Will the Cybertruck only come in a Quad Motor Variant - Pros and Cons

[JBee]

There has been various discussions through this forum about which motor variations will be released first and in particular why people think so.

I have been trying to get the message across to people that unlike a ICE vehicle, adding a extra motor to the CT isn't actually quite such a big deal as many make it out to be. Especially if you consider that Monroes teardown puts the Tesla motor cost at around $750 each.

Because of this I have been thinking that it may well be possible that the CT will only come in a quad motor variant at all, making every CT produced a quad motor.

Let me list the reasons why:

1. The motor used in the MS Plaid uses a CF wrap that allows the rotor to spin at around 23,000RPM and generate 250kW in it's current configuration - If one were to slightly reduce the size of the CF motors to say 150kW but use four of them instead, one for each wheel, then you would get around around 600kW for the non-performance version and at 200kW each 800kW total for a Plaid performance version.
2. Because of the high RPM, CT can use a larger gear ratio for a better suited top speed around 120mph (200kmh), preferably using a planetary gearset, that also has better torque handling capabilities (because it has 3x the tooth area) and is smaller and more compact, and can be mounted inline with the rotor driveshaft for superior packaging. The higher ratio obviously helps put more torque on the wheels for pulling trailers or off-road or use with higher payloads as would be the case for the CT.
3. Quad motor obviously also comes with the benefits of torque vectoring on all wheels that can be used on and off-road for superior traction and handling.
4. The motor speed controllers would be integrated onto each motor individually, and cooled as was done previously
5. There would be no clumsy sized differential or gearing offering better packaging
6. The front and rear motor assemblies could be the same, meaning only one sub-assembly for ease of manufacturing, logistics, parts and assembly
7. Eventually, a rear wheel only version could still be offered by simply not installing the front motor assembly, like they do with MY and M3, so it would still have a two motor in the rear only, with torque vectoring, and essentially a locking differential feature to help it along on loose terrain, and even while towing.

One of the main benefits of the smaller motor configuration is the ability to improve the packaging of the drive units so they are both smaller and lighter than the single motor drive units and so that they take up less under bed space on the CT, and increase it's bed payload capacity. To visualize this lets have a look at how Lucid Air packages it's rear 250kW each dual motor setup:

And how small it is when placed into the vehicle:

As is evident from the images having multiple smaller higher RPM, high power to volume, electric motors might well be the new drivetrain layout for the CT.

What does everyone think?

Sponsored

 

[Arctic_White]

You hit the nail in the head: adding an extra motor is less than $1K cost but will command a very high premium.

 

[rr6013]

190lbs. Koengisegg paired inverter axial flux motors

Overkill on the front of a CT, I expect weight savings will see their adoption. Cybertruck doesn’t need that much power driving the steering axels.

Tesla’s Gen1 dual motor rear drive unit remains my truck preferred embodiment for Cybertruck. The Lucid example is adequate except little gears run on little bearings and small shafts. Not a truck combination until proven.

 

[Bill906]

I agree with your assumptions that they might go with Quad motor only, or quad motor and dual motor options.

I question the part about using high speed motors.
Power is the product of torque and speed. Electric motors have many variables that can be changed to get the speed and torque required as well as what power source is available. Often electric motors have a point on their speed/torque curve where torque drops as speed increases. After that point, the motor is considered to be in "constant power" mode. Now, all motors are different, but if the CF wrapped motors follow typical AC motor curves, it would seem non-beneficial to increase speed at the cost of torque only to use a gearbox to increase torque by reducing speed. I admit I'm assuming the top speed of the CF wrapped motors is greater than the base speed and the model S plaid is in the constant power region at top speed. I feel that is a fairly safe assumption, granted I don't know how far into that region the top speed (≈23,000RPM) is.

Above screenshot stolen from:
https://www.researchgate.net/figure...orque-power-vs-speed-of-IPMSMs_fig1_263272422

 

[JBee]

190lbs. Koengisegg paired inverter axial flux motors

Overkill on the front of a CT, I expect weight savings will see their adoption. Cybertruck doesn’t need that much power driving the steering axels.

Tesla’s Gen1 dual motor rear drive unit remains my truck preferred embodiment for Cybertruck. The Lucid example is adequate except little gears run on little bearings and small shafts. Not a truck combination until proven.

