PIDD Telepace

[willB1]

Trying to develop an easy way to run a vent from fully open to fully closed and all positions in between using a PIDD block. The opening/closing is done to maintain a temperature in a room (set point 75F). Fully closed-open takes 60 seconds. Iv'e set the full and zero at -50 and +50. Each output controls independent relays (1 open and 1 close). Is there a way to use PIDD for this purpose?  I'm obviously missing something because i can't get it to work like i want it.

[Joel_Weder]

Hello WillB1,

This should certainly be possible to manage with PIDD. You'll need a temperature input to the RTU as your Process Value, and two digital outputs. One will pulse to increase temperature, and the other will pulse to decrease temperature. That's probably the most challenging part of this, getting those digital outputs properly wired to your vent control motor. We can't really help you much with that, as it's so specific to your application.

Not knowing whether you've taken a Telepace course or not, I can certainly recommend you sign up for one if not, and if there's one available in your area. However, I'm attaching a few pages from the training manual, covering the PIDA exercise. Unfortunately there's no PIDD exercise, but they operate in a generally similar way. The Help documentation for PIDD is pretty good, though.

[willB1]

Thanks for the fast response. I have the input temperature coming into the PIDD and the outputs wired correctly. I guess the issue is understanding the cycle time vs other tuning inputs to allow the vents to open/close a certain distance (time)/ distance the temp is away from set point 

[Joel_Weder]

@willB1,

Probably the most challenging aspect of PID control is tuning of the parameters. The exercise I attached does include some pointers about initial settings of the Proportional and Integral parameters. (Derivative is rarely used in cases such as this) Also, any scaling in logic can dramatically impact the results if done incorrectly. 

If after working through the exercise and Help documentation for PIDD you're still having issues, please feel free to contact our Tech Support team. They'll want to see a copy of your logic as well as a drawing of the wiring you've implemented.

SupportTRSS@se.com  or

1-888-226-6876

[BevanWeiss]

Are you able to break the problem into two parts (i.e. do you have any positional feedback from the vent?)?

If you were in the RemoteConnect / Unity world, then there is a SERVO function block that would allow for the Vent position control, and then it should be a relatively straight forward approach to wrap the commanded position with a PIDFF to perform the Temperature -> Vent position control.

Or as Joel suggested, you could tailor the scale to a timeduration, and every X seconds (or minutes) trigger the valve movement for Y seconds (based upon the value of the control output from the PIDD).

I'm not familiar with the Telepace platform, but sometimes there is already a split range controller.  There's nothing like that there?

[willB1]

Thanks for the help. I have not configured a PIDD. So here is the issue. The total open-close time is approximately 60 seconds. I'm having trouble with the min/max where the pulse output isn't consistent. Like you, I'd like to have it open/close for say 10 seconds to increase or decrease temperature. Thanks again. 

[PLCs_4_Me]

We use PIDA to control the water level at our pump station at 45 inches via VFD. We use the SCAL function element with a pressure transducer with 4-20mA signal to set the process value we need (45 inches 4mA=0 20mA=415 inches 15 PSI transducer). In your situation it sounds like you need a RTD temperature sensor (measurement range 0-200 degrees F) with a 4-20mA output signal and connect that to your PLC analog input. SCAL the analog input with the RTD. Once that is done, you can set up your PIDD Block to input all the necessary info.  Your set point is 75. Your Gain, reset time, rate time and deadband. Full (%) I assume would be +30 and Zero (%) would be -30 ( for a total of 60 seconds which is the total time to close.) Then you just have the logic compare the set point to the actual temperature reading and create two blocks to control the digital outputs to either open the valve or close it depending on the values (actual temp=100 F > 75 F setpoint open the valve) etc. Let me know if this helps.

[BevanWeiss]

PIDA is generally to control an analog output, and it expects that its output is acted on each cycle.

This is different than a PIDD where the output is a digital output that will be integrated over time to get the control response.

The front part is the same however, so there does need to be an analog input for the temperature, to be compared against the setpoint temperature.

The total time to open/close is only really relevant as it pertains to the gain of the system.  The biggest impact is likely to be the system response to actuator position however.  So when the vent is opened, how long it takes for a temperature change to be detected.

This is where the 'loop tuning' starts to factor in.  With a long system response time, then the system gain should be reduced, otherwise you will end up with overshoot (and likely response oscillation).  I would suggest starting with a very low gain, and working upwards from there.  Likewise, having a low Full % value is likely a good start.  Perhaps as low as 10% (6 seconds of valve movement in each minute).  The cycle time will also relate back to the system response time lag.  No point having the PIDD executing frequently and 'winding up' its response if the previous output change hasn't had any time to propagate yet.

It's also important to remember that a PID expects a somewhat linear characteristic of system response.  This often isn't true, especially when valves are involved.  It may be desirable to perform some linearisation process.  Unsure how that would be done without position feedback on the valve itself... but all of these things become factors in the control system.

For reference, I would consider the best case control to be what I mentioned originally.  Split the system into two loops.

The first controls for temperature against a valve position.  The second controls valve position (based on valve potentiometer feedback or similar) and runs a motor to open/close valve.  This is generally how things like air conditioning systems with variable dampers work.  It really makes the whole control easier.