-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Predicting Battery Sizes
- Thread starter ChiefRollo
- Start date
scottf200
Well-Known Member
scottf200
Well-Known Member
FYI from GhostAndSkaterr post:
"More info we can extrapolate from the EPA data for the Cybertruck
Pretty much confirmed 220S6S battery, or 1320 cells"
"More info we can extrapolate from the EPA data for the Cybertruck
Pretty much confirmed 220S6S battery, or 1320 cells"
AlanSubie4Life
Efficiency Obsessed Member
Only problem (with this assuming 1366 is actually right), is how are those 46 cells configured for the 48V?
I could not make it work.
Lots of assumptions here of course.
Lots of signs pointing to the 123kWh rather than 129kWh but as discussed that is weird too.
scottf200
Well-Known Member
AlanSubie4Life
Efficiency Obsessed Member
Yeah that all makes sense, and 220s6p seems quite conceivable as I said elsewhere. It certainly seems like 6p is for sure, given the approximate voltage (around 924V - 940V max).
Only minor issues:
a)
As I said, 1366, if it is the correct number, doesn't seem to work with this. That leaves 46 cells (1366-220*6) for the 48V system. Only choice is 23s2p which obviously doesn't work at all. Obviously the 1366 number could just not include the 48V system at all, though (48V done with prismatic cells or whatever), and the remaining 46 cells are blanks or something. Or 1366 is just wrong.
b)
110s is called "400V" for the Plaid. So why would 220s be 816V? (Or was it 410V? Can look it up…)
c)
224s6p works better for the "48V inside the pack" issue. (224s6p (main pack) + 11s*2p (48V))
For modules I am not sure which one works better without detailed dimensions of cooling tubes, etc. Model Y 4680 pack was 92s9p. 4 modules of 23 series alongside each other lengthwise (4*23 = 92).
Dimensions: Length: 23 series cells ends up about 35 cells long (2/3 of 35 is about 23) because they're tied in staggered groups of 9 (of width 6!!!), shown here. Width: Ends up being width 6 for 9 cells in parallel. (6 in parallel as in CT will be 4 cells in width)
110 factors as 2*5*11. 112 factors as 2*2*2*2*7.
110s:
So you only have the choice of 5 modules really per each sub pack with 110s (you could do 2 as well but then you end up with very long and skinny). Each module of 22 series cells would be about the same length as the Model Y; about 33 cells long rather than 35. But width would be too skinny; just 4*5 = 20 cells wide compared to Model Y 24 cells wide. Won't work; we need about 28-32 cell width!
This could be rotated and placed perpendicular to travel direction. 33 series cells might just fit across CT with margin?
So then you'd have 5 modules placed crossways across the truck, times two. Coolant flow might not be ideal this way?
But I think it could work, without going through detailed dimensions.
112s:
Go with 8 modules of length 14.
Width is going to be 4*8 = 32 cells (each 6p module is 4 cells wide). Might work if they make the cooling tubes & pack overhead tighter (otherwise will be too wide). Compare to 24 cells wide in Model Y, plus overhead.
Length of each sub-pack is about 21 cells (14s cells goes to 21 length; in Model Y 23s was about 35 cells long). So two end to end would be 42 cells in length (Cybertruck) vs. 35 cells in length (Model Y). That seems like it would fit easily (wheelbase is ~150 inches vs. 113 for model Y)
114s: (works out to be 1368, 114s*6*2). 48V is not composed of 4680s. 1366 is wrong (should have been 1368).
Could work with modules, six modules of 19s6p. Length about 29 cells, could be laid across the vehicle, which would fit the width (maybe too short). The width of the sub pack would then be about 24 cells (6 of 4-wide rather than 4 of 6-wide), similar to Model Y.
I think this ends up being too narrow and too long for the truck?
Anyway hard to know which one works from a practical standpoint without a good model of the dimensions. If you have that it is possibly easier to see which one works better. There is that other thread you pointed to. They can probably work it out.
Not all modules have to be the same - another complication not covered here. So many options.
Last edited:
I don't have anything to say about the rest of the issues, but the palladium S docs say 410v, palladium X says 408v
S: https://dis.epa.gov/otaqpub/display_file.jsp?docid=57010&flag=1
X: https://dis.epa.gov/otaqpub/display_file.jsp?docid=56731&flag=1
In theory both cars should have the same batteries in reality. So 220s should work for 816v.
