Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

MASTER THREAD: 2021 Model 3 - Charge data, battery discussion etc

This site may earn commission on affiliate links.
Is the chemistry of the 87kw batteries the other cells? I'm assuming they are just different cells of the same type and not LFP. (sorry if this is a dumb question, I have only been kinda 1/2 following this since it's been too busy at work)

I have the OBD2 adapter already.. it showed up really fast.. don't even have the car yet. ;)
The 82.1kWh battery has the new 2170L cells. I don’t know if the chemistry is subtly different or whether they just make the cells a little bit denser somehow. I believe they are the same basic chemistry as the other 2170 cells (not LFP), but beyond that I don’t know what they might be doing. I don’t follow the chemistry details.

They arguably screwed up the 2170 cells (though 10-15% capacity loss after a couple years, and just 100 cycles, may still be better than some manufacturers), so maybe they tweaked them a bit to improve longevity. It helps a bit to not charge them as close to 100%, of course, which is effectively happening with these LR cells if they don’t expand the usable capacity to the same as the Performance packs. But I am fairly sure that 100% charging has very little to do with capacity loss observed on the older Model 3s. Time will tell!
 
Last edited:
Been busy for a while and only now starting to catch up, so bear with me here if I skipped a fact in the meantime...
It really is a beautiful way of hiding variable initial pack capacities amongst many vehicles. Nearly everyone starts at 508.5rkm, but not everyone has exactly the same energy.
Is this a new thing? In my understanding that has been done in the previous Panasonic pack in the same exact way.

3) How to explain that 2018/2019 Model 3 vehicles showed zero degradation for a long while, then showed it? (76kWh cap with a 77.8kWh FPWN capacity I suspect.)
Yes exactly this is what I mean:
-Only once the BMS calculates a Nominal Full Pack below 77,1kWh you saw a drop below 499km.
-Only once the BMS calculates a Nominal Full Pack below 80,6kWh you see a drop below 508/509km.
Unless Tesla provides a software update which increases your range (they could claim 370 rated miles if they really opened things up), I don't expect you'll see any indicated range loss until you reach 77.8kWh full pack, even though at that point you will have lost 2.1kWh of capacity (2.5%).
Is this variable rated consumption figure confirmed by data yet? Otherwise I would assume that every Nominal Full Pack below around 79kWh shows a reduction in Rated Range. A later update will probably raise this value to 80,6kWh too and reduce the rated consumption to around 136Wh/km at the same time. Then we see the 590km/370 miles or somewhere in that range.
The degradation really isn’t slower - at all. Unless the 2170L cells have different characteristics.
At least in my case the degradation of BMS calculation is very different to my 2020 Performance. In the exact same use case, driving, charging...all identical...my 2020 Performance saw a drop of the nominal full pack down to 73kWh, while the 2021 Performance saw no change in the nominal full pack at all. Jumping around 80,4kWh since day one.
I don’t know if the chemistry is subtly different or whether they just make the cells a little bit denser somehow.
In my understanding the Nickel ratio has gone up to increase the capacity, but hard to find a confirmation for that.
It helps a bit to not charge them as close to 100%, of course, which is effectively happening with these LR cells if they don’t expand the usable capacity to the same as the Performance packs.
Is there any solid proof that I've missed, suggesting that the LR 82kWh has a capacity limitation in comparison to the performance? Seeing both charge to 100% on SMT I didnt see any difference at all in cell voltage, nominal full pack and nominal remaining values.
 
  • Informative
Reactions: AlanSubie4Life
Is this a new thing? In my understanding that has been done in the previous Panasonic pack in the same exact way.
No, I thought it very likely that it worked this way, for a long time, but I had not seen the clear evidence of it (since you have to have SMT captures from a new vehicle to see how it works).

It sounds like this has also been your understanding for a long time - and maybe you had the supporting data too...
Is this variable rated consumption figure confirmed by data yet? Otherwise I would assume that every Nominal Full Pack below around 79kWh shows a reduction in Rated Range.
No. I don’t know what the cap value/degradation threshold is on these LR vehicles. We would need the screen capture of the energy screen as mentioned from @Stan930 or another brand-new LR owner with a confirmed 82.1kWh pack. As I mentioned above, that would provide us the cap value. We know the degradation threshold is around 77.8kWh for the older 2021 LRs. (I don’t know the precise value - you may be right it is 77.1kWh but energy screen captures with sufficient precision on the % (use trip page with a nearby destination entered to interpolate) should tell us the answer. I base it on 568km*137Wh/km estimate, but off by a fraction and it could be off by 0.5kWh.). Where does your 79kWh value come from?
In the exact same use case, driving, charging...all identical...my 2020 Performance saw a drop of the nominal full pack down to 73kWh, while the 2021 Performance saw no change in the nominal full pack at all. Jumping around 80,4kWh since day one.

