[Ken Streeter]

Originally Posted by baop858

... But has anyone notice that the victor are uneven, like in one direction there's slightly less power than if it was going in the other direction with the same amount of difference from neutral, 127.

Short answer: use 132 for the center, rather than 127. Yes, this is contradictory to everything in the IFI documentation and on their web site. However, our empirical tests bear this out again and again, on different motors. How did we get to this determination?

Well, we observed in 2005 that our robot's right side and left side drive motors seemed to be mis-balanced, and wrote the matter off to the fact that the motors on one side were going forwards and on the other side were going backwards.

In 2006 when programming a PID control loop for shooter azimuth control, we became convinced that the actual center for the victors is 132, rather than 127.

In 2007, we did yet more PID control for a multi-jointed arm, and again found the center to be 132.

In 2008, with PID control for the rack-and-pinion steering on our "Speed Racer" we have again found the PWM center to be at 132. We don't have an oscilloscope to verify this, but have verified this empirically again and again.

For another datapoint, check out the camera control code at kevin.org. Oddly enough, you'll find that the center value used is 132...

Best regards,

--ken