pid temperature control allen bradley

[irondesk40]

go to work on this again this morning. I have attached the file and have a question that i can not figure out.
When monitoring the temperature i see the value from the thermocouple card. For example if it is 90 degress F, it shows 90 degrees F. But watching the screen on my laptop when i raise the setpoint to turn on the heater to reach the new setpoint i watch the value from the input of the thermocouple card and it shows it rising. For example if i go from 90 to 110, i see it changing from 90,91,92,93 etc., but i also see the value go to a wild number such a 120 or even a different number but then goes back to showing the actual increasing of the temp. it is just a heater element hanging in the air at the moment and a j thermocouple mounted and touching it. Any thoughts why i would see this happen. I know there is a lot i do not understand about PID loops but trying to learn as much as possible.
thanks

 

[widelto]

Does the same occurr when you new setpoint is lower than value at the moment.? I mean, does value goes much lower than setpoint.?

 

[rootboy]

Since what you are reading is the process variable, the PID shouldn't have any effect on it.

Have you used thermocouple (extension) wire entirely in the circuit? And the right connectors? Is it an old thermocouple?

None of these things should cause bouncing (a worn out thermocouple might), but you should be aware of the potential problems.

What I would do is to take the thermocouple out and test it in a water bath before doing anything else.

 

[Lancie1]

it is just a heater element hanging in the air at the moment and a j thermocouple mounted and touching it.

Move the TC far enough so that it is not touching the element, and see if that helps. The TC weld could be shorted to the sheath, or the heater element also could be shorted to its outer sheath.

 

[irondesk40]

thanks for replys. have been off for a week and now back to work and back at this. i am now trying to change the filter from 60hz to 10hz to see if that helps.

 

[irondesk40]

changing the filter value seems to have fixed the issue i was seeing. i have attached the current test file here on the bench.
if anyone takes a look and has a suggestion on how i may change it to work better please let me know.
thanks