Paul and Sabrina's Cheap 3 Phase Inverter (AC Controller) with Field Oriented Control

[shaggythegangsta]

Originally Posted by MPaulHolmes

WHAT SHAGGY SAID! HAHA.

By the way, I'm not going to be much help at the moment, since I'm super duper busy with my job right now. But I did get the motor up to about 6190RPM at 339v. The saliency correction helps some, but to really get the speed cranking, you have to advance the rotor flux angle offset. I have no idea why. But it's easy enough to do that.

Hi Paul, I want To ask if thIS Project IS Closed or it will be updated again. .?
Thanks

[e*clipse]

That's good, solid progress IMHO.

Actually, from 4898 @ 375V to 6190 @ 339V is excellent progress. In a practical sense, it's a pretty normal highway commuting speed where I am. (assuming the MGR with its 6.86:1 gear ratio and 24.9" diameter tires)

From now on, extra voltage, flux angle offsets, and maybe some flux capacitors will push me into the speeding ticket zone! Woo Woo!

- E*clipse

Originally Posted by MPaulHolmes

WHAT SHAGGY SAID! HAHA.

By the way, I'm not going to be much help at the moment, since I'm super duper busy with my job right now. But I did get the motor up to about 6190RPM at 339v. The saliency correction helps some, but to really get the speed cranking, you have to advance the rotor flux angle offset. I have no idea why. But it's easy enough to do that.

[e*clipse]

Hello Shaggy! Welcome to the project. As far as I know, it's far from closed - Paul & thingstodo have been making incredible progress.

Regarding your questions about a 6010 board, perhaps I'll take a stab at this. I can certainly appreciate your enthusiasm, and more input is always helpful.

I guess I would like to know your goals with regards to your project. Do you want a sensorless controller, like the an1162 is about? What type of motor do you want to control? These two questions radically affect the code. Code will include assembly language if timing is critial or the code needs to be as efficient as possible. Paul's code includes some assembly and I think it's WAAAAAAAYYYYYY better code than those app-notes use. Also, the code used for controlling an induction motor is much more complicated than the code for a synchronous motor like a BLDC motor.

The challenge with embedded computing is that the software and hardware are very inter-related. With any microcontroller, each pin has multiple uses, which affect the code. Paul's board design, for example, is highly developed for the job of running a 3 phase motor using field oriented code. That integration is very beneficial from both a cost and EMI perspective.

It looks to me like the board design you posted is a very multi-purpose demo board. It includes multiple ways of obtaining position feedback, configuration jumpers, and general purpose I/O connectors. To make a motor controller, you would still need IGBT drivers, current sensors, and hopefully some hardware fault-detection circuits. Believe it or not, that is the expensive stuff.

If you're learning about this stuff, I would HIGHLY recommend getting familiar with C and maybe assembly by doing some much simpler projects first. 3 phase motor controller requires critical timing of AtoD conversions among other things. If you're unfamiliar with PWM and taking analog measurements, I would start there - controlling a DC motor. This can be done with a DSPIC30F2010. Another thing you might consider is getting one of Paul's boards. He's run both induction and BLDC motors using FOC on this board, with both encoders and sensorless code.

I hope this helps,
E*clipse

Originally Posted by shaggythegangsta

first of all sorry for my bad english hahah, i'm not using google translator because it is even bad.. haha

hello guys i have already the pcb and schematic files for dspic30f6010a because
i'm working on AN1162 APP NOTE so i made mine pcb instead of buying mchv boards $$$$ but that application note has very difficult code to study, amix of asm and c code. so i'm thinking to switch to this project.

i want to know if someone is intrested to help me to make the code working on dspic30f6010a

first of all i'm a student, i'm not a programmer but i always desired to learn programming.

[shaggythegangsta]

Originally Posted by e*clipse

Hello Shaggy! Welcome to the project. As far as I know, it's far from closed - Paul & thingstodo have been making incredible progress.

Regarding your questions about a 6010 board, perhaps I'll take a stab at this. I can certainly appreciate your enthusiasm, and more input is always helpful.

