Controller Tuning

Check whether the controller is set for direct-acting or reverse-acting as a first step, then re-post below.

Ya, it's probably all in the setup. This instrument may also have an output limiting feature. A complete parameter check is in order.

Sounds like it may be the conservative settings that autotune usually come up with. If you can afford to play with the process a bit you could try some different things. Generally I halve the Proportional Band, double the Integral and get rid of the Derivative to zero. This assumes that your controller uses % and not Gain for Prop Band, MinsPerRepeat and not RepeatsPerMin for Integral. Also assumes you can afford some overshoot of the setpoint ie: SP=26Deg the PV may go up to 30Deg before cooling back down to 26Deg. If you cannot afford overshoot then you will need some Derivative for certain, Derivative is tricky to tune if you are not a seasoned hand.

It sounds like the Autotune has imposed the least responsive curve (which is the cooling curve) to govern the whole controller tuning values.

-ROK-

Sometimes (just sometimes) these controllers need an additional line of code in them.

The software will try to use the whole duration of the program step to reach the end point. (Like a ramp function.) The controller will sense the temperature rising and only provide enough input power to reach the SP by the end of the program step.

Suggest you re-write the program with an additional line of duration 1 minute with SP at 26 degrees. (Unless you want it to warm over a longer period.)

Learnt this with our first machine as you never knew what the starting ambient was.