Good to know. It does seem like @AAKEE is also seeing much better results. Makes me think they fixed their garbage 2170 batteries. We’ll see! Good news for Tesla, if true! 😆

Is there any solid proof that I've missed, suggesting that the LR 82kWh has a capacity limitation in comparison to the performance?

Circumstantial info only; I have not seen a new LR which charges or reports NFP above 80kWh. I don’t know whether this is due to “binning” of the packs (lower capacity 2170L packs go in LRs), or due to software limitation to keep NFP below 80kWh, or if it is due to an insufficient sample of vehicles (I have only seen a couple captures - you have likely seen more). I also have not seen the voltage readout from an LR 82.1kWh pack charged to 100% (that might clear up if it were an artificial cap since the voltage might be lower if so - but they could also limit the low end cutoff). You probably have better info than I do on this - it sounds like you have seen the new packs charged to 100% - but again the (potentially artificial) capacity limit can be imposed by a bottom-end higher voltage limit as well.

I just figured that if there were no software limit we would have seen an 81kWh NFP report from an LR by now. But no idea. If 81.5kWh NFP LR (non-P) vehicles have been captured by SMT then there is clearly not a limitation. Just a lot of variability!
 
Last edited:
Long Range 2021...no more details:
Do we know more about the Panasonic Battery Pack with an R at the end?

2021 M3L VIN : 913XXX
배터리 : 1104422-00-R
Can I know the capacity with this information?

Is there any solid proof that I've missed, suggesting that the LR 82kWh has a capacity limitation in comparison to the performance?

; I have not seen a new LR which charges or reports NFP above 80kWh. I don’t know whether this is due to “binning” of the packs (lower capacity 2170L packs go in LRs), or due to software limitation to keep NFP below 80kWh, or if it is due to an insufficient sample of vehicles (I have only seen a couple captures - you have likely seen more).

BA 1104422-00-R

It s not only R at the end, but also this "2" at the end of the reference number.
Panasonic signature before this one, on all pictures, for all pack, even old one, was always
1104423- (at my knowledge)
STYLE is different too : never seen BA before.

So for me it seems to be a new Panasonic reference, near previous, but little difference.
@eivissa @wonhy0520 , You confirm it s a LR E3LD, I suppose ?

Like you know reference of performance 2021 82kWh batterie pack is CB 1104423-00-T , and some LR 2021 E3LD has this on. I have picture of that. But no picture of this new panasonic reference 1104422 at this time before this one.
I will ask french forum community to take some picture more on LR E3LD. If someone want to help 😀

It could be interesting to see if there is a lot of this new reference now. Because at the begining of Q2 2021 delivery, i only found CB 1104423-00-T (same M3P battery pack).
 
Last edited:
You are thinking that his constant will drop to 255 Wh/mi as his pack degrades down to 80.6 kWh? Where did you get that 80.6 value and 255 Wh/mi.?
Sorry I did not answer this directly: yes, this is exactly what I think will happen.
Thanks for clarifying. So I think you are saying that there is not only a charge constant but also an energy constant, in this case 80.6 kWh, and apparently this 80.6 value is only calculated from the energy screen, but not found or able to be calculated from the SMT or api data.

I guess I would like to see more definitive evidence of the origin and existence of this "energy constant"

In my case, I can calculate my charge constant to an exact value but I need both obd and api data to be 100% sure of the value. And the value never changes.

However, I don't know exactly how the energy screen calculates its numbers for current efficiency and projected range, even though the projected value in the energy screen is given with decimal point precision from the api, just like rated miles are. It's just not clear exactly how it is derived.

But like I said, maybe @AAKEE will be the proof I need to be convinced, but he might be one of those cases where his battery goes for years without showing any degradation!
I don't get that either. I certainly wasn't one of the lucky ones, as my battery is now at about 15% degradation, and I never did achieve the original rated miles on my car. I am not complaining though, because I know my case is much more in line with the typical experience.
 
I guess I would like to see more definitive evidence of the origin and existence of this "energy constant"

I would too. But it really is not anything new - it’s really just the degradation threshold (display cap, energy constant...whatever you want to call it). When you go below it you lose rated miles.
In my case, I can calculate my charge constant to an exact value but I need both obd and api data to be 100% sure of the value. And the value never changes.

I can calculate my charge constant too. I don’t have the OBD or API data but that is ok. It also never changes for me (~245Wh/mi). And if using the energy screen to calculate it, it never changes for anyone including new owners. Even if they have energy that exceeds the degradation threshold. Because the energy screen scales the energy you have, to make it look like the degradation threshold is the max...which is equivalent to the constant never changing.