There's one thing I left out of the original post and that is if you are using this with a high payload or with a trailer, you really want two motors in the front for effective regen braking, whilst being able to torque vector for safety. In a one motor setup you will need to use the brakes instead as to avoid one wheel locking up, because there is no way to modulate a single wheel using one motor and a open front differential.

As for the two motors being overkill, don't forget I'm suggesting they will make a smaller less powerful CF motor especially for the CT configuration, which I'm expecting to carry over through the rest of the model range.

BTW I like what Koenigegg done with the two axial flux motors. Not sure why Tesla hasn't gone down that route yet.

 

[JBee]

I agree with your assumptions that they might go with Quad motor only, or quad motor and dual motor options.

I question the part about using high speed motors.
Power is the product of torque and speed. Electric motors have many variables that can be changed to get the speed and torque required as well as what power source is available. Often electric motors have a point on their speed/torque curve where torque drops as speed increases. After that point, the motor is considered to be in "constant power" mode. Now, all motors are different, but if the CF wrapped motors follow typical AC motor curves, it would seem non-beneficial to increase speed at the cost of torque only to use a gearbox to increase torque by reducing speed. I admit I'm assuming the top speed of the CF wrapped motors is greater than the base speed and the model S plaid is in the constant power region at top speed. I feel that is a fairly safe assumption, granted I don't know how far into that region the top speed (≈23,000RPM) is.

Above screenshot stolen from:
https://www.researchgate.net/figure...orque-power-vs-speed-of-IPMSMs_fig1_263272422

I'm assuming you have seen how Plaid compares to previous models:

I googled around trying to find a torque, power speed curve to compare yours to the plaid but was unable to find one.

But given the graph above power delivery seems pretty flat, sure torque might drop off, that will slow the rate of acceleration at the top end, but it's unlikely to be configured with gears to run out of torque.

The other thing to consider is that the MS Plaid is already fairly heavy, and the CT is likely to be around 25% heavier, so having a higher gearing ratio that tops out at 120/140mph, will only need to be slightly more. In the end wheel torque will be traction limited by the tyres in any case, and top end acceleration can be neglected if needed.

 

[NJturtlePower]

My ideal survey pick would have been, "A quad motor, tri-motor and dual motor" No single motor variant expected IMO.

 

[Bill906]

I'm assuming you have seen how Plaid compares to previous models:

I googled around trying to find a torque, power speed curve to compare yours to the plaid but was unable to find one.

But given the graph above power delivery seems pretty flat, sure torque might drop off, that will slow the rate of acceleration at the top end, but it's unlikely to be configured with gears to run out of torque.

The other thing to consider is that the MS Plaid is already fairly heavy, and the CT is likely to be around 25% heavier, so having a higher gearing ratio that tops out at 120/140mph, will only need to be slightly more. In the end wheel torque will be traction limited by the tyres in any case, and top end acceleration can be neglected if needed.

Your graph shows Speed vs. Power.
Power is the product of speed and torque.
If power remains constant and speed increases, torque decreases.
From that graph, it looks like torque peaks around 60MPH. After that, you're in constant power region and torque drops as speed increases.
Again, no reason to increase speed if you are losing torque only to trade the excess speed back into torque with a gearbox.
I'm not saying I know how the motor works, or how they are going to do it. All I'm saying is I think the simple "use the high speed motors and put a gearbox on it" may be to simple of an answer to a complex function.

 

[rr6013]

There's one thing I left out of the original post and that is if you are using this with a high payload or with a trailer, you really want two motors in the front for effective regen braking, whilst being able to torque vector for safety. In a one motor setup you will need to use the brakes instead as to avoid one wheel locking up, because there is no way to modulate a single wheel using one motor and a open front differential.

As for the two motors being overkill, don't forget I'm suggesting they will make a smaller less powerful CF motor especially for the CT configuration, which I'm expecting to carry over through the rest of the model range.

BTW I like what Koenigegg done with the two axial flux motors. Not sure why Tesla hasn't gone down that route yet.

