PID Temperature system for thermocycling

[CBMM]

I use FieldPoint 1000 with PWM 520 and TC 120 to make thermocycling with PID functions. I've tried to do that, but temperature is not stable. The temperature graph looks like stairs-shape. When I use Watlow temperature controller, the graph looked like linear or curve line. I attached my file, but I think there is no problem related to it. I am not sure that this problem is from timing or PID gains which are not suitable with thermocycling. If you guys have experience or advise, please let me know. Thanks.

Tae Yoon Lee

[Brian_Beal]

A couple of things are curious.  You have output limits of -50% to 50%, but the case structure would never allow a negative output.  Does your system have cooling ability?  If not, then your limits should be 0 to 100%.  Other than that, problems like yours are almost always related to tuning.  Have you tried to autotune?  What is the hardware in your system?  What PID parameters are you using?

[CBMM]

Dear Brian,

Thank you for your reply. You are right.  The output limits must be 0% to 100%. I will add the cooling module after stabilizing heating source. I've used a kapton thin-film heater. It is very small. I've tried to autotune, but it cannot be stable the temperature. I've attached the processing picture. The hardwares are Fieldpoint 1000 with TC-120 and PWM 520, external power supply connected to PWM module, and a small heater.  I've got PID parameter from autotuing function, P = 241, I = 0.421, and D = 0.105.

Thanks,

Tae Yoon Lee

[Brian_Beal]

I think you have other problems.  The temperature changes that are shown on the chart are impossible.  The temperature cannot step like that.  Does that step pulse width happen to match the pulse width of your output?  I think that you might be getting interference from your heater circuit that is showing up on the t/c input.

[CBMM]

Dear Brian,

Thank you for your comments. What interference can cause to make something wrong? I have tried to find out the problems, but nothing happened.

Tae Yoon Lee

[Brian_Beal]

Try reading temperature without the heater being powered.  Also try it without turning on the output from the PWM.  Then take a look at the temperature plot and see if the readings are smooth.

[CBMM]

I've checked TC-120 without power being and PWM output. Temperature data is very smooth. I've also tried tuning procedure with autotuning function. However, the deviation is 10 degree C. I can't understand what the problem is. I think it had better to buy commercial temperature controller. Hmmm....

Tae Yoon

[Doug M]

Tae Yoon,
I would make sure you are getting good results from your Fieldpoint modules in MAX.  Make sure your temperature readings are accurate and that when you write to your PWM module the desired behavior occurs with your heater.  Also make sure that the PWM module is using its own power supply and is not jumpering power to the TC module.  I also noticed that your PID loop did not have a loop rate set, this is something you may want to add.