I guess I would like to know your goals with regards to your project. Do you want a sensorless controller, like the an1162 is about? What type of motor do you want to control? These two questions radically affect the code. Code will include assembly language if timing is critial or the code needs to be as efficient as possible. Paul's code includes some assembly and I think it's WAAAAAAAYYYYYY better code than those app-notes use. Also, the code used for controlling an induction motor is much more complicated than the code for a synchronous motor like a BLDC motor.

The challenge with embedded computing is that the software and hardware are very inter-related. With any microcontroller, each pin has multiple uses, which affect the code. Paul's board design, for example, is highly developed for the job of running a 3 phase motor using field oriented code. That integration is very beneficial from both a cost and EMI perspective.

It looks to me like the board design you posted is a very multi-purpose demo board. It includes multiple ways of obtaining position feedback, configuration jumpers, and general purpose I/O connectors. To make a motor controller, you would still need IGBT drivers, current sensors, and hopefully some hardware fault-detection circuits. Believe it or not, that is the expensive stuff.

If you're learning about this stuff, I would HIGHLY recommend getting familiar with C and maybe assembly by doing some much simpler projects first. 3 phase motor controller requires critical timing of AtoD conversions among other things. If you're unfamiliar with PWM and taking analog measurements, I would start there - controlling a DC motor. This can be done with a DSPIC30F2010. Another thing you might consider is getting one of Paul's boards. He's run both induction and BLDC motors using FOC on this board, with both encoders and sensorless code.

I hope this helps,
E*clipse

hello e*clipse,

first of all thanks for answering,
my main goal is to make an inverter for induction motor in open loop and closed loop, for my final exam project,
because we don't study c language in my school, my goal was to find an working code, easy to understand and easy to explian, first i found 1162 appnote and until now i was studying that app note.

but then i found paul's inverter and code that is very easy to understand in confront of microchip's code.

my problem was that i have already 10 pcb boards that i have made. and i would like to adapt my board to this code.

my Board is working well with V/Hz mode, also current feedbacks are good.

my board has:

2 current sensors : LEM LTS25-NP (or 3 current sensors and 1 pot)
2 Pot inputs
1 pwm output, driven by a buffer ic
1 fault input
1 fault clearing output.
1 lcd output connector
1 CAN output
1 Encoder input (Filtred)
4 buttons
4 leds

[tajman]

First of all I apologize in advance for the long post but this thread (I read it all) is a very intriguing read. The unfolding drama of the controller being revised to perfection was enthralling! I feel I should be a few IQ points higher just from lurking but whether or not that's true I don't know.

To Paul: I greatly admire your dedication to this project even through some very challenging personal times I saw in your posts and your commitment to keeping the development open. I can type "Great Job!" but it feels like an understatement. I read the entire thread because I wanted to fully understand what was involved. I still don't understand many (many^(a gazilion)) things fully but I can see a lot of work went into this so far. Congratulations on these accomplishments!

thingstodo: Man, I thoroughly enjoyed your testing! The videos are really good, but I actually enjoyed your writing more! It was like you and Paul were the heroes in a novel and the inverter/motor the antagonist - their integration being the challenge to overcome. I was binge reading in the preliminary testing phases because stopping felt like I was at a cliff-hanger. Note: It ain't so bad getting old - It seems you've helped this project along immensely (forgetful or not)

It's great to see the camaraderie among the posters (or should I say team?) with ideas and contributions being thrashed out.

I signed up on this forum exclusively because of this thread. I came to it via the Instructables project Paul posted. I couldn't find the links to the project files and started searching (relentlessly) for it. You see, I've embarked upon a personal mission to do an EV conversion. However in counting the costs, an EV conversion with available commercial products would be the cost of new car in my patch of the globe. So I started looking at what can be done to reduce my costs. The controller (well electronics on the whole), motor and battery pack are the major costs. Obviously the search for alternative (read cheaper) controller brought me here.

Paul, I was able to get controller code from Github but wanted to know are you willing to make the PCB designs available? I would like to find out if I can get the boards done locally - I would like to source as many components as I can locally to avoid (what I anticipate to be) high import costs.

