PID tuning constants conversion

jkerekes

Member

Join Date

Aug 2007

Location

NJ

Posts

2,362

• Dec 9, 2010

• #1

I'm doing a PLC5 to CLX conversion. There are several PIDs in it. Are there any conversion factors that I could use to "translate" the tuning constants used in the PLC5 to the CLX. BTW I'm using the standard Ladder PID instruction in the CLX. Thanks!

 

Honga

Member

H

Join Date

May 2010

Location

Aus

Posts

316

• Dec 9, 2010

• #2

I believe there was a thread about this maybe 2 months ago. If you try a search I believe the PID gurus have been through this with another patient.

 

Peter Nachtwey

Member

Join Date

Apr 2002

Location

No income tax, no capital gains tax. Freedom!

Posts

8,377

• Dec 9, 2010

• #3

What question would a PID guru ask before answering?

The question cannnot be answered until........

 

Honga

Member

H

Join Date

May 2010

Location

Aus

Posts

316

• Dec 9, 2010

• #4

Screenshots! Programs! Lupus!

 

Mickey

Lifetime Supporting Member

Join Date

May 2003

Location

Palmdale,Ca

Posts

10,656

• Dec 9, 2010

• #5

Honga said:

I believe there was a thread about this maybe 2 months ago. If you try a search I believe the PID gurus have been through this with another patient.

Click to expand...

This one..?

http://www.plctalk.net/qanda/showthread.php?t=58158&highlight=ron+PID

 

Honga

Member

H

Join Date

May 2010

Location

Aus

Posts

316

• Dec 11, 2010

• #6

Yeah it was that one mate. Thanks for doing the hard yards.

 

Mickey

Lifetime Supporting Member

Join Date

May 2003

Location

Palmdale,Ca

Posts

10,656

• Dec 11, 2010

• #7

Honga said:

Yeah it was that one mate. Thanks for doing the hard yards.

Click to expand...

No problem, now back to my lawn chair.

 

jkerekes

Member

OP

Join Date

Aug 2007

Location

NJ

Posts

2,362

• Dec 11, 2010

• #8

I did read the link provided. This will be a conversion run through the new conversion utility. All the existing 1771 I/O will be used, so the scaling of inputs & outputs should be the same. The PID gains are independant however.

 

Ron Beaufort

Lifetime Supporting Member

Join Date

Jul 2002

Location

Charleston, SC

Posts

5,701

• Dec 11, 2010

• #9

you still need to consider whether the "triggering/timing" of the original PID was correct ... quite often this is NOT set up correctly - and the original programmer just "fudges" the tuning values to compensate ... (this post about Bubba's paycheck explains the concept) ...

what this means to you ...

suppose that your original PLC-5 has been improperly executing the PID "once per scan" (a very common problem) and the scan had been happening 200 times per second ... suppose that your original tuning values had been "fudged" to make the PID function adequately well ...

now then ...

suppose that you simply convert the "faulty" logic from the PLC-5 over to the ControlLogix platform ... suppose that now your new ControlLogix system begins scanning the PID 800 times per second (it's much faster) ... if you keep the same "fudged" tuning factors, then the TIME-BASED components (Integral and Derivative) will be "skewed" and the tuning will certainly be wrong ...

what all this means to us ...

we really need to see the OLD program from the PLC-5 – and the NEW proposed/converted program for the ControlLogix in order to give you a concrete detailed answer to your question ...

can you post the code we're asking for? ...

 

Last edited: Dec 11, 2010