SLC 5/03 on/off temp control

[chipr97]

Needing some help. I have a machine that was put in storage and has ended up in my lap. It had a 5/02 originally but the card was pulled and I have no program. It is tied to a 2700-K10C3 panelview. I have replaced it with the 5/03. First problem is no program for the plc. It has 3 barrel temp cards and one thermal couple card. I have no experience writing PID loop or on off temp control. Looking for help or samples if possible.

 

[OldChemEng]

I can give you some basics starting with the following process assumptions:


a) 3 Barrel Temp Cards indicate to me that this is an extruder
b) The 3 Barrel Temp Cards are Discrete Output Cards that fire Solenoid Valves for Cooling & Solid State Relays (SSRs) for electric resistance heating.


This application is typically done with Proportional + Integral + Derivative (PID) feedback control. The SLC-500 processors have a PID instruction.


Basics for PID feedback control with Time Proportioning Output:



1) Response of these type of Process PID control loops is slow


Example:Electric Resistance PID update times are in the 1.0 to 2.0 second range compare to <10 msec for a Current Minor Loop in Digital DC Drive (Reliance AutoMax)


2) The SLC-5/02 cycle time will probably be in the 10 to 20 msec range.


3) The Barrel Temperature Input from the Thermocouple card is the PV input to the PID instruction & should be filtered (ie:Moving Average) to minimize Common Cause Variation.


4) The output from the PID instruction will be 0.0% to 100.0%.


5) The PID input output will be "Split Range" which means 0.0% is Max Cooling / 50.0% is No Cooling & No Heat / 100.0% is Max Heating


6) The Analog 0.0% to 100.0% PID instruction output will modulated by "Time Proportioning" the Discrete Outputs to the Solenoid Valves & SSRs.


Example: a) With an Output Cycle Time of 2.0 sec & a PID output of 37.5% (50.0% - 37.5% = 25.% of 0.0% to 50.0%) the Cooling Discrete Output would be ON for 0.5 sec & OFF for 1.5 sec.
b) Wind an Output Cycle Time of 2.0 sec & a PID output of 75.0% (50.% of 50.0% to 100.0%)the Heating Discrete Output wouldbe ON for 1.0 sec & OFF for 1.0 sec.


SLC-500 PID instruction programming:


1) The update time of the PID instruction in SLC-500 & PLC-5 PLC MUST be controlled by the PROGRAM. The PID instruction DOES NOT know the time period between successive executions.


Note: I have seen MULTIPLE SLC-500 / PLC-5 PLC programs written with a PID instruction update time of 1.0 sec that execute the PID instruction unconditionally / every scan. This error can be easily detected if the PID control loop is unstable & the P / I & D constants are very small (0.000something).



2) If the rung conditions before the SLC-500 PLC instruction are FALSE then the PID output is set to 0.0%.


Note: This is different than the PLC-5 PID instruction so programming a SLC-500 PID instruction must be done differently.


3) Therefore, put the PID instructions in a SLC-5/02 separate program file & call the PID program file at a time period equal to the PID instruction update time.



4) Make the Moving Average time period equal to the PID instruction update time so any PID output calculation is based on 100.0% new PV input data.


5) PID control action for a Heating / Cooling control loop = Reverse Acting which means the when the PV input increases the PID output decreases & vice versa when the PV input decreases.



6) Scale the PV in engineering units (degrees F, psig, etc) & make sure the SP & PV scaling is the same. There is a Scaling instruction in the SLC-500 PLC processors & you might be able to scale the PV input on the Thermocouple Input Module (maybe).



7) In the PID configuration, turn Zero Crossing Deadband OFF & set the Deadband to 0.1% to 0.5% of the PV scaling range to start. Adjust the Deadband as needed when tuning the PID control loop.


7) Start with a Proportional Gain of 1.0 = Proportional Band = 100.0%


8) Start with a Integral Gain of 0.1 to 0.05 sec-1 (10 to 20 seconds).


Note: This will probably result in an "Overdamped" / slow responding PID control loop but will reduce the potential for an over temperature condition that could damage equipment.


9) Two (2) timers for each PID control loop will be required & the presets will have to be calculated from the PID output & Time Proportioning cycle time which can be greater than the PID update time.


Example: For a PID Output of 12.5%, 75.0% & Time Proportioning Cycle Time of 4.0 sec:


Cooling Output On Time = [(50.0% - 12.5%)/50.0%] x 4.0 sec = 1.0 sec
Cooling Output Off Time= 4.0 sec - 1.0 sec = 3.0 sec



Heating Output On Time = [(75.0% - 50.0%)/50.0%] x 4.0 sec = 2.0 sec
Heating Output Off Time = 4.0 sec - 2.0 sec = 2.0 sec


Note: In Extruder Barrel temperature control using Split Range output the Cooling & Heating ranges sometimes "Overlap":


Cooling Output ON = PID Output 0.0% to 55.0%
Heating Output ON = PID Output 45.0% to 100.0%


This will change the Denominator in Output On time equations above.



10) Timers Presets for the Heating & Cooling Time Proportioning Output are equal to the Time Proportioning Cycle Time.


11) From the example in #9, when the Timer Accumulator is LESS or Equal To the calculated On time then turn the Discrete Output ON & when the Timer Accumulator GREATER than the the calculated On time then turn the Discrete Output Off.


I would recommend reading up on Tuning PID control loops but remember that when tuning a PID control loop the object is to match response of the PID control loop to the natural built in process response (not what you want to be). What I have listed are the basic for PID control with an Allen-Bradley PLC-5 or SLC-500. I would recommend reading some good books on PID control.

 

[KingPresident]

^ Bookmarked this one ^

 

[chipr97]

Thanks for all the information. It’s a pin paste wax machine. Janke is the make. I’ve only been at it about a day. All the btm cards are reading and I have got one scaled where I can do a simple on off. Looks like I’ve got some studying to do.