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

[P-hack]

it is an igbt half bridge on the prius, but yah, pretty simple. For high power charging (to reduce charge times) you do need the pfc stage though (or if your battery voltage is too high relative to the wall).

I'm not worried about software errors, that is a testing issue. I posted a link to hobbits page where he gets into some control details a few posts back. Personally I think avoiding regen unless the brake pedal is touched is simplest algorithmically and has its benefits in driving efficiency (it glides by default, encouraging gliding behavior). the brake light narrows it down to the battery buck or battery boost quadrant, and could be controlled by slip alone probably (just turn the top switch on) for regen (but not AC power).

<50 is about $22 apiece, >10000 is about $6. Add your own wire.

Edit: only talking about induction motors here, I don't know much about PM/reluctance regen, other than it is gonna be putting out some volts weather you want it to or not in regen, so the top buck switch is probably needed there, but single variable slip seems easier to implement otherwise.

[cts_casemod]

Looking forward for some progress then

[e*clipse]

That's why I really like those SOT-227 modules. They're much smaller than the "bricks" and they still use bolt-on busbars.

Regarding SiC diodes, my old spreadsheets (this was done about 10 years ago) indicate a forward voltage of 0.86V for a 1kW boost converter. I know parts have improved since then, for example I can't wait till this is available:
All-SiC and Si IGBT + SiC Rectifier CoPack from GeneSic Semiconductor

That Infineon IC that I used was truly amazing. It tracked the input current and adjusted the output so that the Power Factor was always near 1. The switching frequency I designed for was 100kHz, and I used a torroidal inductor.
In addition, Infineon also provided some good information for designing boost converters.
http://www.infineon.com/dgdl/Product...12b427caa43ccf
http://powerelectronics.com/site-fil...g/410pet21.pdf
ICE2xXXX ... - Datasheet Search Engine Download

- E*clipse

Originally Posted by cts_casemod

You're not the only one there. The brick modules make life much easier. No issues paralleling devices or crazy DC-LINK inductance calculations.

And for that, one has to stick with 30KHz or less, although for a charger the smaller modules may be a good option. Say a modular approach with 2KW per converter or so.

e*clipse what was the voltage drop of the SiC PFC diode on your boost converter?

Speaking of which, my PIC has two additional PWM channels for PFC correction (or whatever else you feel you want then for), in addition to the motor controller. Perhaps your device has an option to do the same, Paul

[cts_casemod]

Originally Posted by e*clipse

That's why I really like those SOT-227 modules. They're much smaller than the "bricks" and they still use bolt-on busbars.

Regarding SiC diodes, my old spreadsheets (this was done about 10 years ago) indicate a forward voltage of 0.86V for a 1kW boost converter. I know parts have improved since then, for example I can't wait till this is available:
All-SiC and Si IGBT + SiC Rectifier CoPack from GeneSic Semiconductor

That Infineon IC that I used was truly amazing. It tracked the input current and adjusted the output so that the Power Factor was always near 1. The switching frequency I designed for was 100kHz, and I used a torroidal inductor.
In addition, Infineon also provided some good information for designing boost converters.
http://www.infineon.com/dgdl/Product...12b427caa43ccf
http://powerelectronics.com/site-fil...g/410pet21.pdf
ICE2xXXX ... - Datasheet Search Engine Download

- E*clipse

Thats quite good.
High frequency PFC diodes I had have a drop of at least 1.1V unloaded.

I use this converter in my 1.5KW modules. Current is set via a shunt resistor.

MC34262-D

[cts_casemod]

Originally Posted by P-hack

Edit: only talking about induction motors here, I don't know much about PM/reluctance regen, other than it is gonna be putting out some volts weather you want it to or not in regen, so the top buck switch is probably needed there, but single variable slip seems easier to implement otherwise.

We're on the same boat then.

The voltage rise on a PM is more contained than that on an IM as the rotor flux is constant and the generated emf is mainly dependent on the rotor speed.

With an IM flux is dependent on the slip as well. If the generated waveform commands a negative slip of enough amplitude the voltage shoots to an undefined value. And as the voltage boosts, more current flows into the motor. One needs a very fast voltage sampling similar to that used on dynamic bus ripple compensation, sampled at every PWM cycle.

That would ideally be achieved by the same micro as the motor control so that both are in sync.

[P-hack]

Originally Posted by cts_casemod

That would ideally be achieved by the same micro as the motor control so that both are in sync.

Yup, I'm experimenting with a 72mhz 100 pin STM32F103VCT6 (huebner code compatible) dev board. Hoping to integrate everything BMS/Charger/controller/instrumentation feed (maybe buck/boost), someday, for no good reason So at least my experiments should be more port-able, someday. I doubt it will be timely though as there are far less challenging ways for me to get rolling.

[cts_casemod]

Originally Posted by P-hack

Yup, I'm experimenting with a 72mhz 100 pin STM32F103VCT6 (huebner code compatible) dev board. Hoping to integrate everything BMS/Charger/controller/instrumentation feed (maybe buck/boost), someday, for no good reason So at least my experiments should be more port-able, someday. I doubt it will be timely though as there are far less challenging ways for me to get rolling.

Huebner setup is cool. I never got to past the toolchain configuration, tough. Are you using his code?

I proposed him to make a compatible setup with some stuff I was working in the past to incorporate charger and EVSE communication. Since he wasn't very receptive I decided to roll my own.

Have you had a look at the piccolo series from Texas?

[MPaulHolmes]

What is a toolchain configuration?

[cts_casemod]

Originally Posted by MPaulHolmes

What is a toolchain configuration?

The toolchain is like MPLAB for PIC (The IDE).
ST doesn't provide a supported compiler, so one needs to use a 3rd party compiler, debugger, linker, peripheral libraries, (hence toolchain) etc. and set them up for that particular CPU.

Unless one has a degree on programing or some good links with someone who has its kinda tricky to get started.

[P-hack]

@paul, it is all the tools you need for edit/compile/run(flash) for a given microcontroller, technically. Though most just think of it as a cross platform compiler.

@cs, yup, I managed to compile his source with the gnu compiler and libraries and flash it on the ST, did some minor hacking, saw there wasn't much left in the way of i/o pins, grabbed the 100 pin version of the cpu. The piccolo is interesting, but the ST is more than adequate for now, and a very popular platform, so I'll probably stay with ST till I get some sense of mastery of it.