Also, is there still a way to contribute to the development? I saw there was a donation link very early in the thread but it doesn't seem functional anymore.

Tarquin.

[thingstodo]

Originally Posted by tajman

First of all I apologize in advance for the long post

Did I miss something? That post was quite short compared to some of mine

The controller code is a moving target. I think Paul does changes every day or every second day at least. I'm trying to convince him to release a version 1.10 and leave it as the stable, tested release ... then mess around with 1.10a, or 1.20, or whatever.

The latest AC board is components on one side only, I think. Some of the parts are pretty tiny and you will need tweezers to move them, and solder paste to keep the parts from wandering. But DIY can be done.

Paul just received the last of the parts for my DC controller, so I'll be over there tomorrow.

[tajman]

Thanks for the quick reply thingstodo.

Originally Posted by thingstodo

The controller code is a moving target. I think Paul does changes every day or every second day at least. I'm trying to convince him to release a version 1.10 and leave it as the stable, tested release ... then mess around with 1.10a, or 1.20, or whatever.

I gathered as much but I don't mind playing around with it for now. I'm in the "research phase" right now but I like how things are going here.

Originally Posted by thingstodo

The latest AC board is components on one side only, I think. Some of the parts are pretty tiny and you will need tweezers to move them, and solder paste to keep the parts from wandering. But DIY can be done.

Is this the design that includes the integrated (software selectable) resolver to encoder? I'm not afraid of solder or solder paste The PCB and SCH files I have are from 2014 - from the OpenRevolt wiki.

[thingstodo]

Originally Posted by tajman

Is this the design that includes the integrated (software selectable) resolver to encoder? I'm not afraid of solder or solder paste The PCB and SCH files I have are from 2014 - from the OpenRevolt wiki.

The resolver to encoder circuit is on the same board. I don't remember if it is software selectable - you've just read through everything so you probably know more than me at this point.

The resistor dimensions and pads are small. Past reading glasses for me. We're into the lighted magnifier range. A tiny dab of paste is put on each pad with a syringe, then the parts placed on top and the paste kinda sticks it there until you put it in the toaster oven to do the soldering.

The single-sided board was just completed ... Mar 1? The boards are here now.

[e*clipse]

Paul - is it ok to use your code for this?

If so, I could put my "teacher" hat on and help migrate the code to a 6010. I think that might also help efforts for this project.

shaggy - how much time before it's due? I'm really involved in another project right now, and this will have to be a side effort.

This would require shaggy to learn C. It's very do-able; even I did that while learning how to program PICs. It really helps if you know <any> programming; C is pretty easy to learn.

Thoughts?

- E*clipse

Originally Posted by shaggythegangsta

hello e*clipse,

first of all thanks for answering,
my main goal is to make an inverter for induction motor in open loop and closed loop, for my final exam project,
because we don't study c language in my school, my goal was to find an working code, easy to understand and easy to explian, first i found 1162 appnote and until now i was studying that app note.

but then i found paul's inverter and code that is very easy to understand in confront of microchip's code.

my problem was that i have already 10 pcb boards that i have made. and i would like to adapt my board to this code.

my Board is working well with V/Hz mode, also current feedbacks are good.

my board has:

2 current sensors : LEM LTS25-NP (or 3 current sensors and 1 pot)
2 Pot inputs
1 pwm output, driven by a buffer ic
1 fault input
1 fault clearing output.
1 lcd output connector
1 CAN output
1 Encoder input (Filtred)
4 buttons
4 leds

[shaggythegangsta]

Originally Posted by e*clipse

Paul - is it ok to use your code for this?

If so, I could put my "teacher" hat on and help migrate the code to a 6010. I think that might also help efforts for this project.

This would require shaggy to learn C. It's very do-able; even I did that while learning how to program PICs. It really helps if you know <any> programming; C is pretty easy to learn.

Thoughts?

- E*clipse

T

Thanks eclipse for all effort that you are putting... knowing that i'm not a programmer... there are only few people's that helps person like me..
I'm saying this because I wrote on microchip''s forum and I got no help...