Short answer Tesla already has nice, decent motors. Elon simply doesn’t see anything axial motors can contribute to Tesla, yet…
Tl;dr
The long answer Cybertruck doesn’t need high RPM + horsepower that CF enables, anyway. So especially in the front axel, no application there for CF, smaller or axial. There’s such a thing as being too light in the nose.
Yes, CF in the rear drive unit for Plaid ‘cuz high torque = extended HP is insanely better. But I’d be wary if I was EM of CT high lift suspension and high speed == bad combo. Tesla can regulate speed in software but suspension dynamics at speed kills.

When Tesla launches the Cybercar variant, that is the day I expect smaller CF wrapped rear drive unit motors and axial flux on the front AWD setup would be applicable for better handling there. But Roadster is the natural fit front and rear for crazy performance gains as a Koenigegg hypercar.

Reducing motor weight becomes more influential smaller the vehicle. Axial definitely starts eating into induction and PM motors marketshare as production goes down market. Tesla expects to meet axial flux as its next competitor. Not sure, wonder why…Tesla hasn't gone down that route.

 

[TheLastStarfighter]

No. I enjoy your thoughts and analysis, but Tesla had plans a couple years ago for 1, 2, 3 motor variants using a similar setup to their other cars. While we know they will add a fourth motor up front, I expect it will be similar to what they already did to add the the second in the rear in the plaid. I don't think they've had the time - or necessity - to really re-work things. They'll keep it as inline with other models as possible.

For proof of why this will be the case, we know two things: 1. Elon has said he hopes for 250k units per year. 2. We know a German company has built the equipment that builds the motors, and they have built equipment to build 750k motors per year. That means Tesla plans to use on average 3 motors per Cybertruck. Possibly less if they produce more than 250k vehicles per year. If it were 4 motors for all cars, the goal production would be down considerably.

 

[JBee]

Your graph shows Speed vs. Power.
Power is the product of speed and torque.
If power remains constant and speed increases, torque decreases.
From that graph, it looks like torque peaks around 60MPH. After that, you're in constant power region and torque drops as speed increases.
Again, no reason to increase speed if you are losing torque only to trade the excess speed back into torque with a gearbox.
I'm not saying I know how the motor works, or how they are going to do it. All I'm saying is I think the simple "use the high speed motors and put a gearbox on it" may be to simple of an answer to a complex function.

As i said I tried to find a comparable graph for you to compare but haven't yet.

Regardless, I'm not sure what your actual concern is. A gearbox only magnifies torque but not power by trading rotor speed. A higher gearing than the Plaid would mean a ratio of around 12:1 instead of 9:1, this means the torque taper starts earlier, at a lower vehicle speed with the CT but it will not run out before it hits its lower top speed limit. The CT is also not going to be that much heavier in the grand scheme of things.

With the CF the higher power comes from two things, a smaller rotor gap producing greater flux and the ability to run at higher RPM without it expanding. Are you trying to say that Tesla need not have bothered with the higher RPM part?

That plaid graph above is highly idealised anyway. Reality is that Plaid is traction limited, not power or torque limited when accelerating and in essence the only negative of hitting the torque taper, it it does at all (we don't know atm) is that the vehicle won't accelerate with as much g-force as it does at lower speeds.

Operating at the traction limit means that Tesla is "cheating" with their motor output anyway, in that the only reason why they have such high outputs is because these motors are only capable of producing peak power and torque for a few seconds, before the motor would melt into a puddle. The motors are NOT rated for these power and torque levels for continuous use whatsoever, and can only achieve the performance by precooling the system, and trying to cool them as best as possible whilst driving.

By doing so the thermal mass of the various interconnected parts of the drivetrain cooling system, including the thermal mass of the battery, are used to absorb the peak heat being generated by the power train for the few seconds it takes to launch because of being at the traction limit. The cooling system then dissipates that heat through the cooling system in preparation for the next event. This was particularly evident in the Ludicrous Nürburgring laps, where after successive accelerations it would thermally throttle output power, and is the whole reason why the Plaid has 6x better cooling capacity, so it can do it more often to be competitive on the track.

Its actually the upgraded cooling system upgrade that gives the Plaid better lap times, not the motors themselves.