Last edited:
AlanSubie4Life
Efficiency Obsessed Member
The existence of 110s already does seems to speak to them just doing that again. The only issue is that might work fine for 18650 but maybe the dimensions don’t quite work out right for 4680.
I assume the 110s is done with 5 modules? I guess if so they just duplicate that.
It does make sense. Seems like 220 probably works. The dimensions I mention in that post just must work out.
But then there is talk about there being four “200V” pieces.
Who knows. It’s like reading tea leaves because it is hard to tell what info (1366 cells, four 200V packs) is correct.
GhostSkater
Member
First thing I did was try to line up 48 V LV batteries with that number, even two 48 V batteries for redundancy, nothing lined up so 1366 must be wrong and also not including the 48 V LV battery, then I started expanding the search to see what would fit, and in the end 220S6P lines up perfectly, specially with EPA data out now
220S7P would also work if not for the EPA data, because with 122.4 kWh, it would mean each cells is 79.48 Wh, which in considerably worse than gen 1 4680s and doesn't line up with Drew comments of 10% higher energy density
It seems like the gen 2 4680s are 92.7 Wh, or 7% improvement in energy. If the 10% comment is accurate, it also means the cells are 344 g, vs 355 g before, so 269 Wh/kg
AlanSubie4Life
Efficiency Obsessed Member
Would be cool to see a footprint of how it fits, using dimensions and approximate structure from Model Y AWD. Relatively easy once you have the building blocks of a diagram (Model Y pack size and layout is well understood), but I don’t have time for that!
Model Y uses only top cell connections, I think.
Why not 224s6p? That leaves 22, 11s2p, for the “48V” if that design (4680s for the 48V) is being used. It’s close to 48V!
Anyway doesn’t matter, we’ll find out eventually.
GhostSkater
Member
Oh so you mean the 48 V pack integrated into the pack? That is a nice idea, don't know if good or bad in terms of redundancy, specially with the steer by wire, might be good to have two power sources, as in, if the HV pack fails catastrophically, the 48 V LV battery is in another location safely providing steering, and if the 48 V LV fails catastrophically, the DC/DC converter can provide 48 V for the steering
Or there might be two LV batteries for redundancy
For pack layout, each of the 4 "modules" have to have a pair number of row for the cooling channels, 6 cells like on the 4680 Model Y seems too narrow, but possible
8 would mean a pack that is too square and doesn't fit the pictures we've seem, so a Model Y arrangement but longer seem plausible, but maybe too long
AlanSubie4Life
Efficiency Obsessed Member
It’s (MY) 6 cells wide per module with a cooling snake down each bandolier pair. That gives 9 cells parallel due to wiring pattern using 3 cells for each 2 cells making up the width.
Cybertruck if using same collector wiring scheme would be 4 wide per “sub module” with two cooling snakes.
But there are so many permutations and possible enhancements it is hard to know what we will get.
Yeah I think biggest issue is HV isolation when sharing coolant loop, but still seems like there might be ways with heat exchangers or something. Best part is no part!
I’m not suggesting it is done that way. We just have to wait and see.
GhostSkater
Member
Not even worthy trying to draw since hopefully we will see one tire apart pretty soon
I don’t think they will have cooling for the LV battery, even if it’s inside the pack
Or better put, if the 48 V needs cooling it will be inside the pack, which honestly make sense because I tried but couldn’t spot a LV battery in any of the pics, not even where it would go
Could Tesla have ditched it altogether? Best part is no part, it’s possible to design a DC/DC converter that is efficient in high and lower power drain, as in vehicle at idle vs steer by wire working at full power
Not totally related, but I worked with some inverters that have stand by current draw lower than the self discharge of a lithium cell, different scenario because stand by meant zero output, not a few 10s or 100s W at idle
AlanSubie4Life
Efficiency Obsessed Member
For heating primarily. I have no idea how they manage the prismatic cells in current “12V” but maybe they do not need to be above freezing to charge. Or it heats itself. No idea. It’s not probably necessary to share a loop as long as it can heat itself somehow. (If using 4680 NCA.)
I think the big issue with entirely ditching it is ensuring that complete HV isolation is possible. For that you need to be able to open/close contactors which requires a power source.
GhostSkater
Member
Good points, I had seen the datasheet for the cells they use there floating around a while ago, and there was a allowable charging current even bellow freezing, different than most cells
That is true, if you lose isolation and have to open contactors you lose LV power
Yep, and with the steering being entirely dependent on the 48v system being alive, they need to make it so that you still low voltage available even if the HV system goes down.
