Cybertruck will be 800V

[AlanSubie4Life]

Also makes me wonder whether Model Y 4680 is set up the same way and will also get a boost from V4 800v charging.

It’s not. It’s 92s9p.


If they get enough 350kW-capable stations it seems like there are plenty of vehicles (possibly not old Model Y AWD) which would get marginal benefit if they increase the rate. Don’t want mess it up!

Cybertruck definitely can benefit. And Model S should obviously charge faster at this point.

There’s no doubt that CT can charge faster, but they have to be conservative and careful about the high power stuff.

It’s going to be configured as (most likely) 110s12p when configured for 400V, so it can charge at 33% higher current at 20% higher voltage, so minimum should be capable of 60% higher rate than Model Y AWD at 400V chargers that are not limited by current or voltage - and that is assuming the v2 “Cybercell” is at least as capable as v1 (and it should be slightly more capable, assuming rational design, since it is higher capacity by at least 10%).

 

[JulienW]

So one day the CT may reach the level of a "children's toy".

 

[mongo]

Actually, I was hoping that the V4 superchargers would have a max charge rate of 500kW or better. After all, if the voltage is doubled from 400 to 800, they should be able to deliver twice as much power as the V3s. From Lars's statement, it sounds like they're going to be 350kW, at least at first. Or maybe with the Cybertruck's 123kW battery pack, the max rate that the pack will be able to charge at is 350kW. I guess we'll have to wait and see.

One v3 cabinet can only pull 350kW from the grid. Connected to two others on the 1kV DC link, it can output 900kW or so.
The same basic cabinet powers the Semi chargersn 2 oer ar Pepsi, 4 per post at Sparks. Either all of them can do 1kV and are software configured not to, or the higher voltage modules are a manufacturing time selection.
V3 cabinet + V4 posts = 350kW Cybertruck charging @ 800V.

Would someone explain how the Cybertruck battery arrangement works for 400V and 800V charging?

Jason Cammisa explained:

Four 200V units are arranged in parallel for 800V operation for all 3 motors. Is that a typo? 200V in parallel should yield 200V and 200V x 4 in series would yield 800V shouldn't it?

For 400V charging, there's a switch to take them all from series to parallel: I assume one 200V in series with another 200V to get 400V and do the same with the other 2 units, then combine those 400V in parallel for 400V charging when 800V V4 stations are not deployed?

Is the switch mechanical or electronic?

Mostly mechanical, there is motor on the bottom of that shot that actuates the two switches.

It’s a typo. Four in series for 800 volts (aka 920V). Split in half 2 series, 2 parallel for 400V charging.

Max voltage of V3 Supercharger is allegedly 480V so 960V pack will work.

Switches look mechanical - like contactors I would think?

This four module thing is interesting. That might speak against the 220s pack. Because that would require a 55s pack if there really are four 200V elements. That 55s pack is too long for a single module I think (more than double the length of Model Y pack). So would have to be 5 modules of length 11 which seems odd.

Anyway we’ll see.


Predicting Battery Sizes

Y modules are 3 bandoliers of 69 cells for 207:23s8p
55s6p could be 4 bandoliers of 82.5 cells (2 83, 2 82?). Y has padding on the sides, so that might fit.

They called out 4 modules to do the 400/800 conversion.

 

[AlanSubie4Life]

207:23s8p

23s9p you meant…

I’ve given up on speculating (lol). But since they grab groups of three from the bandolier, if they do the same as done for Model Y, the size of the bandolier would be 3x the number of series cells in the module. So 165 if there are 55s (this would be very long:82.5 cell pitches). Or 33 if there are 11s.

Yeah if there are four equal modules (sub-packs?) they’d be 55s. But could be composed of “submodules” of lower length. I don’t think 55s fits if using Model Y scheme. It’s 82.5 cell pitches long minimum (Model Y pack is about 34.5 cell pitches long for comparison) and that won’t fit unless it goes beyond the wheelbase.

Obviously they could use some different scheme for wiring on Cybertruck. (Tap groups of 6 from a bandolier for example.) I think the issue then is you can’t easily connect to the negative terminals on the next brick with the current collector. But there are probably SOME other options - have not thought about the geometric limitations.

And of course 55s can be broken into 5x11s.

There are other options if turns out to be 56s/112s/224s. No idea what actually works.

 

[Mrbrock]

C-rate already takes into account the size of the battery. If you charge a 50 kWh battery at a 2C c-rate, that means you're charging at 100kW. To charge a 100 kWh battery at the same c-rate, it needs to charge at 200 kW.

