Sponsored

electruck

Well-Known Member
Joined
Oct 6, 2019
Threads
69
Messages
3,495
Reaction score
6,454
Location
Dallas, TX
Vehicles
2023 Rivian R1S
Even GMC's own press room has conflicting information so it's no wonder the general press can't keep things straight. This could be a reflection of plans that have changed over time.

From March 4, 2020:
https://media.gm.com/media/us/en/gm/home.detail.html/content/Pages/news/us/en/2020/mar/0304-ev.html
Ultium-powered EVs are designed for Level 2 and DC fast charging. Most will have 400-volt battery packs and up to 200 kW fast-charging capability while our truck platform will have 800-volt battery packs and 350 kW fast-charging capability.
And from October 21, 2020:
https://media.gm.com/media/us/en/gm...20/oct/1021-hummer-edition-1-performance.html
HUMMER EV will have the ability to take advantage of the industry’s fastest 350-kilowatt DC fast chargers4. This vehicle will offer a GM-estimated 350+ miles7of range based on preliminary testing, with a unique ability to switch its battery pack from its native 400-volts to 800-volts for charging. A disconnect unit and mechanization within the pack enables the battery to switch from “parallel” to “series,” allowing it to add nearly 100 miles of range in 10 minutes of charging8.

The Ultium battery system will control the electrical currents at the charging station via clever algorithms that ramp down the current and switch to the other half of the pack, when necessary, and increase the rate of charge. The charging system is also designed to support legacy 400-volt charging infrastructure, without the need for a converter box or other accessories, allowing HUMMER EV to use a variety of charging stations.
EDIT: Do the above quotes actually reflect a change? Or is the latter simply more precise in its wording?
Sponsored

 
Last edited:

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
I haven't read the patent. Hopefully it is about more than just serial vs parallel circuits. Did they offer anything "novel" in terms of how the circuit might be dynamically reconfigured?
There are 30 claims! In essence it's just connecting them in series for charging and parallel for running. There is tons of patenese in there "and embodiment in which a single pole double throw switch... and "and embodiment in which the single pole double throw swtich is controlled by a controller..." and on and on. A a patent attorney acquaintance of mine once said "The job of a Jedi patent attorney is to claim as much as possible while revealing as little as possible". There may be something "not obvious" in some of the fault detection/BMS claims but the basic idea seems pretty obvious to me and thus would not be patentable.

The only thing I thought at all clever was the notion that one of the two halves should remain connected to the car while charging in order to support all functions normally supported during charging.
 

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
I have read a couple of other articles that dispute this and claim an 800V system is used. I am not 100% sure this guy is correct.
No one will be 100% sure until he sees the guts but I have never known this particular guy to be wrong. And he did thank GM for letting have a preview peek.
 

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
Even GMC's own press room has conflicting information so it's no wonder the general press can't keep things straight. This could be a reflection of plans that have changed over time.

From March 4, 2020:
https://media.gm.com/media/us/en/gm/home.detail.html/content/Pages/news/us/en/2020/mar/0304-ev.html


And from October 21, 2020:
https://media.gm.com/media/us/en/gm...20/oct/1021-hummer-edition-1-performance.html


EDIT: Do the above quotes actually reflect a change? Or is the latter simply more precise in its wording?
I would hardly call the second statement precise. The first makes sense to me. The second does not.
 

electruck

Well-Known Member
Joined
Oct 6, 2019
Threads
69
Messages
3,495
Reaction score
6,454
Location
Dallas, TX
Vehicles
2023 Rivian R1S
I would hardly call the second statement precise. The first makes sense to me. The second does not.
Engineer speak paraphrased by marketing is rarely "precise"... but they give it their best.

The first quote might lead some to jump to the conclusion that there are distinct 400V and 800V system architectures. However, the second quote seems to expand on the first by stating the input voltage to the pack may be switched between 400V and 800V with the 2 halves of the pack operating at 400V either in parallel or series depending on the input to the pack. The reference to the "native" 400V pack would lead to a reasonably logical conclusion that the system architecture is 400V. My point was simply that the 2 quotes really aren't at odds.
 

Sponsored

DucRider

Well-Known Member
First Name
Gary
Joined
Oct 21, 2019
Threads
17
Messages
1,652
Reaction score
3,143
Location
ORegon
Vehicles
Polestar 2, Ioniq, R1S
Engineer speak paraphrased by marketing is rarely "precise"... but they give it their best.

