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

Investor Engineering Discussions

This site may earn commission on affiliate links.
dww12 said:
So we beat the yoke to death, how about the carbon fiber wrapped rear 20k rpm motor? How long before one of the beast rear motors makes it to the 3/Y to increase top speed and efficiency? Seems doubtful all that motor engineering was just for the S/X. Is the next step for the 3P and YP in the bag?
I wish we had an separate investors engineering thread for this type of discussion, because it has the potential to devolve into long posts, if we do more than scratch the surface.

Some have ruled this out on the basis of cost or performance, they may be right or wrong, but it is hard to make a quick judgement.

On costs, Tesla makes the motors, and makes a lot of the machines which make the motors, the capital cost of machinery for a high volume program like Model 3/Y seems unlikely to be a problem.

The other issues are scale and the cost of raw materials. We don't know how the carbon sleeve is made, how that process scales, or what it costs.
Tesla finding better ways or making the carbon sleeve, or alternative materials would not surprise me.
Tesla tapping some of the advantages of Plaid without a carbon sleeve, would not surprise me...

I don't think the carbon sleeve has negative mass, so it isn't the only secret sauce in Plaid, and I expect Sandy Munro to be very happy when he pulls a Plaid drive apart..

If there are advantages in the Plaid architecture than means motors can be lighter and/or smaller, that might be applicable to a wide range of models,.
That comes back to a why motors are different sizes and weights.

There is more in the Plaid Model S that might eventually migrate down to the Model 3/Y, the 12V battery being one obvious thing. Again I expect Sandy to be very happy. I bet Sandy can hardly wait to start...

  • Like
 
I recommend also moving the discussion on FSD to this thread...

I remember Tesla did an take-over of a small organisation that specialised in optimising Neural Nets, to acquire the talented team..

With NNs and FSD, after it is working, considerable effort can be expended to optimise it. And it is likely some significant savings can be found.

While HW4 might be more powerful, Tesla will try hard to get FSD working with HW3. It is too early to make a call on that.
 
It's possible of course but the 3/Y motors are already so efficient and powerful I can't imagine much benefit. Unless the cost is minimal, or Tesla decides they want to significantly upgrade the specs of the Performance versions.
Yar, and to get the performance would take a dual motor rear drive unit which would bump up the cost more than minimally.
 
I suspect the existing 3/Y motors are capable of more performance as is and are just software limited. Tesla doesn't want them getting too close to the S.
Sometimes the aim may be reducing weight, rather than increasing power.

If that reduced weight means more range from a smaller battery pack, that brings compounding benefits.

I am sure a Plaid variant of a Model 3 would make a great track car, but Tesla would be wary of competing with Model S.
Plaid Model S is in a great slot at the moment, good value for money, if you want that kind of performance, and I'm also sure margins are good.
 
I suspect the existing 3/Y motors are capable of more performance as is and are just software limited. Tesla doesn't want them getting too close to the S.
That would align with the semi doing 0-60 in 20 seconds with 4 motors at 80k lbs...
5k lbs is 1/16 the mass, 2 motors is 1/2 the power, so 2.5 second 0-60 (ignoring pack, weight distribution, gearing, aero)
 
  • Helpful
Reactions: JRP3
That would align with the semi doing 0-60 in 20 seconds with 4 motors at 80k lbs...
5k lbs is 1/16 the mass, 2 motors is 1/2 the power, so 2.5 second 0-60 (ignoring pack, weight distribution, gearing, aero)
Power doesn't scale like that because 3P is traction limited most of the run. Semi is power limited, even with only 40% of the weight on the driven wheels.
 
I'm also interested in the power curve of the new drivetrain. I wonder if this is directly due to the carbon sleeving and/or tighter tolerances, or if there's some other special sauce perhaps in the inverters...

(Incidentally, I also find seemingly conflicting information on the back-EMF characteristics of switched-reluctance[1] motors)

[1] Of which I assume the Plaid motors are...
 
I'm also interested in the power curve of the new drivetrain. I wonder if this is directly due to the carbon sleeving and/or tighter tolerances, or if there's some other special sauce perhaps in the inverters...

(Incidentally, I also find seemingly conflicting information on the back-EMF characteristics of switched-reluctance[1] motors)

[1] Of which I assume the Plaid motors are...
My personal theory is that the volts/RPM value was dropped way down since the peak torque needed was reduced (50% ish due to one wheel driven, offset by the steeper gear ratio 9.734:1 to 7.5:1). This, plus the higher pack voltage pushed the point at which back-EMF limits power higher and lowered the slope (unless the minimal slope is due to other factors).
Reducing turns to drop back-EMF and torque/AMP would also reduce winding inductance which allows for faster field switching.

I've had trouble with good SR back EMF data also, but concider: to get mechanical power out, electrical power must be put in. Without back-EMF, there is no power in. So there is back-EMF (when excited), however, that still leaves open the possible drive waveforms to minimize BEMF for a certain power/torque operating point.
 
