Good night people, I'm starting with a design for a frequency inverter. I want to control the motor using the space vector modulation, but I am in doubt of how to write the vector will be rotating between sectors 1 through 6. Can anyone help me? Thank you.
Any particular reason why SPWM can't be used instead? What arduino are you planning to use?
am also try to develop for svm inveter..
can tell me how to set a counter for audrino uno
"The motor" - what kind of motor?
sVPWM is advanced pwm tech...i am using induction motor..
annsmathew:
sVPWM is advanced pwm tech...i am using induction motor..
By far, the hardest part on accomplish such a project is to find a suitable chip with a motor control PWM module. This makes sure the different phases are synchronized and generally features a hardware override input, in case the IGBT/MOSFET driver detects a short circuit.
Implementing such features in software is possible, but a good recipe for disaster.
Other than the specialized AT90PWM chips, some newer attinys have this feature. The due has this feature too. It all goes down to the required performance. Once you get this running it's 'easy' enough to implement SPWM. SVPWM could be a next step if you wish to take the project further, but at this point you'll probaly wish to migrate to ARM/AVR32.
I should finsih by mentioning this is not a stardard arduino project. It will require some low level programming working directly with the device registers, which is a reason this is not a common DIY project. Best of luck.
The Mega has some counters that drive 3 PWM pins which is convenient for
generating 3-phase PWM, and on the Uno you can synchronize the counters
so they run in lockstep with care.
Lets be clear which method you are using though, as the literature can be
confusing here.
Are you selecting the output drive cycle-by-cycle (also know as hysteresis
or bang-bang driving), or are you using a PID loop to select a rotor-referenced
phasor which gets converted to stator coords and sent via standard 3-phase PWM?
For hysteresis control you typically use table lookup based on actual (estimated)
and desired rotor current phasor and the rotor angle.
The PID approach people typically separate into two loops, one for phase and
one for magnitude, but I don't think that's very smart, you can use a single
PID loop in phasor space directly which reduces the amount of inter-conversion.