15 minute average charging c-rates = (kWh added in 15 minutes * (60/15)) / battery capacity kWh

Model S / X (Panasonic 1865) = (47 kWh * (60/15)) / 100 kWh = 1.9C
Model 3 / Y LR / P (Panasonic 2170) = (38 kWh * (60/15)) / 82 kWh = 1.9C
Model 3 RWD (CATL LFP) = (29 kWh * (60/15)) / 60 kWh = 1.9C
CyberTruck (4680) = (49 kWh * (60/15)) / 130 kWh = 1.5C

I understand but please stop. This discussion wasn’t with you and your c rate has nothing to do with gtx (not gtg, you) commenting on the Cybertruck having the worst charging characteristics of any Tesla vehicle because it’s 40% replenishment time was too long. Just because the time happened to be 15 minutes doesn’t mean gtx was talking about c rate. There was nothing mentioned about c rate. At no time was any comment directed to you (gtg) unless you (gtg) also post as gtx in which case this is even more ridiculous.

And then you thumbs downed and disagreed with me saying you needed actual energy to determine c rate. not just percent. In all of your equations, what do you call 47kwh, 100kwh? That’s energy. Not energy rate but energy storage or energy amount. Without kw or kWh, you can’t calculate c rate.

And the C rate on It’s own doesn’t tell you Jack crap about the charge rate other than it is X times the battery capacity. If you don’t know the battery capacity it means nothing. Knowing that the Cybertruck charges 40% in 15 minutes doesn’t tell you anything unless you know battery capacity or the equation of the charge curve or average kw of the 40% charge session.

Take this example. Company X has a new vehicle coming out with a 0.5C rate. Would you say that is horrible? Without any other known information you cant comment. Now if I said over that 0.5C, it averaged 400kw, then you can say the battery is 800kw, a full charge is 2 hours and it is actually quite a fast charging vehicle. You need more than just the c rate.

 

[Mrbrock]

One v3 cabinet can only pull 350kW from the grid. Connected to two others on the 1kV DC link, it can output 900kW or so.
The same basic cabinet powers the Semi chargersn 2 oer ar Pepsi, 4 per post at Sparks. Either all of them can do 1kV and are software configured not to, or the higher voltage modules are a manufacturing time selection.
V3 cabinet + V4 posts = 350kW Cybertruck charging @ 800V.



Mostly mechanical, there is motor on the bottom of that shot that actuates the two switches.


Y modules are 3 bandoliers of 69 cells for 207:23s8p
55s6p could be 4 bandoliers of 82.5 cells (2 83, 2 82?). Y has padding on the sides, so that might fit.

They called out 4 modules to do the 400/800 conversion.

How are the V3 cabinets outputting 800V? The DC DC bus is still 350/400v between cabinets. The AC to DC converters output at the 350/400V. I am unaware that the cabinets can combined and step up 400v to 800v. That equipment and capability is what will determine V4 cabinets. If line voltage was 960V then it would be easy since it would still be a step down but it’s 480 so it changes the process.

 

[mongo]

How are the V3 cabinets outputting 800V? The DC DC bus is still 350/400v between cabinets. The AC to DC converters output at the 350/400V. I am unaware that the cabinets can combined and step up 400v to 800v. That equipment and capability is what will determine V4 cabinets. If line voltage was 960V then it would be easy since it would still be a step down but it’s 480 so it changes the process.

The intercabinet DC bus is specified as 880V-1000V.
First section of the AC power input is a power factor correction (PFC) boost stage.

Interesting, that's definitely different. And as expected it also has a different part number. Here's one with "4 Output", but I don't recall which specific location it comes from:

View attachment 658867

 

[Nack]

When it's engineered properly, a DC/DC converter is as reliable as any other electrical device. I'm sure that Tesla could easily design their own to work on the CT. Standard electrical engineering would call for the converter to be on its own circuit so that if something went wrong with it, it would just trigger a circuit breaker for that one circuit.

Not in an automotive or commercial vehicle environment. The DC/DC is a weak point for sure. I have been in many FMEA/FuSa where this is discussed in relation to powering safety critical devices with power on a bus that's supplied by a DC/DC converter. Keep in mind if something goes wrong with the converter and that circuit is dropped....then you lose safety features like steering or brakes, etc.

There are lots of DC/DC converters in the various ECUs obviously, but rarely to power a system bus. It could work as long as that system bus did not control anything besides convenience features.

We can't think of vehicles like simple smartphones or gadgets. If something goes wrong with that...oh well. Vehicles can kill people when stuff goes wrong. You can't send the vehicle an OTA update to make someone live again.

 

[Nack]

We kinda get hung on voltage but in reality there is no such thing as actual pack (or any) voltage. The voltage is always dynamically changing. For instance a Lit-ion cell can vary from 3V at low charge to 4.2V at a full charge. Also we know that the CT is a split pack so in the filing that would be 408V each. So 408V could be 100 (or couple more or less) in series which would be between about 300V to about 420V with 408V being nominal.

