Interesting that Munro said that the one thing he was not able to reverse engineer was the Halbach array in the motor. He implied that it would require observing it being assembled and he did not know who made it.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Interesting that Munro said that the one thing he was not able to reverse engineer was the Halbach array in the motor. He implied that it would require observing it being assembled and he did not know who made it.
This is probably ignorance on Sandy's part. The halfbach magnets have been used in accelerators for a long time. I think the major part of the improvements is in the power electronics that control the winding current to achieve efficiency, reduce noise, increase driving performance etc. I could be wrong, though, but I couldn't imagine the halbach array is such a big secret. Well, maybe there is some secret material needed, but the major breakthrough is in the electric engineering, the controller electronics.
Great Great stuff! Had breakfast with some engineer friends this morning and we had an interesting discussion. Bunch of retired guys mostly that meet for early breakfast on Sundays. Permissivity was not a familiar term so had to do some research on that which seems to be also known as permittivity in some circles I think.
Ok, here I think the functional event is the ability of the high permissivity Alnico magnet to have it's field swap polarity at the brief pulse of the electromagnetic coil whereby the flux lines become external in an enduring fashion. Very interesting. I ran into a different but similar thing with a safety related solenoid valve where it could retain some residual magnetism and "latch" - a very bad thing. Needed to be modified with a capacitor (Not a flux capacitor!) that would send out a brief reverse charge to the solenoid when the solenoid needed to transition - worked great.
So, too soon?
Great stuff. So interesting to learn about these motors and how these permanent magnets can be manipulated. Munro talks about the Halbach Array delivering "root of 2" improvement - 1.41. And I have to be excited about the reduction in needed cooling in the SRMs. They really are going after a million mile drive unit.
Have to say that this tech seems applicable to a linear drive unit for the hyperloop - efficient, reliable, cool.
I am trying to educate myself on the components used in the inverters that are state of the art electronic components. Got stuff to learn.
Like your super bottle
Couldn’t resist. Love it when tech comes with coolness and humor. In one little glyph they have captured attention to their technical accomplishment. Great engineering very much deserves it.
I was so happy when Munro dragged out that superbottle from his bag. If I was a short, seeing that elegant solution so enrapture Munro, I would be terrified I’d made a huge mistake.
I bet there is more to come as Model 3 gets a deeper and deeper examination.
Artful Dodger
"Neko no me"
"Sokath, his eyes open!"
THAT is the great, great stuff. The Tesla Model 3 is a wealth, no a treasure chest, of new ideas, approaches, and applications. (I'm slightly mad at my physics teacher for failing to educate me properly in my youth).
CH3ERS!
EDIT:
Oh you're right of course. I should make it a more frequent practice to link my keywords, like Permittivity:
"More specifically, permittivity describes the amount of charge needed to generate one unit of electric flux in a particular medium. Accordingly, a charge will yield more electric flux in a medium with low permittivity than in a medium with high permittivity. Permittivity is the measure of a material's ability to store an electric field in the polarization of the medium."
Last edited:
Artful Dodger
"Neko no me"
Me too. I suspect (without proof because I can't test my theory) that the Model 3 motor uses a switched reluctance magnet like we've discussed in the video above, but with the added feature of having one-sided external magnetic flux when switched to the 'on' state.
We'd need to see the endcaps for the magnets (munro didn't show any other components in his recent Autoline interview), and we'd need to test the magnets with a pulse electricity wrapped in a coil around them.
Also, I'd like to test the horizontal bars in the rotor shown in pictures in this article. It's not clear if those bars are magnets (as the article author claimed without being present at the teardown he discusses). These might just be steel bars joining multiple steel plates into a rotor of desired dimensions.
Otherwise, its not clear (to me at least) how a permanent magnet in that position would enhance the rotor's torque. I'd need to do tests to identify the materials used per component, and see a computer simulation/analysis of the 3-D magnetic flux leakage to really understand what's going on with this motor.
Wowsers, this '3' is some kind of 21st century time capsule of future technology. A real Gobstopper.
CHE3RS!
If it had end caps, they would not be in the main flux direction. Based on the end view from the recent interview, the resultant field is either toward or away from the circumference.
Given the 54 stator slots, I think the magnets are there to couple with the additional 2 stator pole located between the 60 degree aligned ones. (6 rotor poles × 3 grooves per stator pole × 3 statir pole per rotor pole). It may be the magnets are set in pairs to loop the inner pair of poles to each other. N.s.n.S.n.s.N where the capitals line up to the main rotor path and the lower case are the middle field looping through the magnets.
They are the magnets Sandy showed in the recent interview. There are no other objects of that shape in the motor. The author was in contact with Munro who supplied the photo directly, so I'd expect the details are accurate.
SageBrush
REJECT Fascism
Please allow me to interrupt this great discussion (that is mostly over my head) to ask a simple question:
The head of Tesla motor development mentioned that PM-SRM was a trade-off. Clearly they gained efficiency, and presumably price. What was given up ?
The head of Tesla motor development mentioned that PM-SRM was a trade-off. Clearly they gained efficiency, and presumably price. What was given up ?
