Originally posted by Elcheapo
View Post
Not knowing the details of your setup, I offer this as general information to you, and everyone else.
If you are using PWM, then speaking from experience pulse wave modulation of current did not work for me. Why? I have a theory and that theory is that PWM is not at all the same input as constant DC current at a certain value. Just look at the PWM scope trace in the ms range. Figuera and Buforn did not use hundreds of full current pulses spaced with 0 current to feed the coils.
Also speaking from experience, and also from analyzing the patents, breaking the input into equal time slices for each current level does not duplicate the signal that the commutator of part G delivers. It does not look like the usual AC sine wave. Draw the commutator out with the jumpers shown in the 1914 patents. Each commutator segment is a time slice. Graph it out yourself with the brush always in contact with two segments, as per the patent, and you will see. Even the size of the brush makes a difference. Using the proper time values made a HUGE difference in my early build even with PWM.
If you are attempting a build then build it as exactly as the patent shows as possible. Weigh any advice or theory you read on the forums against the patents and judge accordingly.
CM
Leave a comment: