Sign up

[sousvide59]

Given that pretty much any temperature or pressure you measure is only a proxy for the temperature at the surface of the coffee puck, there will always be some transfer function, and in particular there will always be an offset. Since the function of the Integral loop in a PID controller is to slowly null out the offset that's inherent in the P control loop, wouldn't it always be better just to include both of these offsets in the choice of a set point? Am I missing something, because it seems to me that the only effect of a nonzero parameter in the I loop will be to slow down the system's arrival to steady-state.