
Well, let's get a calculator out and do the math.

Correct me if i'm wrong, but i believe the current 2070 capacity record holder is the lg m42..
LG M42 @ 58g..
3.64v nom x 4.2ah = 15.288
15.288 / 0.058 = 263.586 whrs/kg.

..which is very close to what you see in 18650 chemistries.. 250-260whrs/kg.

We don't have a certain weight for the Tesla cell yet, but one group thought the cell is 4.8AH.. so let's be a little pessimistic and say 60g.
3.65v nom x 4.8ah = 17.52
17.52 / 0.060 = 292 whrs/kg.

Another poster seems to be very sure it's 5.75AH and 66 g, which would calculate to 322.3 whrs/kg
https://www.reddit.com/r/teslamotor..._2170_battery_cell_specifications_calculated/

I'm kinda surprised some of these cells haven't slipped out the back door and found their way into the hands of a battery geek yet.
 
I've already seen a handful of 5Ah 20170's, but 5.75Ah (and dirt cheap) seems unlikely to me. If they really achieved it, hat's off to them.
 
If sanyo's 20700 a and b cells are used as a yard stick then 4.8ah sounds the likely one to be on the cards, the sanyo cells only have a 4c rate nominal for higher c rates the capacity drops significantly on the a version. Panasonic have messed about in the cell so it's a guessing game till one is documented under test.
If 4.8ah and 4c was the cell specs i'd be happy with that I could double the range on my ebike and not change weight, I could effectively have an A2B metro that does over 70 miles on the internal stock pack alone if rode at bike like speeds or foot down and easily keep up with cars for 15miles or so, plenty for my local trips and to get out to the seaside etc.
 
neptronix said:
.

Correct me if i'm wrong, but i believe the current 2070 capacity record holder is the lg m42..
LG M42 @ 58g..
3.64v nom x 4.2ah = 15.288
15.288 / 0.058 = 263.586 whrs/kg.

..which is very close to what you see in 18650 chemistries.. 250-260whrs/kg.

We don't have a certain weight for the Tesla cell yet, but one group thought the cell is 4.8AH.. so let's be a little pessimistic and say 60g.
3.65v nom x 4.8ah = 17.52
17.52 / 0.060 = 292 whrs/kg.
But the Tesla cell is a 21700 , (bigger than the LGM42, which is actually 20650), and as i said, other 21700 cells have weighed in at 69 gm.
Also , we know the pack is 80.5 kWh with 4416 cells..so each cell is 18.23 Wh or 4.8 - 4.9 Ah as stated by the OP
Now your calculator will tell you its still at 263 Wh/kg
Just to add..
Panasonic 20700B. (Still smaller diameter,..and 10% less volume) weighs 63 gms
 
neptronix said:
I read a report that indicated they were somewhere between 300 and 333whrs/kg.
If that's true, i wonder what we're all doing with these lower density cells laying around.

The gravimetric energy density is actually slightly worse then the NCR18650GA
 
liveforphysics said:
I would love for you to be right. Awesome hotroding potential.

okashira said:
Based on some more speculative estimates on my part, the Model 3 pack is probably capable of more power then even the P100D pack. Perhaps as much as 650kW in short bursts.

Cells are up for it. Bondwires, contactors, connectors, couldn't tell you :-D
 
okashira said:
The gravimetric energy density is actually slightly worse then the NCR18650GA
that does not have to be a problem. dont forget the cells need to be attached to something. the old style (model S) packs have steel plates on both sides of the cells. the new model only has it on the top.
dont focus so hard on the Wh/kg or other stuff. having a lower count cell battery means less support material and the overall weight can be lower. it all depends on the design and cost coices they make. less cells also means less cooling loops and other stuff.

right now i am designing a 30+kWh pack for a customer. i can make the pack from a huge single block. but that would require REALLY heavy struts, metal sheets and crossmembers to make a flat pack strong enough so it wont bend and break the interconnects between the cells and the cell holders. so i am leaving that idea and use several smaller modules that have flexible interconnects and fuses between each module so the pack can be more simple to build, design and repair.
 