So I think you are saying that there is not only a charge constant but also an energy constant, in this case 80.6 kWh, and apparently this 80.6 value is only calculated from the energy screen, but not found or able to be calculated from the SMT or api data.
Yes. But on some (but definitely not all) cars it appears to be close to or the same as the FPWN value.

However, I don't know exactly how the energy screen calculates its numbers for current efficiency and projected range, even though the projected value in the energy screen is given with decimal point precision from the api, just like rated miles are. It's just not clear exactly how it is derived.


I believe the energy screen product is given by the formula I gave earlier. Let’s assume for a second the car energy is below the degradation threshold, for simplicity. Then the relationship between the numbers is:

Proj Range * Recent Efficiency = Buffer * SoC% + Usable Remaining.

Where Usable remaining does not include any buffer. It is the MaxUsable*SoC%.

This is equivalent to the formula we know to be true (I think):

Proj Range * Recent Efficiency /SoC % = FullPack. (Just rewrite it and it is the same as above.)

And if you’re above the threshold there’s a scaling applied to make your Nominal Full Pack appear to be equal to the degradation threshold/ display cap/energy constant.

Could be wrong; this is just my current understanding. At this point, I think I have said the same thing several different ways, so I guess I will stop and see if the data bear this theory out.
 
Last edited:
The 82.1kWh battery has the new 2170L cells. I don’t know if the chemistry is subtly different or whether they just make the cells a little bit denser somehow. I believe they are the same basic chemistry as the other 2170 cells (not LFP), but beyond that I don’t know what they might be doing. I don’t follow the chemistry details.
The 2170L has lower amount of cobalt than 2170.
The later 18650 Tesla and the ”old” 2170 use 10% cobalt. This is already about half cobalt content compared to NMC-cells used by some EV car makers. The standard was 622( part of materials in tents, last 2 means 20% cobalt). I think NMC811 with 10% cobalt is coming( not sure how much these is already in use).
2170L has ”a new chemistry” with lower cobalt than before. Tesla/Panasonic have not disclosed how much the reduction in cobalt is.
Cobalt is used to increase energy density power and keep the structure in the chemistry so basically resist degradation.

The issue with low power at low SOC/cold battery in the 2170L probably cones from the lowered cobalt conent. (Initially Tesla also seemed to limit the voltage at the low point during use, but the lower power seem to still be there).
 
  • Like
Reactions: AlanSubie4Life
This is equivalent to the formula we know to be true (I think):

Proj Range * Recent Efficiency /SoC % = FullPack. (Just rewrite it and it is the same as above.)
I agree. I just find that even when using the very precise numbers from OBD and API, when I use that formula, I get some small variation from the SMT full pack value reading which I don't understand, but that is a minor detail.
 
I agree. I just find that even when using the very precise numbers from OBD and API, when I use that formula, I get some small variation from the SMT full pack value reading which I don't understand, but that is a minor detail.
I think the only variation I have seen is due to the roundings. I did some comparing between the energy app metod and SMT values, and they all was within the rounding error( mainly the percentage that only show whole percent). When the SOC was taken from SMT it gave me virtually the same numbers( before my Nominal Full pack went above 80.5 or so, then I see a missmatch).

When driving, one can aim to hit close to the correct percentage number, either by taking the values as the screen percentage changes down to a new number, then you most probably have about % +0.5% or driving some 2- 3km/1.5-2miles furter after a downshift in percentage before taking the values.
I did this for Alansubie the other day coming from the SuC, only fault was that I had good speed and a little slope so after hitting just below 90% ( used SMT values to try to hit 90% exactly)
I ended up with 90.4% before I had stopped on a parking lot.
 
So does this mean the LR models currently being shipped across North America do indeed have the Panasonic 82kwh battery?
At least some of them do. It would be very nice to know what the degradation threshold/energy cap is from one of these new vehicles, from the energy screen. (@Stan930 ?). Preferably using interpolation method as well for accuracy...so three quick pics would be all we need...

If it is still around 78kWh, then they could potentially just use both battery types. It would not really matter (except for degradation characteristics and future unlocks).

We’ll find out. Hopefully they have switched everything over to the new cells at this point.
 
Last edited:
  • Like
Reactions: travis3000
BA 1104422-00-R

It s not only R at the end, but also this "2" at the end of the reference number.
Panasonic signature before this one, on all pictures, for all pack, even old one, was always
1104423- (at my knowledge)
STYLE is different too : never seen BA before.
1104422 is not new, it's used in North America.
1104423 is used in Europe.
Here's a service bulletin mentioning this difference for an older pack revision:
HV battery SB-20-16-004
 
  • Like
Reactions: FredMt
When the SOC was taken from SMT it gave me virtually the same numbers( before my Nominal Full pack went above 80.5 or so, then I see a missmatch).
You can’t do it anymore to calculate your energy since you exceed the threshold/cap but I do think the Energy -> Trip interpolation method should give you close to three significant figures. It looked like the % lined up pretty well with SMT.
 