Can you point me where it is confirmed its a split pack? I am curious. Is the Semi a split pack as well as they were supposed to share a lot of similar architecture and in the Leno video they say "near 1000V max." The semi stuff is still fairly secret I think so its probably not known for sure.

You are correct the pack voltage varies throughout the SOC curve which is mostly like a sigmoid function. At really low and really high SOC the voltage changes a lot with slight increase, in the middle, its roughly constant. Generally you refer to the pack voltage and the system with nominal voltage from my experience as this is what voltage level you are at majority of the time.

 

[mongo]

Can you point me where it is confirmed its a split pack? I am curious. Is the Semi a split pack as well as they were supposed to share a lot of similar architecture and in the Leno video they say "near 1000V max." The semi stuff is still fairly secret I think so its probably not known for sure.

You are correct the pack voltage varies throughout the SOC curve which is mostly like a sigmoid function. At really low and really high SOC the voltage changes a lot with slight increase, in the middle, its roughly constant. Generally you refer to the pack voltage and the system with nominal voltage from my experience as this is what voltage level you are at majority of the time.

Semi doesn't split.
Cybertruck can per Drew and the switch was shown on the Carmudgeon video.

https://twitter.com/x/status/1730391023799386480

 

[mongo]

Not in an automotive or commercial vehicle environment. The DC/DC is a weak point for sure. I have been in many FMEA/FuSa where this is discussed in relation to powering safety critical devices with power on a bus that's supplied by a DC/DC converter. Keep in mind if something goes wrong with the converter and that circuit is dropped....then you lose safety features like steering or brakes, etc.

There are lots of DC/DC converters in the various ECUs obviously, but rarely to power a system bus. It could work as long as that system bus did not control anything besides convenience features.

We can't think of vehicles like simple smartphones or gadgets. If something goes wrong with that...oh well. Vehicles can kill people when stuff goes wrong. You can't send the vehicle an OTA update to make someone live again.

DC-DC isn't an issue, the problem with relying on it solely is that if you lose the HV pack, the converter is useless and you have no low voltage.

 

[GregD60]

Not in an automotive or commercial vehicle environment. The DC/DC is a weak point for sure. I have been in many FMEA/FuSa where this is discussed in relation to powering safety critical devices with power on a bus that's supplied by a DC/DC converter. Keep in mind if something goes wrong with the converter and that circuit is dropped....then you lose safety features like steering or brakes, etc.

There are lots of DC/DC converters in the various ECUs obviously, but rarely to power a system bus. It could work as long as that system bus did not control anything besides convenience features.

We can't think of vehicles like simple smartphones or gadgets. If something goes wrong with that...oh well. Vehicles can kill people when stuff goes wrong. You can't send the vehicle an OTA update to make someone live again.

By definition, any DC/DC converter that isn't reliable for it's application isn't properly engineered. I don't know what vehicles you're referring to, or what the applications are, but DC/DC converters are used everywhere including mission critical applications in airplanes and spacecraft. Note that these applications have batteries as well, since you always want a backup for anything mission critical.

I know that the heat pump in Teslas uses high voltage to drive it, but is it the full battery pack voltage or a lower voltage? I've heard some people say that the CT's steer by wire system uses 48V, and that was part of the motivation for Tesla to go to a 48V low voltage system. Pretty much all EVs have DC/DC converters to charge the batteries for their low voltage systems. I haven't heard of any manufacturer having any particular issue with these converters.

 

[GregD60]

DC-DC isn't an issue, the problem with relying on it solely is that if you lose the HV pack, the converter is useless and you have no low voltage.

Which is why it's normal to have a battery for the low voltage supply on a vehicle. For the application that initially started this discussion, how to power 12 volt accessories on the CT, a battery isn't necessary since these aren't critical applications.

 

[JulienW]

Which is why it's normal to have a battery for the low voltage supply on a vehicle. For the application that initially started this discussion, how to power 12 volt accessories on the CT, a battery isn't necessary since these aren't critical applications.

....but in ALL EV's the 12V (48V battery in CT) are charged and supplemented by a DC-DC converter from the high voltage pack. So all EV's contain DC-DC converters.

 

[terranx]

....but in ALL EV's the 12V and 48V battery in CT is charged and supplemented by a DC-DC converter from the high voltage pack. So all EV's contain DC-DC converters.

The risk tolerance is different. If a DC-DC converter that's used to charge the LV battery fails, it doesn't mean immediate loss of mission critical safety systems. The LV battery will still power those things more than long enough to pull over to a stop safely.

If there's no LV battery and you're depending on a DC-DC converter for all LV systems, then a failure means immediate loss of steering, ABS, stability control, etc.

