on heat rate control by changing PID parameters

[DeepSOIC]

Miriam seems to have this MELT mode, which (as far as i know) is supposed to raise the temperature very slowly. It is achieved by setting different p,i,d parameters. This is a very strange design, and the resulting heating rate is very unpredictable.

The way i think it should work is to have one set of PID values, tuned for best response (fastest settling to the setpoint with minimal overshoot). Then, if you need to heat it in a controlled way (for example, slow linear heating), you vary the setpoint according to how you want it to heat up (Tset = T0+time*heating_rate), and the pid loop is to take care of following the requested trajectory. At least, that's how we are doing it here for thermoprogrammed desorption spectroscopy. This should make heating process highly predictable, reproducible and easy to set up.

PS. A miriam with all cells occupied may need slightly different p,i,d settings than one with just one cell occupied, because heater response will change, potentially making the control loop unstable, or too slow.