Analog Input of Heater Fluctuating

G

Thread Starter

GlennAvans

Let me introduce our project first, before I ask any questions.

Our project is consisting of two water tanks, pipes, valves and a pump. We have to make a working process control. We are dealing with a pressure control, flow control, level control and temperature control. We have to make a SimuLink model in matLab so that we can find the ideal values to control the system. The analog input of the heat sensor (Siemens 7MC1006) is fluctuating with +/- 1000. What could be causing the problem?

We are open for comments and answers.

Greetings,
Team B, Avans
 
Open lead or connection?

Connected to wrong terminals?

Improper setting (e.g. set to TC, not RTD)?

You not giving us much to go on.
 
Why are you bothering with a silly process control project when you've achieved a miracle in physics and should be wrangling for a Nobel Prize?

Absolute zero is 273.15° but you've achieved -1,000 Deg X!

Ordinarily engineering units are critically important in any engineering or scientific work, but I'm sure the Nobel committee will take either C or F or K, given your landmark achievement.
 
I think the sensor is defect, try replacing it with another sensor. It could be a programming fault. Are you watching the value or did you manipulate it at first?

Good luck with the project.
 
If the oscillations are not due to Smart_S's reasons or a defective sensor, but are due to a PID loop chasing its tail, then you can reduce the oscillation due to the PIDs aggressive tuning.

First, see if you can't configure the range of the temperature units associated with the analog input down to some reasonable range, nothing negative unless you have refrigeration control.

Second, to stabilize the PID:

Turn the rate/derivative D term off, set it to zero.

Turn off the reset/integral I term, set it to zero.

Halve the gain value (multiply by 0.5) in iterative steps until the oscillations stop. If the setup uses proportional band (PB) rather than gain, double the PB (multiply by 2) in iterative steps until the oscillations stop.

It doesn't matter the PV droops below or above the setpoint, it needs to be stable first.

Once you have some stability, you can add in whatever reset/integral factor you need to correct for the droop. Start with a relatively high 'minutes' value, like 2, and watch what happens. The lower the minutes reset value, the greater the reset action.

Then you can play with rate/derivative term and get it oscillating again.
 
Top