12/8/2013 11:53 AM  
Joined: 10/1/2012 Last visit: 12/12/2021 Posts: 13 Rating: (0) 
Dear all Here is the case: 1) there is a ACDrive for driving the Kiln's Filter Fan with analog (420mA) setpoint and actual of speed (055Hz) 2) There is a Pressure transmitter to sense negative pressure in kiln inlet with the range of [15 to +15 mbar] 3) I need to fix the negative pressure around 0.4mbar with fluctuations of +0.1mbar [0.3 to 0.5 mbar] with varying the speed of Filter Fan. 5) The problem is how to use FB41 in this case or any other solutions with other FBs could benefit to this issue? 6) Should I normalized the input or output? 7) Are these two parameters (Pressure and Speed) have linear proportion or logaritmic proportion and is it important to know and involve in calculations? 8) Should I involve the kiln input tonnage as an effective parameter in negative pressure and how? Please let me know the best solutions. Best Regards Mehdi Zarean 
Last edited by: mzarean at: 12/8/2013 12:01 PMLast edited by: mzarean at: 12/8/2013 12:01 PMLast edited by: mzarean at: 12/8/2013 12:00 PM 

12/8/2013 5:05 PM  
Joined: 1/28/2009 Last visit: 2/21/2023 Posts: 6794 Rating: (1341)

Hello, This is not an application that I have worked with before, but there are some points about PID controllers in SIMATIC range of products:
I hope these comments help you Hamid Hosseini 
This contribution was helpful to1 thankful Users 
12/8/2013 5:22 PM  
Joined: 10/1/2012 Last visit: 12/12/2021 Posts: 13 Rating: (0) 
Hamid jan thanks for reply We can assume that they have linear proportion , what to do so? The only thing I know(even not sure) is that The SP_INT = 0.4mbar and I've been confused what could be the PV? 
This contribution was helpful to1 thankful Users 
12/8/2013 6:09 PM  
Joined: 1/28/2009 Last visit: 2/21/2023 Posts: 6794 Rating: (1341)

Hello again mzarean, and thanks for your feedback Linear process guarantees troublefree operation of the whole controller and when a nonlinear element exists in your system, you are getting more confused by changing PID parameters with no success. Again, I insist on using normalized values, It is based on results in practice that the algorithm works better if the PV and SP_INT are normalized. To give you a basic insight of using PID controllers, they are hardware or software technological objects that are coupled to a process to regulate. Controllers manipulate process by an output called LMN and it can be continuous on control valves or binary on heaters/integrating actuators. The physical changes are read by transmitters and are internally connected to PV.SP_INT is your desired point and normally a constant over a long period of operation , process value should reach that with a minimum oscillation and over(under)shoots.PV and SP_INT have same physical dimension and they are preferred to be normalized. This task is automatically done if you process PV from the "Peripheral Processing" section of FB41.When machine is in standstill, sign of SP_INT  PV determines the action of controller. and this is only a small portion of tips on regulation , read more similar topics in the forum for expanded practical samples. I hope you enjoy practicing regulation with SIMATIC Hamid Hosseini 
Last edited by: hdhosseini at: 12/8/2013 8:56 PM// typo 

This contribution was helpful to2 thankful Users 
12/9/2013 7:33 AM  
Posts: 2348 Rating: (259)

As Hamid already mentioned it's better to use normalized values. But just to make things clear  you can have setpoint in mbars (0.4 in your case) and input value in mbars (+15 in your case) also, but manipulated value can be anything  in your case it's 420mA, but generally it can be anything as soon as it can influence input value. If something influences PID regulation it's generally connected to DISV input. But i recommend going for simple PID regulation first (without disturbance) then just see if your regulator can handle disturbance as it is, and only in case it can't  you can try to solve it by changing PID. 
This contribution was helpful to1 thankful Users 
12/9/2013 8:59 AM  
Joined: 10/1/2012 Last visit: 12/12/2021 Posts: 13 Rating: (0) 
Thanks dears hdhosseini and Aret up to here from some formulas and equations about fan I've reached to this equation which shows the relation between frequency and pressure : f = k . (Q.p) / (F.r) f : Freq k : Constant Q : db p : pressure F force [N] r : Radius So from the equation and my personal sight observation the conclusion is sth like below proportion : f ~ p . (ton) . (Burner gas) but about the negative pressure I have no idea how to deal with it in pid and could it have direct proportion with negative gain? Generally speaking I need to understand how to involve in PID parameters these two examples : 1) X = Y 2) X=1/Y Best Regards Mehdi Zarean 
Last edited by: mzarean at: 12/9/2013 9:02 AM 

This contribution was helpful to1 thankful Users 
12/9/2013 9:12 AM  
Posts: 2348 Rating: (259)

What for you trying to solve your process equation analytically? It doesn't needed for building PID. Please normalize inputs  setpoint =0.4 mbar of range +5 bar is 46% and have nothing common with negative gain. Negative gain just means "if (input < setpoint) we should decrease output"  nothing else. Please mention in this pseudocode unnormalized input and setpoint can be both negative or positive  generally it doesn't matter. 
This contribution was helpful to1 thankful Users 
5/26/2014 9:41 AM  
Joined: 4/24/2009 Last visit: 2/23/2023 Posts: 2685 Rating: (135) 
New question splitted to How to use FB41 (CONT_C)? Best regards O_Moderator 
Follow us on