AC Induction Motor FOC control

[maju]

Hi, I'm completely new there. It's my first post, so feel free to tell me if I did anything wrong. I wonder if this is the proper category, but I had to start somewhere and it seems like a good place to go.

I'm trying to build my own electric car, with custom motor driver, battery pack etc. And yes, I know that this is VERY hard thing to do, but I've been collecting information about the process for about a year now. Since the title of the topic talks about the motor control, this will be my main focus, but I will provide some more information, maybe they're useful.

So... I want to use an AC induction motor in my vehicle. About 20 kW. It's rated for 55V AC in delta configuration and has 2400 rpm rating. It comes from an old traction lift. I consider other motors, but I will talk about it later. I wasn't able to find any reliable information about whether 55V delta voltage stands for amplitude of line voltage or RMS line voltage. Or maybe something else?

My plan was to use IRFP4468 for my power stage. Is this a good selection, or should I look for something else? I actually own a few bags of these, so I'd prefer to stay with them. I initially wanted to connect 2 or 3 of them in parallel for each segment of 3-phase H-bridge (so 12 or 18 total) and use IR2110 connected in parallel to drive them, but I've read somewhere that although it will work, it's not the best solution possible. So do you guys have any suggestions what I should consider?

Okay, here starts the hard part: the control board. I designed my very crude SPWM inverter using constant V/Hz, but (surprise...) it didn't work out well. I was using STM32F103 as microcontroller, and using one of it's timers to generate 3 PWM signals. It can post the code if you want, but I don't think it's necessary, since I abbadoned this idea. I then found the STSW-STM8020 library. It features the V/Hz control of ACIM drives. But then I found the paper describing EV car conversion with Field Oriented Control of AC induction motor (ACIM). I will leave link to it at the end. So after some Googling I found the STSW-STM32100 library, which supports FOC. Unfortunately, it supports only BLDC/PMSM FOC, not ACIM FOC. I found that the previous version of it had supported the ACIM FOC. However, even the succesor of STSW-STM32100, which is MC SDK for CUBE-MX doesn't support ACIM FOC! Could anyone explain me what's going on? Did I miss something, or is ACIM FOC some stupid idea that is totally unsuitable for application?

Talking about FOC: the real reason why I'm writing this post is to get information on how to implement it for ACIM. I know the basic principle: you convert the 3-dimensional coordinate system to the 2-dimensional, convert current vectors to the rotation vectors, then move the rotor vector, calculate the new current vector values and convert it back to the 3-dimensional coordinate system and then feed PWM with the calculated voltage values. However, it is just too complicated for me. I found some implementations on PIC microcontrollers, but I have never programmed PIC's before. I know Atmel and STM a bit (I'm not an expert, but I've already done many projects on them), but not Microchip. So if anyone could provide me with working example of ACIM FOC implementation, or at least with the required equations and explaination on how to implement them im STM32/ATmega, I would be really, really grateful.

Maybe you're asking yourself 'why won't he just use the BLDC/PMSM motor?'. The answer is simple: the price. The AC motor I'm talking about costs around 200 US dollars, my DC alternative (6 kW, 48V) also around that, while BLDC's are much, much more... I could of course use the commercial solution, but point of this whole thing is to make something by my own. I'm aware of this project difficulty, but I will be really happy if you made an effort into explaining me how to make it working and fulfil my ambition. If it will be too hard - it's okay, I have the working plan B with DC motor.

Links:
STSW-STM8020 https://www.st.com/en/embedded-software/stsw-stm8020.html
STSW-STM32100 https://www.st.com/en/embedded-software/stsw-stm32100.html
Paper describing car conversion (that's how I would like my project work like, except that it doesn't have to be sensorless) https://www.ajol.info/index.php/ijest/article/viewFile/64009/51809
IRFP4468 datasheet https://www.infineon.com/dgdl/irfp4468pbf.pdf?fileId=5546d462533600a40153562c73472019

Thanks in advance!

[Benta]

If you really have a 6 kW DC motor on hand, use that instead. Its characteristics are much better for an electric vehicle than an induction motor, FOC or not.

[schmitt trigger]

I wasn't able to find any reliable information about whether 55V delta voltage stands for amplitude of line voltage or RMS line voltage. Or maybe something else?

 

Thanks in advance!

Induction motors are always rated for RMS, line-to-line voltage.

55 volts.......that is an oddball voltage. Where did it come from?
Also, what is its rated frequency and RPM? Does it have a plate you could take a photo?