Magnets aren't free, so there is a cost factor there, including in manufacturing. The AC motor squirrel cage may be cheaper. However, they then add the coaxial cooling to the rotor, so system level cost is ???
It is possible to demagnetize a PM with a large enough field (depends on the magnet's coercivity) so maybe they sacrificed peak torque? (I say sacrificed, but it's more like "here is a design factor you can adjust other things to compensate for")
SageBrush
REJECT Fascism
Yeah ... I forgot that the motor in my LR car maxes out at ~ 192 kW and the car only manages 0 - 60 mph in a rather pedestrian 5.1 seconds (compared to the other Tesla speed demons, anyway.)
Anyone care to guess how much mass of rare earth metals in the PM-SRM compared to say, the GM Bolt ?
The Bolt looks like the rotor is a center shaft with magnets mounted to the outside, so mostly magnet. Tesla is mostly steel, a fifth or less depending how many magnets are along the length, maybe??
After thinking about it more I think the approach is this:
Normal SR motors have few rotor poles and a slightly larger number of stator poles (1.5x or 2x). This low number causes high torque ripple due the angular distance from the next pole. To reduce this ripple and increase torque, Tesla used a more AC type winding and a higher number of stator poles. So 18 poles at anyone one time with 3 phases = 54 electrical switch points per 360 degree commutation. The problem with this is that the additional poles would short circuit the flux paths through the rotor. Instead of the flux traveling though 60 degrees of rotor, it would pass through only 20.
To combat this and boost efficiency, they add a pair of magnets between the rotor's SR poles. These interact with the additional two intermediate poles like a normal PM motor. With a delta winding configuration with two connections driven the stator field pole groupings would be:
nNnsSsnNn sSsnNnsSs nNnsSsnNn ... and shift one letter each electrical commutation.
The large Upper case coils have double the current/ flux of the lower case steady state(R, vs 2R yields 2I vs I)
The rotor main SR path connects the first N and S of each group, with the magnets linking the other pair.
The back side of the magnets would normally counter the effect of the front side, (n-s and s-n) so the the use of the Halbach array minimizes the back side field while coupling the flux between the pairs.
Net results: inner flux loops and outer flux loops, low ripple, high torque.
Normal SR motors have few rotor poles and a slightly larger number of stator poles (1.5x or 2x). This low number causes high torque ripple due the angular distance from the next pole. To reduce this ripple and increase torque, Tesla used a more AC type winding and a higher number of stator poles. So 18 poles at anyone one time with 3 phases = 54 electrical switch points per 360 degree commutation. The problem with this is that the additional poles would short circuit the flux paths through the rotor. Instead of the flux traveling though 60 degrees of rotor, it would pass through only 20.
To combat this and boost efficiency, they add a pair of magnets between the rotor's SR poles. These interact with the additional two intermediate poles like a normal PM motor. With a delta winding configuration with two connections driven the stator field pole groupings would be:
nNnsSsnNn sSsnNnsSs nNnsSsnNn ... and shift one letter each electrical commutation.
The large Upper case coils have double the current/ flux of the lower case steady state(R, vs 2R yields 2I vs I)
The rotor main SR path connects the first N and S of each group, with the magnets linking the other pair.
The back side of the magnets would normally counter the effect of the front side, (n-s and s-n) so the the use of the Halbach array minimizes the back side field while coupling the flux between the pairs.
Net results: inner flux loops and outer flux loops, low ripple, high torque.
Awesome discussion. My assumption is that the SRM does not play well with being turned off as would be the case with the induction motors on the S and X. So something gained and something lost. The SRM in the 3 I presume to be in use continually. The induction motor on the AWD front position plays nicely with being unused at times. So if this is correct then this would be a functional trade-off for the design adding complexity in some configurations.
Speculatively, it may be that the induction motors are more fully friendly to being modified by SW while the SRM is unable to have some modifications done over the air.
It will be interesting to see if there are still some induction motors in the Semi.
A pure SR motor is fine with being spun when off. The Tesla version with PM in the rotor will produce some drag due to induced currents in the stator. Likley one of the reasons the front motor is AC induction.
No issues with changing motor drive software in either case.
Young Kelvin over on Twitter quotes an unnamed supplier, that the magnet quad is not actually a Halbach array:
Kelvin Yang on Twitter
PS. Pre-market trade up @ 342.41$
Tweet:
My understanding: Separation makes sense for steel laminations to reduce eddy currents. However ceramic magnets would not benefit from separation. It would also not inherently increase flux.
Seems like Sandy would notice one side having more flux than the other. It may not be a normal Hallback configuration if they are routing flux between statir poles.
I know this is an old thread but I was wondering whether you'd be willing to update your "nearly useless" comments. Apparently Tesla has invented a way to mitigate the disadvantages of a SR motor and has installed them successfully in hundreds of thousands of Model 3s, including mine. I am always concerned when experts apodictically lock into a given belief and then seem unwilling to accept that technology is a moving target.
Last edited:
nwdiver
Well-Known Member
Brando
Active Member
side note: Synchronous motor - see wikipedia even Reluctance Motors are referenced
Synchronous motor - Wikipedia
Also, IF you haven't read this entire thread, it would be good to go back and review.
take care
good luck
make friends and help them when possible, they may well return the favor.
Similar threads
