Incorrect use of setIZone()

[brettle]

RobotCode2024/src/main/java/org/carlmontrobotics/subsystems/Arm.java

Lines 180 to 184 in 820310f

	if (armAtSetpoint() || getArmPos() > setPoint.position) {
	armPIDMaster.setIZone(Double.POSITIVE_INFINITY);// turns off pid once it reaches the setpoint
	} else {
	armPIDMaster.setIZone(IZONE);
	}

I'm not sure where this came from but:

1. The IZone only has an effect if kI != 0.
2. Setting the IZone to infinity does not turn off PID, it says to use the integral term of PID no matter how far away from the setpoint you are, which is typically not what you want.
3. The current code calls setIZone() 50 times a second, which the docs don't recommend.
4. The variable name IZONE does not imply any units but it's units are radians.
5. The current value of IZONE is 4. setIZone(4) says "only" use the integral term if you are within 4 radians of the setpoint. But 4 radians is over 180 degrees, so you'll always use the integral term.

I recommend removing all of this code for now.

In the future, if you are going to use an integral term (ie kI > 0) to remove any error that remains after the arm stops.

1. Fix the variable name.
2. Set the IZone once (like the other PID constants).
3. Set it to a value that is slightly larger than the largest error that remains when the arm stops.
4. Increase kI from 0 slowly until it closes the rest of the gap.

[FriedLongJohns]

Fixed in:
ddf7573
and
2f33075

[brettle]

IZONE_RAD is currently 0.8 radians. If there is anywhere near that much remaining error, you have other problems. Recommend just setting it to 0 or infinity for now with a comment that says to change it before using it. Perhaps referencing this issue.