Simply put the more traction you have the shorter the acceleration run, the shorter amount of time heat is generated by the motor and the more "peaky" the power you can get out of it. You then end up heading towards the battery output as the weakest link again, and at some point the controllers.

This reminds me of another point why making motors smaller is viable pathway, and that is because they have a better cooling surface area to volume ratio. If you look at the Lucid setup you'll notice they've optimised the motor cooling jacket especially to do this and is why their packaging is so compact.

RE peak vs continuous motor power: At a continuous rated power they will only do a fraction of that. Consider that cruising at highway speeds with a M3 uses 130wh/km or 13kWh/100km and the motor is using just 13kW of power on average to do so. For normal driving use it will probably not even see 1-200kW as it becomes uncomfortable to drive that hard.

As is evident there are lots of parameters that overlap and cascade onto eachother that affect performance. There's another one I came accross recently; the reason the Plaid does the cheetah stance and lowers for launching is so it reduces the angle on the CV joints and doesn't bind up or break the halfshafts on launch. Apparently shuddering is a common problem that needs repairs. I'm pretty sure the controllers factor this in and limit output not to snap things as well.

On traction: at some point I expect we will see active aero to create downforce using a fan so that vehicle have more traction, to once again push the peak motor power acceleration envelope.



I hope this highlights that there are lots of moving parts that need to be optimised against eachother to leverage eachother for best performance. And making things smaller, lighter and more optimised by understanding the driving envelope It operates in.

 

[JBee]

No. I enjoy your thoughts and analysis, but Tesla had plans a couple years ago for 1, 2, 3 motor variants using a similar setup to their other cars. While we know they will add a fourth motor up front, I expect it will be similar to what they already did to add the the second in the rear in the plaid. I don't think they've had the time - or necessity - to really re-work things. They'll keep it as inline with other models as possible.

For proof of why this will be the case, we know two things: 1. Elon has said he hopes for 250k units per year. 2. We know a German company has built the equipment that builds the motors, and they have built equipment to build 750k motors per year. That means Tesla plans to use on average 3 motors per Cybertruck. Possibly less if they produce more than 250k vehicles per year. If it were 4 motors for all cars, the goal production would be down considerably.

I think you're operating on the assumption that all Tesla motors are the same. They are not, some are magnetless induction some are permanent magnet etc. Therexs around 5 or 6 of them already. They are much more different to eachother, even the speed controllers, than changing the diameter or rotor length of a CF motor will ever be.

Remember that this is no ICE, there a an order of magnitude less parts and are much simpler to gradually optimise on.

As for the parts numbers being evidence I think even EM gets updated projections on a weekly/monthly basis, let alone we don't know where they are destined for, MY/M3 is also a year behind orders etc.

 

[TheLastStarfighter]

I think you're operating on the assumption that all Tesla motors are the same. They are not, some are magnetless induction some are permanent magnet etc. Therexs around 5 or 6 of them already. They are much more different to eachother, even the speed controllers, than changing the diameter or rotor length of a CF motor will ever be.

Remember that this is no ICE, there a an order of magnitude less parts and are much simpler to gradually optimise on.

As for the parts numbers being evidence I think even EM gets updated projections on a weekly/monthly basis, let alone we don't know where they are destined for, MY/M3 is also a year behind orders etc.

We do know that the equipment is for the Cybertruck. The machines are scheduled to start producing middle next year, up to 750k units.

I'm not under the assumption that all of the motors are the same. They are, however, fairly interchangeable.

 

[charliemagpie]

125,000 dual motors + 125,000 Quad motors

= 250,000 trucks , average 3 motors.

Whether they are the same motors , i will leave that in the lap of the experts. Maybe those 750,000 'drives' out of Germania aren't all the same. Yeahhh Baby !!

 

[JBee]

We do know that the equipment is for the Cybertruck. The machines are scheduled to start producing middle next year, up to 750k units.

I'm not under the assumption that all of the motors are the same. They are, however, fairly interchangeable.

And they have 1.5m CT orders? I don't see a correlation between 750k motors a year and 250k CTs. You could make that connection and number up with any combination of motors, including 2 and 4 motor variants as proposed, especially considering we're still waiting for ramp to full production for a few years.

And the motors being designed to be "interchangeable" is an argument for the change to QM from my perspective.

Sponsored