Hello Guys,

i have read an article about the Model 3.So some of the German automakers have bought a used model 3 from gray market(fro 100k usd) and dissembled it .They came to conclusion that the parts to build the model 3 costs around 18k usd and additionaly 10k for the production or assembly of the car.Another interesting thing, what has been mentioned is the disassembly of the cells.The Analysis shows that the newest cell do have only 2.8% Cobalt in their cathods compared to the other ones ,which have around 8%.According to Bmz Ceo this is a big competitive edge .you could read the article in the https://www.wiwo.de/my/technologie/mobilitaet/tesla-model-3-zerlegung-eines-rivalen/22625200.html which is paid, or here the short version:
https://www.golem.de/news/tesla-ingenieure-zerlegen-das-model-3-und-halten-es-fuer-profitabel-1806-134703.html

Use google translate since it is in german.

Regards Sava
 
okashira said:
Ps my model 3 due for delivery this week. :D


Hey bud.. Will you have cells in stock for sale ever again ?? thnx
 
Hey guys, I thought I would add to this older thread instead of making a new one with similar info.

How are people using these cells in custom packs? Does spot welding a cap on the top seem like a good idea or has anyone used cells with an added cap?

As mentioned in the first post, the Model 3 cells are recessed or concave on the top and there is no cap or cover, which would be a problem for spot welding. The middle of the top is about 0.8 mm lower than the sides of the can. It would be about 1 mm lower with an insulator ring. Some auction vendors have started selling capped, wrapped, and insulator ring Model 3 cells. I bought some to check out exactly what they were doing.

IMG_1064.jpg
IMG_1066.JPG

After removing the wrap and insulator ring you can see the cap is added by spot welding around the edge.
IMG_1066 - Copy.JPG

The cell on the right side is straight from a module and the jb weld like glue is still present.
IMG_1068.JPG

Here are some other cells for comparison. The blue one far right is a junk 18650 button top, but the one to the left of that in blue is a wrapped and insulated Model S 18650. That one is about 0.2 mm recessed due to adding the insulator ring.
IMG_1071.JPG

I got this pic from a good batterybro article on cylindrical cell protections. This shows the top cover or cap is normally crimped by the can in the gasket seal next to the PTC.
18650.jpg
https://batterybro.com/blogs/18650-wholesale-battery-reviews/18306003-battery-safety-101-anatomy-ptc-vs-pcb-vs-cid
 
I guess nobody is, I was wondering the same thing...

Gigawatts said:
Hey guys, I thought I would add to this older thread instead of making a new one with similar info.

How are people using these cells in custom packs? Does spot welding a cap on the top seem like a good idea or has anyone used cells with an added cap?

As mentioned in the first post, the Model 3 cells are recessed or concave on the top and there is no cap or cover, which would be a problem for spot welding. The middle of the top is about 0.8 mm lower than the sides of the can. It would be about 1 mm lower with an insulator ring. Some auction vendors have started selling capped, wrapped, and insulator ring Model 3 cells. I bought some to check out exactly what they were doing.

IMG_1064.jpg
IMG_1066.JPG

After removing the wrap and insulator ring you can see the cap is added by spot welding around the edge.
IMG_1066 - Copy.JPG

The cell on the right side is straight from a module and the jb weld like glue is still present.
IMG_1068.JPG

Here are some other cells for comparison. The blue one far right is a junk 18650 button top, but the one to the left of that in blue is a wrapped and insulated Model S 18650. That one is about 0.2 mm recessed due to adding the insulator ring.
IMG_1071.JPG

I got this pic from a good batterybro article on cylindrical cell protections. This shows the top cover or cap is normally crimped by the can in the gasket seal next to the PTC.
18650.jpg
https://batterybro.com/blogs/18650-wholesale-battery-reviews/18306003-battery-safety-101-anatomy-ptc-vs-pcb-vs-cid
 
Back
Top