## Driving a three phase inverter using DSP f28335 C2000

Started by 8 years ago2 replieslatest reply 8 years ago473 views

Post deleted by author

[ - ]
Reply by November 19, 2016

So, you're depending on The MathWorks to write your code for you, and it doesn't work?

How are the PWM signals from the dspace 1104 and the f28335 different?  How do they not behave the same way?  Are they different PWM frequencies, different sine wave frequencies, different amplitudes, what?

You say "asynchronous machine" -- do you mean an induction motor?

Are you using current feedback?  If you're making a variable frequency drive for an induction motor (or, for that matter, a synchronous motor), you need to vary the peak-peak voltage along with the frequency -- ideally you'd be monitoring current and adjusting the phase of the current (I'd like to tell you what to adjust for, but I haven't done this sort of thing in real life so I could be wrong -- surely there's a book, though).

Do you have current feedback?

Are you using it?

The general method to take when debugging something like this is to start by understanding how it's supposed to work, then compare what it's doing with what it's supposed to do, then for each instance of it not doing what it's supposed to do, formulate a theory for why, try it, and repeat until you have that problem fixed.

[ - ]
Reply by November 19, 2016

I don't understand your way of working. First, can you explain what kind of system you are designing *precisely*? What are the preliminary requisites?

Simply said, driving an asychronous motor from a 3P inverter is extremely easy to do. You just need a PWM modulator driven from linear U/F law. The law can be adapted after the design phase, in order to compensate the motor characteristics (most U/F laws used in industrial inverters have a non linear part on bottom, in order to get a more efficient magnetization at low speed, to avoid excessive sliding)

My second question : why do you want to use a DSP for that? The first inverters I designed 20 years ago were based on 80C32 microcontrollers, and they even included tacho feedback support. I only use DSP when very specific feedback (like current feedback is needed)

My third question is : what do you expect from your software tool? You say you get something different. Different from what? Getting different results between two simulations is not something extraordinary when the target is different. I would say that you do not know what to expect in advance, so when the tool gives you two different results, you can't tell if one of them is acceptable, because you can't tell what you expect. You should apply the same rules as we use in software design world : UNITY TESTING! Before using a simulation tool, compute and define what you expect to get. Then define how you will check that you get the expected results. Then, when you get a result from simulation or prototype, you can check if it is acceptable or not.

By the way, defining IGBT maximum operating frequency as a single limit is a non-sense. Never consider IGBTs alone. Always IGBT+GDU (Gate Drive Unit). The IGBT capabilities are defined at 40% by their GDU. Never expect a GDU to be able to put the IGBT at its maximum capabilities.