Tesla is showing impressive progress in ramping up production of 4680 battery cells. According to the company, this is no longer its bottleneck and it even has cells in stock in its warehouses.
Let me stop you here and say you’re going to have a bad time if you attempt to have this explained to you by a text generator.
The first problem is you asked it about the form factor rather than the cells, manufacturing process, and chemistry themselves. The shape makes perhaps the least difference of all. Well, except that Panasonic (the company that actually invented Tesla’s “revolutionary” battery tab mechanism) already documented a ton of issues with assembling cells reliably when you’ve got a bunch of tabs. What you want in that case is actually a pouch cell or a prismatic.
Bard also missed a pretty major one under “Lower production cost”. Massive volume makes 21700 cells cheaper too, since all of your costs are divided by a huge number. This is also why 18650s are beating the 21700s. Also, fun fact, Tesla’s top performing pack for charge and discharge rates is built in 18650 NMC cells.
Wider compatibility
That’s a meaningless argument produced by a text generator.
Faster charging
Not some cases. All cases. All of Tesla’s products based on 21700 cells from all of their suppliers charge faster than any of the packs based on Tesla’s in-house cells. Charging “infrastructure” has absolutely nothing to do with this, it’s a nonsense hallucination by a text generator. This is one of the major reasons to not use one.
Lower energy density
This is just an outright lie, which you could determine by doing the math. Tesla’s early Model 3 cells were 17.3 Wh each, while their in-house 46800 are 86 Wh. The Model 3 cells have improved over time, but let’s stick with the earliest of both products. So this means their 21700 cells were 247 Wh/kg while the 46800s are 244 Wh/kg. Taking the volume of each cell into account, the old Model 3 cells are 713.55 Wh/L and the 46800 cells are 651.5 Wh/L. So the simple math clearly shows it’s lower density in both dimensions.
Bard slurped up text from marketing efforts and mangled it with some other words and gave it to you. This is a warning not to use these things to learn new topics, they’re less reliable than Wikipedia.
More complex manufacturing
This is another nonsense argument. First and foremost, Tesla uses stamped collectors on the Y and 3 packs based on 21700 cells so there’s really no more complexity in manufacturing. On the S and X packs, they use bonding wires. Bonding wires robots do this work insanely quickly, so it’s a foolish argument there as well. Bard has again slurped up nonsense from a fan blog and repeated it, because there’s no intelligence in what Bard does. Trusting these things blindly is a problem.
As an example of what bonding wire robots do, BTW, they are used in the production of every single processor and microprocessor on the planet. Products numbering in the tens if not hundreds of billions per year, with precision requirements wall beyond anything Tesla needs for their old packs from 2012, let alone the newer stamped collectors from 2018.
Limited scalability
I already demonstrated that the energy density bit was made up nonsense. But this part is also complete horse shit. Cell chemistry tends to improve approximately 5% per year averaged over a 5 year window. Tesla didn’t increase the pack dimensions when going to the 82 kWh Model 3 pack, they simply changed cell chemistry. If I used the cells from that pack in the calculations above, it would make this argument outlandishly pointless.
4680 Simpler manufacturing
The cell isn’t “tabless” bit instead has a shitload of tabs. Like I said already, Panasonic invented this tech and they abandoned it because it’s a frequent source of cell failure. Their patents on this are from over a decade ago, they documented all of their issues pretty well, and Tesla’s own internal documents when leaked showed they have massive yield problems in their Texas battery factory.
Potential for faster charging
Potential. And instead what we see in real products on sale is the opposite. so that potential didn’t materialize.
Let me stop you here and say you’re going to have a bad time if you attempt to have this explained to you by a text generator.
The first problem is you asked it about the form factor rather than the cells, manufacturing process, and chemistry themselves. The shape makes perhaps the least difference of all. Well, except that Panasonic (the company that actually invented Tesla’s “revolutionary” battery tab mechanism) already documented a ton of issues with assembling cells reliably when you’ve got a bunch of tabs. What you want in that case is actually a pouch cell or a prismatic.
Bard also missed a pretty major one under “Lower production cost”. Massive volume makes 21700 cells cheaper too, since all of your costs are divided by a huge number. This is also why 18650s are beating the 21700s. Also, fun fact, Tesla’s top performing pack for charge and discharge rates is built in 18650 NMC cells.
That’s a meaningless argument produced by a text generator.
Not some cases. All cases. All of Tesla’s products based on 21700 cells from all of their suppliers charge faster than any of the packs based on Tesla’s in-house cells. Charging “infrastructure” has absolutely nothing to do with this, it’s a nonsense hallucination by a text generator. This is one of the major reasons to not use one.
This is just an outright lie, which you could determine by doing the math. Tesla’s early Model 3 cells were 17.3 Wh each, while their in-house 46800 are 86 Wh. The Model 3 cells have improved over time, but let’s stick with the earliest of both products. So this means their 21700 cells were 247 Wh/kg while the 46800s are 244 Wh/kg. Taking the volume of each cell into account, the old Model 3 cells are 713.55 Wh/L and the 46800 cells are 651.5 Wh/L. So the simple math clearly shows it’s lower density in both dimensions.
Bard slurped up text from marketing efforts and mangled it with some other words and gave it to you. This is a warning not to use these things to learn new topics, they’re less reliable than Wikipedia.
This is another nonsense argument. First and foremost, Tesla uses stamped collectors on the Y and 3 packs based on 21700 cells so there’s really no more complexity in manufacturing. On the S and X packs, they use bonding wires. Bonding wires robots do this work insanely quickly, so it’s a foolish argument there as well. Bard has again slurped up nonsense from a fan blog and repeated it, because there’s no intelligence in what Bard does. Trusting these things blindly is a problem.
As an example of what bonding wire robots do, BTW, they are used in the production of every single processor and microprocessor on the planet. Products numbering in the tens if not hundreds of billions per year, with precision requirements wall beyond anything Tesla needs for their old packs from 2012, let alone the newer stamped collectors from 2018.
I already demonstrated that the energy density bit was made up nonsense. But this part is also complete horse shit. Cell chemistry tends to improve approximately 5% per year averaged over a 5 year window. Tesla didn’t increase the pack dimensions when going to the 82 kWh Model 3 pack, they simply changed cell chemistry. If I used the cells from that pack in the calculations above, it would make this argument outlandishly pointless.
The cell isn’t “tabless” bit instead has a shitload of tabs. Like I said already, Panasonic invented this tech and they abandoned it because it’s a frequent source of cell failure. Their patents on this are from over a decade ago, they documented all of their issues pretty well, and Tesla’s own internal documents when leaked showed they have massive yield problems in their Texas battery factory.
Potential. And instead what we see in real products on sale is the opposite. so that potential didn’t materialize.
Thanks for this. It was insightful.