[jmelson]

Hmmm, I have been interested in this for a LONG time.  I did build an electric car some years ago.  I got a friend's donor VW beetle (that's the ORIGINAL beetle) and swapped in a jet engine starter/generator (1300 A for 1 minute for starting, 400 A continuous at 28 V.  I built a PWM regulator for the field, and ran it off 4 trolling motor batteries.  With the field full on, it idled at 3000 RPM and made a huge racket.  I wanted to make it into a hybrid, and adapted (badly) a stratofortress generator to a Honda 350 motorcycle engine.  But, the coupling was not built well enough, and would have had to be re-done.  Anyway, the body was MASSIVELY rusted out, so there was no way it could have passed inspection here.  So, the project just died.  But, I DID drive it around some, and it worked amazingly well.  It seemed to be "zippier" than the stock VW engine.

So, it seems that most AC induction motors are MASSIVELY heavy, like several hundred Lbs. for a 20 - 30 Hp unit.  If your motor was intended for vehicle use, then maybe it is built lighter.  I have been thinking a machine tool permanent magnet spindle motor might be a better place to start.

Jon

[Zero999]

Mains frequency induction motors are bulky because they run at such a low frequency. It might be possible to overdrive it at a higher frequency, speed and voltage, assuming the insulation and bearings can take it and get a lot more power in the same size.

[schmitt trigger]

The heft of an induction motor, for a given power rating, increases with motors rated for continuous duty vs intermittent duty.

Another factor is whether they are open or closed frame. The latter which does not have the benefit of direct air cooling the windings, but has to cool them down via thermal conduction thru the frame, also adds heft.

[maju]

If you really have a 6 kW DC motor on hand, use that instead. Its characteristics are much better for an electric vehicle than an induction motor, FOC or not.

Could you please explain me why? Tesla, Nissan and other "big guys" run on AC or BLDC. If DC was better, why would they do this?

Induction motors are always rated for RMS, line-to-line voltage.

55 volts.......that is an oddball voltage. Where did it come from?
Also, what is its rated frequency and RPM? Does it have a plate you could take a photo?

Thanks for the information. Photo in the description. It comes from an old forklift.

Motor weight isn't a problem in this case (DC motor wieghts the same), I just wonder whether benefit of implementing FOC to this motor is worth it or if I should just go with DC (I don't consider static V/Hz control since I know that it won't work as intended).

I think I made it clear enough in my first post, but my main question is if anyone out there has an access to the working implementation of FOC on STM32 or Atmel. That would be really, really helpful.

[maju]

Hmmm, I have been interested in this for a LONG time.  I did build an electric car some years ago.  I got a friend's donor VW beetle (that's the ORIGINAL beetle) and swapped in a jet engine starter/generator (1300 A for 1 minute for starting, 400 A continuous at 28 V.  I built a PWM regulator for the field, and ran it off 4 trolling motor batteries.  With the field full on, it idled at 3000 RPM and made a huge racket.  I wanted to make it into a hybrid, and adapted (badly) a stratofortress generator to a Honda 350 motorcycle engine.  But, the coupling was not built well enough, and would have had to be re-done.  Anyway, the body was MASSIVELY rusted out, so there was no way it could have passed inspection here.  So, the project just died.  But, I DID drive it around some, and it worked amazingly well.  It seemed to be "zippier" than the stock VW engine.

So, it seems that most AC induction motors are MASSIVELY heavy, like several hundred Lbs. for a 20 - 30 Hp unit.  If your motor was intended for vehicle use, then maybe it is built lighter.  I have been thinking a machine tool permanent magnet spindle motor might be a better place to start.

Jon

I don't know if I got you right: did you use engine starter as a drive motor for the vehicle? I considered it at first, but everyone said that it's impossible because it will die quickly. It's a pity that it eventually died, but I'm really impressed by your creativity.

[Dave]

Talking about FOC: the real reason why I'm writing this post is to get information on how to implement it for ACIM. I know the basic principle: you convert the 3-dimensional coordinate system to the 2-dimensional, convert current vectors to the rotation vectors, then move the rotor vector, calculate the new current vector values and convert it back to the 3-dimensional coordinate system and then feed PWM with the calculated voltage values. However, it is just too complicated for me. I found some implementations on PIC microcontrollers, but I have never programmed PIC's before. I know Atmel and STM a bit (I'm not an expert, but I've already done many projects on them), but not Microchip. So if anyone could provide me with working example of ACIM FOC implementation, or at least with the required equations and explaination on how to implement them im STM32/ATmega, I would be really, really grateful.

You're going to need a proper DSP microcontroller with the appropriate PWM peripherals and ADCs to drive the motor responsively. The vector transformations and rotations are actually the easy part, implementing a good FOC has a whole plethora of other intricacies that need to be kept in check and calculated with the frequency of the control loop.

Not to discourage you, but doing something of this scale usually requires a whole team of experts. I honestly don't see a one man band being capable of pulling this off.