You can’t do it anymore to calculate your energy since you exceed the threshold/cap but I do think the Energy -> Trip interpolation method should give you close to three significant figures. It looked like the % lined up pretty well with SMT.
Yes, the cap is crystal clear :)

Yes, It seem to be spot in with normal matemathic rules for rounding.
 
  • Like
Reactions: AlanSubie4Life
1104422 is not new, it's used in North America.
1104423 is used in Europe.
Here's a service bulletin mentioning this difference for an older pack revision:
HV battery SB-20-16-004
Very interesting thank you 😉 I didn't know sorry. So question stay about -R at the end. "New" release (?) but before "T" (?) On my side i will check pack reference of Q2 2021 LR.
 
Last edited:
I base it on 568km*137Wh/km estimate, but off by a fraction and it could be off by 0.5kWh.). Where does your 79kWh value come from?
Yeah, we do the same math then:
2020 LR -> 499 x 152,5 -> 77,1kWh threshold
2021 LR -> 568 x 137,0 -> 77,8kWh threshold
"2022" LR -> 568 x 140,0 -> 79,5kWh threshold??? On SMT it looked like 79kWh Usable remaining was the point at which the maximum range was shown.

Makes me think they fixed their garbage 2170 batteries.
Were they really? The degradation reports by users in Germany with the old 2019/2020 LR/P cars are not bad at all. My capacity in the 2020 P might not have been lost, but the BMS didnt like my charging habbit of 20%-90% every three days or so.
The new 82kWh packs seem to degrade less though or their BMS is just smarter.

have not seen a new LR which charges or reports NFP above 80kWh. I don’t know whether this is due to “binning” of the packs (lower capacity 2170L packs go in LRs), or due to software limitation to keep NFP below 80kWh, or if it is due to an insufficient sample of vehicles (I have only seen a couple captures - you have likely seen more). I also have not seen the voltage readout from an LR 82.1kWh pack charged to 100% (that might clear up if it were an artificial cap since the voltage might be lower if so - but they could also limit the low end cutoff). You probably have better info than I do on this - it sounds like you have seen the new packs charged to 100% - but again the (potentially artificial) capacity limit can be imposed by a bottom-end higher voltage limit as well.

I just figured that if there were no software limit we would have seen an 81kWh NFP report from an LR by now. But no idea. If 81.5kWh NFP LR (non-P) vehicles have been captured by SMT then there is clearly not a limitation. Just a lot of variability!
Well, I have posted screenshots of the new Long Range hitting 80,5kWh Nominal Full Pack. Apart from its cell imbalance, all the voltage and capacity values were exactly identical to my Performance 2021. I have seen a few Performance 2021 models having below 80kWh down to 79,1kWh from the factory. Also we havent seen much SMT data from the new 82 LR. Not even from the Performance actually. AAKEE and me provided some data, but the sample size is pretty low so far. Also the way I charge the car does not balance the BMS to the highest possible capacity in contrast to AAKEE, who does exactly that.

I am pretty sure that in the next weeks we will see a lot more data from the new long range and we will find that the battery pack is reaching the exact same peak voltages, cell imbalance and usable capacity as the performance pack...even having the same sticker glued onto it.

You confirm it s a LR E3LD, I suppose ?
No, I cant. The same user posted the same picture in a German Forum without giving details to which car that sticker belongs to.
 
  • Like
Reactions: FredMt
Were they really?

I mean, they are fine. I only have 10% capacity loss after 100 cycles which seems fine but maybe worse than expected. Just seems like a lot of variability. We’ll see how the new ones do.

On SMT it looked like 79kWh Usable remaining was the point at which the maximum range was shown.

I’d definitely be curious to see these captures. But remember with the way this works, maximum range is always achieved when the pack reaches near max capacity. When above the threshold, the point at which this happens is independent of the threshold for capacity loss. It’ll reach max range at 79kWh, and then later in life at 77.8kWh (for a threshold of 77.8kWh). Also, really want to see that energy screen for a new LR,since pretty sure that tells you the threshold for capacity loss.

The only question that is being answered is when the vehicle will SHOW capacity loss in rated range. It’s not that important but it is interesting to me since it implies what minimum capacity they can ship with. The battery could start at 81kWh and that loss threshold where loss (in miles/km) starts to show could be as low as 75kWh (it is not that low - probably closer to 77.8kWh but no idea); it is really up to Tesla.

I am pretty sure that in the next weeks we will see a lot more data from the new long range and we will find that the battery pack is reaching the exact same peak voltages, cell imbalance and usable capacity as the performance pack.
Hopefully! Thanks for the other data - I will go back and review your posts - I tend to forget things...
 
Last edited:
  • Like
Reactions: FredMt and eivissa