Agree Tesla is in prime position to be one of the winners if there is a shortage given how long they have been prioritising battery supply compared to their rivals - as they're likely to have locked in available supply before their competitors.

From what I've read and watched (and I'm no expert) there aren't that many shortcuts to spinning up new mining production - There's a ton of regulation and impact assessments and scaling issues that just take time to work through. That said, I'm sure Tesla's process would be faster than average timelines.

It will be interesting to see how it plays out.

I'm replying here so I don't derail the main thread...

What follows is speculation, but speculation 100% compatible with what Elon and Drew said on Battery Day, my main assumption is that they were telling the truth and they had something worth discussing.

Key Assumption - Tesla had a lab scale process working before Battery Day, IMO lab scale only requires taking some buckets filled with clay to a lab extracting the lithium and then returning the clay back to site...

When I thought about the process to extract lithium from clay my conclusions was Electrowinning is one possibility.
Essentially using electricity to plate the lithium metal...

Alkali metals can be extracted in this way, and that odiously includes lithium...

This might not be the process Tesla are using, but most of the rest of the post stays the same,

Mining process:-
  • Remove signicantly plants to a greenhouse and try to breed additional stock.,
  • Scape back topsoil.
  • Extract clays with earthmoving equipment and trucks.
  • Transport clays to processing plant (on-site)
  • Return clay to the same location (approximately)
  • Put back top soil and plants.
The mining process described above is "light-touch" and the lithium clays are not high value ecosystems, approvals still need to be granted.

Processing on site.:-
  • Dissolve clays in a mixture of salt and water (sea water can possibly be used, but needs to be transported some distance)
  • Extract lithium using some process powered by electricity with little or no additional chemicals.
  • Extract salt from remaining clay
  • Extract water from remaining clay
  • Return clay to site.
  • Water and salt captured for reuse, no on site emissions.
Processing at Austin:- (if using Electrowinning)
  • transport cathodes with plated lithium to Austin.
  • Add calthodes to water producing lithium hydroxide and hydrogen.
  • Additional purification steps, probably based on boiling points.
  • Return cathodes to mine site for reuse.
While this is completely hypothetical and speculative, it shows why I am keen on Electrowinning.
it is a clean "light-touch" process powered by electricity, that doesn't use additional chemicals, doesn't have exhaust stacks, and helps ensure a high purity end product.
Processing on site can be powered by on site solar,,,

This might not be the process Tesla is using, but I think they have a process, and they have a target timeline...

The lithium coming from Australia is hard rock, I expect Tesla to just process that at Austin.
That supply and other supplies have to support the ramp until the clay extraction process is working..
 
I got to thinking today, while pondering my enthusiasm about the CT all wheel steering (AWS: not the AMZN kind).

Humor me;
Elon recently Twatted about ‘steer by wire’…how awesome would it be to have steer by wire in an AWS CT.

I’ve heard complaints about Teslas inferior (u) turning radius. I love the steering of my wife’s 3P, but if steering was adjustable; could we retain that steering performance at med/high speeds without sacrificing low speed turning radius.

I’m sure that people smarter than I, have put thoughts into this. Curious what your conclusions are.
Turning radius is driven by max wheel angle, track width, and wheelbase. Steering ratio determines the number of turns full left to full right.

Electronically adjustable variable ratio steering can also decouple steering wheel movement from steering angle. It is not that hard to add while retaining a failsafe mechanical connection (extra motor at the wheel, or motor plus differential or planetary inline). Fixed ratio steering plus a fail connected clutch/ brake in the steering shaft would also allow for FSD and manual modes.

Rear wheel steering is frequently done by wire to allow for both crab (rear wheel same direction as front) and tighter turning (rear wheels opposite front).
 
  • Informative
  • Helpful
Reactions: FlyF4 and FireMedic
Continuing the 4680 pack discussion some selection from TheMightyFuji who seems to be very credible:

I don't think I can answer without being vague since I'm not sure if there is any public information regarding the existence of a non structural 4680 pack so I wouldnt be able to confirm or deny that. Bill Wri seems to be pushing limits with what he's saying and I'm not going to just spew insider information.

With that said, Bill Wri is still incorrect because this is neither 4680, nor is it structural.
To me this sounds as if he's confirming the existence of a non structural 4680 pack, i.e. since there is no public company information about a non structural 4680 pack he won't confirm its existence.

I was reading his tweet like he was being sincere. I am now reading it as sarcasm, which kind of makes him sound like an asshole, especially since this Alex guy is 100% correct and found evidence to proved the photo wrong in an intelligent way. I guess I had too much faith that people would admit when they're wrong. Either way, dude is wrong. 4680 pack is much taller than this plaid pack. I honestly don't understand how he could possibly mix them up if, as he claims, "he's seen them and the components inside".

The 4680s will come. I work with them everyday (I won't say for which application) but just know that no matter what is in your car, its the best you can buy.

Not that it matters for anything other than clarity but my signature is not pictured. I signed after this photo was taken. The pack was a lot more crowded by that point.