Hello folks,
I have two questions I can't answer myself and would like to politely ask you guys for help:
Question one would be:
During the FOC loop the spacevector-modulation takes place.
For this the respective mosfets are switched on and off accordingly, as it is nicely described for example here: [link]
My first question is: Does the duty-cycle in FOC-Mode control the duty-cycle of the S(pace)V(ector)M(odulation)?
Or in other words: Does the duty-cycle in FOC-Mode control the magnitude and thus the length of each of the six base-vectors (which then results in a defined current flowing through the stator-windings?)?
My second (and for me most important) question would be:
What is controlled by the "switching frequency" (20 khz by default) exactly?
Is it the switching-frequency of each mosfet?
I am asking this because I'd like to understand how the switching frequency affects the electrical rounds per minute (and thus finally the mechanical rounds per minute).
I would like to understand how these two quantities are linked together.
The faster I drive or the higher my erpms get the less time a mosfet has got to switch on or off.
How often does a mosfet have to be turned on and off until one can speak of a well modulated vector (as the combination of two base-vectors)?
Is there a factor?
I hope I did not express myself too bad..
Is there anybody who can help me out?
Thanks a lot.
-Pete!
Anybody ? :(
I would say your questions need to be answered by Benjamin himself. In general, switching losses should be kept low so in most cases it it is advisable to go for lower switching frequencies.
20Khz is a very good default value, high enough for smooth FOC operation, low enough to avoid a lot of switching losses. Precision (resolution) will go up with higher switching frequency, but losses will also go up.
Some motors need higher switching frequency BTW! It really depends on if you run a low inductance 2 Pole Pair Inrunner with a high KV value or a a torque rich Outrunner with many poles pairs and low KV. value.
One thing that you can do is go from 20 to 30 KHz and monitor the Mosfet temperature and how the switching frequency will affect the temps.
Thank you for your respond.
I understand that the switching-frequency affects the switching-losses.
I do not have any problems at the moment it is just my curiosity. I want to get a better understanding.
The answers are in the code, I should start with "mcpwm_foc.c" I guess..
If there is anybody who can give me a little explanation on the questions above I would be thankful.
Maybe I'll be able to understand the code and answer the questions by myself.
(If that's the case I will post them here)
And since I've forgotten to say it so far, this is a really great project! Big shout out to everybody who pushes it, especially to Benjamin of course!