The first quote might lead some to jump to the conclusion that there are distinct 400V and 800V system architectures. However, the second quote seems to expand on the first by stating the input voltage to the pack may be switched between 400V and 800V with the 2 halves of the pack operating at 400V either in parallel or series depending on the input to the pack. The reference to the "native" 400V pack would lead to a reasonably logical conclusion that the system architecture is 400V. My point was simply that the 2 quotes really aren't at odds.
When charging, they would be connected in a way that for all practical purposes it was an 800V pack. The real question is does it operate at 800V when powering the motors?

The second statement has a bunch of detail implying not.
 

electruck

Well-Known Member
Joined
Oct 6, 2019
Threads
69
Messages
3,495
Reaction score
6,454
Location
Dallas, TX
Vehicles
2023 Rivian R1S
When charging, they would be connected in a way that for all practical purposes it was an 800V pack. The real question is does it operate at 800V when powering the motors?

The second statement has a bunch of detail implying not.
Exactly. My point however was that without the details contained in the October press release, the March press release left some things open to interpretation and some may have jumped to the conclusion that the pack input and output voltages were the same (for both 400 and 800 volt packs) as I believe that has been the norm until Rivian and now GMC.
 

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
I just had a real Homer Simpson style "Doh!" moment concerning the difference between a 800V motor and a 400V motor. This came about from flicking through the Grainger cataloge motor section. Is you building wired for 230V and you need a 12HP motor? That's a catalog 5PGK4. Suppose you open a new facility that is wired for 460V and need another 12 HP motor for that. Well it's catalog 5PGK4 again. It's the same motor! The main difference between the two voltages is that in the second building you can wire this motor with AWG 12 whereas in the first building you would have had to have used AWG 8.

Readers who have worked with motors know how this works. For the others: a 230V/460V motor has a pair of 230V windings on each pole. These are connected in series when used at 460V. When used at 230V they are connected in parallel.

Thus if we suppose that Rivian (and Tesla) originally set out to design 400 V cars they would have developed (or had developed for them) 400V motors to run in them. As the push for faster charging has become a bigger and bigger part of the competitive picture the necessity for higher voltage battery packs has become clear. Does this require a higher voltage motor? No. You can reconfigure the battery pack in series for charging and parallel for operation. But that requires a kluge of switches. So why not go back to your motor lab and tell them to wind your motor for double the voltage and your inverter lab to design a 2*V inverter/rectifier etc.? It isn't something they can do overnight. They do have to think about getting more turns of finer wire into the slots and the insulation has to be able to withstand higher voltage. But it does not mean starting with a clean piece of paper either.

The Ultium battery system will control the electrical currents at the charging station via clever algorithms that ramp down the current and switch to the other half of the pack, when necessary, and increase the rate of charge.
Still can't figure that one out. All vehicles control the current from the charging station. What other half?


The charging system is also designed to support legacy 400-volt charging infrastructure, without the need for a converter box or other accessories, allowing HUMMER EV to use a variety of charging stations.
If the pack's "native configuration" is 400 V that shouldn't be much of a challenge.

The motors are intrinsically 400/800 and the battery packs are intrinsically 400/800 and high voltage solid state swtches are reliable and cheap so that an OEM has lots of flexibility in what it can do and its marketing department has lots of flexibility in the gobbledy-gook they can publish into trying to dupe consumers into thinking their product stands at the front of the class.
 

electruck

Well-Known Member
Joined
Oct 6, 2019
Threads
69
Messages
3,495
Reaction score
6,454
Location
Dallas, TX
Vehicles
2023 Rivian R1S
The motors are intrinsically 400/800 and the battery packs are intrinsically 400/800 and high voltage solid state swtches are reliable and cheap so that an OEM has lots of flexibility in what it can do and its marketing department has lots of flexibility in the gobbledy-gook they can publish into trying to dupe consumers into thinking their product stands at the front of the class.
All true. Lucid also suggests that they are also able to optimize for efficiency by leveraging a high voltage powertrain. I think this is where the real distinction between "native" 400 or 800 volt architecture comes into play. Rawlinson would likely argue that switchable 400/800 achieves the faster charging objective but leaves some efficiency on the table.

https://spectrum.ieee.org/cars-that...-motors-says-ecar-design-all-about-efficiency
 

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
Well he would argue, based on the IEEE article you linked, that the 400V battery is less efficient because it must supply twice the current and that the losses to the internal impedance of the battery are quadrupled. This is the IEEE and they didn't challenge him on that!

Suppose you have two 4.2V cells in your hand and you need to get 8.4 W to drive a motor. First you connect the cells and motor windings in series so that the battery is delivering 8.4V at 1 ampere i.e. 8.4 watts. The internal impedance of the cells is 1 so that I^2R in each cell is 1*1 = 1. There are 2 cells so the total I^2R loss is 2. Now you connect motor winding and batteries in parallel for a 4.2V system. Obviously the motor now draws 2 amps, 1 from each cell. The cell current is thus the same and the I^2R losses the same. Mr. Rawlinson has exhibited this kind if disingenuity before.

