2004 Robot Controller

[Mike Soukup]

Originally posted by GregTheGreat
The one thing I don't like is that they are keeping the same OI, so whatever your transmit rate is (the 28 ms...I think) is the same rate you will transmit this year. That basically eliminates the whole use of the fast PWM's (except for auton mode)

Not true. The fast PWM's are quite valuable for feedback loops even in user control mode. A lot of the latency in feedback mechanisms such as crab wheel positioning & arm angle is in how often the motor speed is updated.

Think about it this way. The driver of a crab style robot doesn't change the desired direction of the wheels every 28 ms, and as far as the driver is concerned the delay between moving the stick & the robot getting the signal 28 ms later is negligible. The big problem is when the driver moves the stick and the wheels overshoot their target or when start oscillating back & forth. This is where the fast PWMs come in. Analog & digital inputs on the robot are updated every program loop, so we can take the driver's desired target (which is constant for 28ms) and the fast updating crab position and use those to control our feedback loop. Now we get to update our desired motor speed every time through the program cycle and the fast PWMs allow us to send that value to the speed controllers. Does that make sense?

Mike