PDA

View Full Version : How "smart" are these e-vehicle controllers?



CliC
08 July 2012, 1715
Merely a curiosity question. I deal with industrial variable-frequency AC induction drives at work, and they''re quite sophisticated pieces of equipment. Most now have various modes of control (vector, constant torque, etc.) to meet a wide variety of industrial applications (which of course wouldn't be required on e-vehicles).

One thing they do now is to run some tests on the motor at first startup, and build a computer model of the motor/cable system, which is then computed while running by a DSP chip and used for the control. Are the e-vehicle controllers that sophisticated, or are they more V/Hz type things (ignoring for the moment the temperature-dependent current limiting and other protective functions)? Would there be a big advantage gained if they were that sophisticated (if they're not)?

Doctorbass
08 July 2012, 2144
Merely a curiosity question. I deal with industrial variable-frequency AC induction drives at work, and they''re quite sophisticated pieces of equipment. Most now have various modes of control (vector, constant torque, etc.) to meet a wide variety of industrial applications (which of course wouldn't be required on e-vehicles).

One thing they do now is to run some tests on the motor at first startup, and build a computer model of the motor/cable system, which is then computed while running by a DSP chip and used for the control. Are the e-vehicle controllers that sophisticated, or are they more V/Hz type things (ignoring for the moment the temperature-dependent current limiting and other protective functions)? Would there be a big advantage gained if they were that sophisticated (if they're not)?

These BLDC controller are also quite sophisticated. One of our guys on the Endless-Sphere forum developped a variable timing interface that compensate for the inductive negative effect that make the timimng a little bit off as the rpm increase.. so his device have a programmable timing map for different throttle positions!.. I have one of these that i installed and it is really impressive but require a bit of work to get the best efficiency out of it. It connect on the hall sensor wires between the motor and the controller. The guys name is Burtie on the forum

Doc

podolefsky
08 July 2012, 2227
Nothing like that in the usual controllers we use (Alltrax, Kelly, Curtis). Simplest controllers, like Alltrax, are just buck converters with some parameters for throttle mapping and current limiting. Curtic AC induction controllers have a lot of parameters and they're tuned to the motors that come with each package. You can set them for torque of speed control, tune the field weakening and stuff like that. On start up they just check for shorts and such, nothing like what you're describing.

Not sure about the more expensive stuff like Rinehart, or what goes in commercial EVs like the Leaf. If that kind of processing could get better efficiency or power, that would be great.

CliC
09 July 2012, 2253
Thanks for the responses. Sounds like the Curtis models are closest to what I've dealt with.

All the modeling in the ones I've seen is only done on the first startup, or when manually initiated afterward.

I wasn't sure that power or efficiency gains achieved were worth the extra design effort, and perhaps chip cost. The ones I use at work probably are built the way they are in order to be adaptable to a wide variety of industrial applications. There are hundreds of parameters in them, but many can be left at default for standard pump/fan applications.

One cool feature I saw on the last one I commissioned is a form of regen braking -- not to brake the motor, but to extract a bit of energy to keep the drive powered up on loss of input power as long as the motor is spinning above a certain rate. It allows an instant automatic restart and ramp up of the motor when input power is restored.