So better battery efficiency is BS and better motor efficiency is BS leaving the inverter. To be honest I don't know. I don't doubt that SiC is a better semiconductor for switching than IGBT but I am not personally aware of how a high voltage inverter would be more efficient than a low voltage one. If I look at load lines for high and low voltage switches casually it looks as if losses would be about the same. You'd have to know much more about the designs of the inverters and the devices they use than I do in order to be able to give a sound answer on this. That still leaves wiring and there is no doubt that the buses running around the vehicle benefit from higher voltage in that they can be half the weight at twice the voltage.

My perspective on this keeps growing as discussions like the ones here take place but I find I keep coming back to the same place: the primaary advantage of higher voltage is the ability to charge faster with secondary benefit in bus copper savings. I don't doubt that there may be some tertiary advantages to higher voltage too. I just haven't figured out what they might be.
 

Sponsored

electruck

Well-Known Member
Joined
Oct 6, 2019
Threads
69
Messages
3,495
Reaction score
6,454
Location
Dallas, TX
Vehicles
2023 Rivian R1S
My perspective on this keeps growing as discussions like the ones here take place but I find I keep coming back to the same place: the primaary advantage of higher voltage is the ability to charge faster with secondary benefit in bus copper savings. I don't doubt that there may be some tertiary advantages to higher voltage too. I just haven't figured out what they might be.
Not saying the following is the case, just throwing it out for consideration: perhaps you are expecting individual gains to be bigger than they are?

I think one of the things that has really differentiated Tesla and Lucid (and to a degree Rivian) is that they recognize that inefficiency is really a death by a thousand cuts phenomenon and have chosen to capitalize on efficiency via a thousand optimizations. Rawlinson also highlights synergistic gains such as the ability to improve aero as a result of being able to downsize the cooling system capacity. If one is only focused on optimizing individual components, the system as a whole may not be truly optimal.
 
Last edited:

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
Not saying the following is the case, just throwing it out for consideration: perhaps you are expecting individual gains to be bigger than they are?
The sort of thing I am thinking of are subtle things like having to build a transistor to higher voltage tolerance results, as icing on the cake, lower parasitic capacitance so that it can switch faster thus reducing switching losses and making an inverter built with these components more efficient therefore. Note that this example is entirely hypothetical

I think one of the things that has really differentiated Tesla and Lucid (and to a degree Rivian) is that they recognize that inefficiency is really a death by a thousand cuts phenomenon and have chosen to capitalize on efficiency via a thousand optimizations.
One of the OEMs may have better systems engineers than another but I think you can be sure that each of them had a detailed spreadsheet listing every fraction of a watt of loss in the vehicle. And I think we can bet that they sit around conference rooms staring at that list wondering which of those losses to attack next. Obviously they go for the low hanging fruit first but they don't stop when they have mastered those.



Rawlins also highlights synergistic gains such as the ability to improve aero as a result of being able to downsize the cooling system capacity. If one is only focused on optimizing individual components, the system as a whole may not be truly optimal.
While power semiconductor engineers look for ways to make switches less lossy any systems engineer worth his salt would immediately recognize that having to pump less heat away from the inverter would potentially mean a smaller compressor in a heat pump thus realizing further savings.

At this point BEV technology is far enough advanced (systems efficiencies are in the high 90's - there isn't much room left to improve) that engineers are after the harder to reach fruit.
 

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
I think it stems from people wondering about what they might learn about the technology from the configurator. I think the answer is not much unless it comes standard with 250 kW fast charging but can be optionally enhanced to charge at 350.
 

ajdelange

Well-Known Member
First Name
A. J.
Joined
Aug 1, 2019
Threads
9
Messages
2,883
Reaction score
2,317
Location
Virginia/Quebec
Vehicles
Tesla XLR+2019, Lexus, Landcruiser, R1T
Occupation
EE Retired
I know some of you have had enough of the tech but I guess my years of writing and reading specs has left an indelible mark. It occurred to me that the entire HPC set of chargers can be very compactly and completely specified in 1 sentence. I couldn't resist posting it:

An HPC class charger shall be capable of supplying, as a minimum, any voltage in the range from 200 V to 920 VDC, any current in the range from 5 A to 500 A and in any combination such that the product of voltage and current is less than 150 kW for the HPC150 class, less than 250 kW for the HPC250 class and less than 350 kW for the HPC350 class.

After reading this one can quickly construct the envelope for the particular class or, even without doing so, calculate the voltage limitation at any current or the current limitation at any voltage.
Sponsored

 
Last edited:
 




Top