 

[GregD60]

....but in ALL EV's the 12V (48V battery in CT) are charged and supplemented by a DC-DC converter from the high voltage pack. So all EV's contain DC-DC converters.

Which is what I stated in the post prior to the one you responded to. And these DC/DC converters have been very reliable from everything that I've heard.

 

[GregD60]

The risk tolerance is different. If a DC-DC converter that's used to charge the LV battery fails, it doesn't mean immediate loss of mission critical safety systems. The LV battery will still power those things more than long enough to pull over to a stop safely.

If there's no LV battery and you're depending on a DC-DC converter for all LV systems, then a failure means immediate loss of steering, ABS, stability control, etc.

In my experience, having a battery backup is pretty much standard on any mission critical application. Certainly steering, ABS, stability control, etc., would be considered mission critical on a car, and there will be a battery on the power supply that can drive them directly.

 

[ItsNotAboutTheMoney]

So one day the CT may reach the level of a "children's toy".

View attachment 996751

For what it's worth the "children's toy" comment was about the _chargers_, NOT the _charging_. People jumped on it thinking it was about the charging.

The V3 architecture allows output of a _lot_ more than 350kW. The cabinets are more limited on the _input_ side.

The approach used by almost other networks/charging companies has a 350kW input _and_ output charger, and no power sharing.

I've had to wait to charge my Kona (max 200A) at Electrify America because a Bolt (max 150A) was charging. The chargers have 2 CCS plugs.

For people who're worried about the onslaught when Tesla opens the network, remember that Supercharger sites have more stalls than power for good reason.

 

[gtg465x]

I understand but please stop. This discussion wasn’t with you and your c rate has nothing to do with gtx (not gtg, you) commenting on the Cybertruck having the worst charging characteristics of any Tesla vehicle because it’s 40% replenishment time was too long. Just because the time happened to be 15 minutes doesn’t mean gtx was talking about c rate. There was nothing mentioned about c rate. At no time was any comment directed to you (gtg) unless you (gtg) also post as gtx in which case this is even more ridiculous.

Who the hell is gtx? I went back through the thread and saw no one by that name.

And then you thumbs downed and disagreed with me saying you needed actual energy to determine c rate. not just percent. In all of your equations, what do you call 47kwh, 100kwh? That’s energy. Not energy rate but energy storage or energy amount. Without kw or kWh, you can’t calculate c rate.

Sure you can calculate c-rate with percent... if the battery charges to 100% in 1 hour, that's a c-rate of 1C, 100% in 30 minutes is 2C, 50% in 15 minutes is also 2C.

And the C rate on It’s own doesn’t tell you Jack crap about the charge rate other than it is X times the battery capacity. If you don’t know the battery capacity it means nothing. Knowing that the Cybertruck charges 40% in 15 minutes doesn’t tell you anything unless you know battery capacity or the equation of the charge curve or average kw of the 40% charge session.

You're wrong about c-rate not telling you anything. At the very least, c-rate tells you how well the cells perform in comparison to other cells. You may not be able to say with 100% certainty that a car with lower c-rate charging adds less miles per time than another car with higher c-rate without knowing any other details, but generally, it's a pretty good indicator of what charging performance is going to be like.

And if all we knew was c-rate, maybe I wouldn't have claimed that the charging performance wasn't good. But that's not all we know. We also know how many rated miles are added in 15 minutes, and sure enough, it's worse than every other Tesla apart from the Model Y with the same 4680 battery cells.

Take this example. Company X has a new vehicle coming out with a 0.5C rate. Would you say that is horrible? Without any other known information you cant comment. Now if I said over that 0.5C, it averaged 400kw, then you can say the battery is 800kw, a full charge is 2 hours and it is actually quite a fast charging vehicle. You need more than just the c rate.

Maybe you think it's good, but I would never want to road trip any EV that takes 2 hours to DC fast charge. Physics dictates that this vehicle is going to use a proportionally large amount of energy to move, so even if it adds 100 kWh in 15 minutes, it's still not going to go very far on that 100 kWh unless it defies the laws of physics and is magically an order of magnitude more efficient than other EVs. A Tesla semi has a rumored 900 kWh battery, but can only go up to 500 miles, so if it charged at a 0.5C rate, that means it would add 62 miles of range in 15 minutes, which is still bad, even though it added a whopping 112 kWh. Tell me that a vehicle added 112 kWh in 15 minutes and I couldn't tell you if it's the best charging EV ever or the worst, but tell me that it charges at 0.5C and I can make a very solid guess that it's not going to perform very well in terms of miles added per time.

 

[Nack]

FWIW the batteries usually have different maximum C rate capabilities over the SOC of the battery and cell temperature.

Maybe its well known here and we are just using the avg C rate or something for approximation.