Great work! Do you have FOC with separate D and Q currents? Didn't you try "fixed slip speed" scalar control algoryth? It's rather workable! Do you plan to make and publish photos of busses and IGBTs of new inverter on your blog? It's very interesting for power engineers like me. Thanks!
Thx for the question! Yes it is FOC with separate Id and Iq. I have never heard of "fixed" slip speed type control. We use a type of FOC called indirect FOC, whereas the slip speed is calculated by the motor currents and the motor parameters in real time. This is a "feedforward" method of control which does not require feedback - such as a in a sensorless algorithm. This is more accurate and better response, assuming we are doing our calculations right of course. It is called "indirect" because the rotor flux position is not measured such as in some sensorless algorithms for ACIM (or a PMAC with an encoder), it is calculated and predicted from the rotor speed (from the encoder) and the calculated slip speed (which is based on motor currents and current motor parameters).
Well I think Tesla uses the same algorithm. Main challenge is to select the right flux for every speed and torque to reach best effectivity. Constant slip speed is very simple but powerfull scalar algorithm. Inverter always outputs frequency propornional to RPM plus slip speed. But voltage is varied by PID for speed to reach target RPMs. Slip speed itself is selected to be so that motor gives maximum torque at fixed stator current, it's simple experiment to do with blocked rotor. At this frequency inductive and active resistances in rotor are equal and current and torque is maximized. Keeping slip speed the same at all RPMs and torques lets you easily control ACIM in rather good way. Not very far from MTPA, as I read. Several years ago I made CNC spindle inverter with such algorithm. Now days I develop PMSM invertors and I'm fery interested in insides of new tesla inverter visible on your videos.
Cool. Yeah, you can learn a lot about the motor from a locked rotor test. Actually, selecting the correct flux is not really important. Nor is selecting the torque for an EV. First the torque.. well that is selected by the driver with their foot.. so that's not something one needs to optimize for an EV :-) In terms of selecting optimal flux, this is of much lesser importance then ensuring your slip speed calculation (and thus your rotor flux position) is accurate. A quick rule of thumb for selecting flux is to set Id = Iq. That way rotor flux is equal to stator flux (assumimg equal inductance), and you will get good torque per amp. This will offer slower response for some motors But the Tesla motor has a very low rotor time constant so it is not an issue for an EV. By far the most important aspect is ensuring your rotor flux position is accurate. It is also the most difficult aspect.
Yeah, in my controllers for PMSMs I measure the rotor position very precisely with complicated analog encoder and algorythm, a little eroor in radians leads to many hundred amps different in currents. I hoped in ACIMs such a pricision is not needed, because you need to know exactly only the speed, not angle. But it seems not. One more question about tesla inverter, is it from front axle? And what about it's photos?..... May you send me smth via e-mail? I would very appreciate this! Also you can wath videos of my creation on my channel, that's electric motorcyles, ATV and even glider. Thanks!
Yeah, flux position in ACIM is no less important then PMSM. Slip just makes it more difficult. You can go to the website and find my email there and we can talk. edisonmotors.net
Sure, we will be selling drive units with our controller integrated. I am also considering offering a core exchange service where one sends their invert-er if they already have a drive unit and we send back a hacked inverter with a user manual. edisonmotors.net
Interesting choice using a dspic. Are you using Lebowski codebase?
Hello I have akso Tesla electric engine. Can you help me to how i connect the cable to start the engine same as you?
Which parameter is controlled to control the speed of motor
Great work!
Do you have FOC with separate D and Q currents?
Didn't you try "fixed slip speed" scalar control algoryth? It's rather workable!
Do you plan to make and publish photos of busses and IGBTs of new inverter on your blog?
It's very interesting for power engineers like me. Thanks!
Thx for the question!
Yes it is FOC with separate Id and Iq.
I have never heard of "fixed" slip speed type control. We use a type of FOC called indirect FOC, whereas the slip speed is calculated by the motor currents and the motor parameters in real time. This is a "feedforward" method of control which does not require feedback - such as a in a sensorless algorithm. This is more accurate and better response, assuming we are doing our calculations right of course.
It is called "indirect" because the rotor flux position is not measured such as in some sensorless algorithms for ACIM (or a PMAC with an encoder), it is calculated and predicted from the rotor speed (from the encoder) and the calculated slip speed (which is based on motor currents and current motor parameters).
Well I think Tesla uses the same algorithm. Main challenge is to select the right flux for every speed and torque to reach best effectivity.
Constant slip speed is very simple but powerfull scalar algorithm. Inverter always outputs frequency propornional to RPM plus slip speed. But voltage is varied by PID for speed to reach target RPMs.
Slip speed itself is selected to be so that motor gives maximum torque at fixed stator current, it's simple experiment to do with blocked rotor. At this frequency inductive and active resistances in rotor are equal and current and torque is maximized. Keeping slip speed the same at all RPMs and torques lets you easily control ACIM in rather good way. Not very far from MTPA, as I read.
Several years ago I made CNC spindle inverter with such algorithm.
Now days I develop PMSM invertors and I'm fery interested in insides of new tesla inverter visible on your videos.
Cool. Yeah, you can learn a lot about the motor from a locked rotor test.
Actually, selecting the correct flux is not really important. Nor is selecting the torque for an EV.
First the torque.. well that is selected by the driver with their foot.. so that's not something one needs to optimize for an EV :-)
In terms of selecting optimal flux, this is of much lesser importance then ensuring your slip speed calculation (and thus your rotor flux position) is accurate.
A quick rule of thumb for selecting flux is to set Id = Iq.
That way rotor flux is equal to stator flux (assumimg equal inductance), and you will get good torque per amp.
This will offer slower response for some motors
But the Tesla motor has a very low rotor time constant so it is not an issue for an EV.
By far the most important aspect is ensuring your rotor flux position is accurate. It is also the most difficult aspect.
Yeah, in my controllers for PMSMs I measure the rotor position very precisely with complicated analog encoder and algorythm, a little eroor in radians leads to many hundred amps different in currents. I hoped in ACIMs such a pricision is not needed, because you need to know exactly only the speed, not angle. But it seems not.
One more question about tesla inverter, is it from front axle? And what about it's photos?..... May you send me smth via e-mail? I would very appreciate this! Also you can wath videos of my creation on my channel, that's electric motorcyles, ATV and even glider. Thanks!
Yeah, flux position in ACIM is no less important then PMSM. Slip just makes it more difficult.
You can go to the website and find my email there and we can talk. edisonmotors.net
Do you have a email, website or contact info would like to learn more how you accomplished this!!
Sure, we will be selling drive units with our controller integrated. I am also considering offering a core exchange service where one sends their invert-er if they already have a drive unit and we send back a hacked inverter with a user manual.
edisonmotors.net
:56 it's Alive! 😱
I want my bird