Converting PLC5 to Studio5000 PIDs

Hey all,

I'm pretty new to PLC programming and am on a team doing a PLC5 to Studio5000 conversion. After doing the built in conversion, I've begun to debug what didn't transfer correctly. One thing my mentor brought up is that the PIDs may update the way that I intend them to. I'm wondering if any of you have experienced this and what to watch out for, both in PIDs and general PLC5 conversions.

Thanks!

Welcome to the PLCTalk forum community !

PID equations require an accurate measurement of time in order to function correctly. In the PLC-5 PID, you had to both enter the Loop Update time and write logic that executes the PID once every time that Loop Update time elapsed.

The execution was sometimes done with Timer instructions, and sometimes with interrupt routines (STIs).

And often it was done by coincidence. Folks just put the PID equation on an un-conditioned rung so it executed every scan of the PLC-5 program, and then did their best to tune it with gain values that would work with that scantime.

The ControlLogix ladder-logic PID works the same way; you both enter a loop update time and execute the instruction periodically. You can use a Timer or a Periodic Task.

But if your logic was written without consideration of the loop update time, then the much faster ControlLogix will execute the instruction much more frequently, and the tuning gains won't work.

The ControlLogix Enhanced PID (PIDE) is a little more subtle; it is set up so that it inherits the loop update rate from the Task it's executed in. Those instructions are in Function Block routines, and are generally executed in Periodic Tasks.

There was a great thread on the topic a few months ago, and I'm sure other folks will have more input. But that's the basics.

something else to think about ...

are you keeping the PLC-5 platform's I/O (1771 style) - or replacing it all with new ControlLogix style hardware? ...

if you're keeping the older hardware - for example 1771-IFE analog input modules and 1771-OFE analog output modules - then I'd strongly suggest that you get those pieces of the puzzle working BEFORE you start tackling the PID instructions ...

TIP: the more DETAIL you give us about your project - the more DETAIL we can put in our answers ...

and welcome to the forum ...

That's helpful! It makes sense that the timing would be the significant difference between the two, I would imagine that trial and error would be a large part of figuring out how to adjust your values. I'll read up on it some more and see how they did it in the original program.

Thanks!

Ron,

Thank you for the reply. We are completely replacing the 1771 chassis with 1756, all components will be new. The plan is to consolidate as much as possible (16 Inputs to 32 Inputs and whatnot) then directly move I/O cables to their new counterparts.

Using the project migrator was helpful because it's a short time-frame but the challenging part is discerning what different bits of info are. The tool just pulls in the data file names from PLC5 and their comments and turns them into ControlLogix tags.

As far as PIDs go, would it be worth editing the PID subroutine to be a periodic subroutine?

Not to bump an old thread but something to keep in mind.

A system I am working on right now also still has to communicate with another PLC 5 rack after the conversion. I found it much easier to scale everything in the controllogix to the 0-4095 used in the old code so that makes everything line up when being transmitted to the other rack. This allowed a lot of things to fall in line with regards to the inputs and outputs of the pid instructions and marrying them up to registers slated to be transmitted out to other plc5 racks.

I also found it much easier to have aliased IO tags that "look" like the old program tag structure. This allows personnel on site that may have experience looking at the old code to recognize the new code after conversion. A few more hoops to jump through that makes it easier for the customer after I leave site.

These aliased tags also match up with any old schematics that may still have the old addressing on them.

Just my two cents.

I'll rebump this thread.


I'm working on PLC5 to ControlLogix conversion as well. Does anyone know if the PID tuning constants can be brought straight across from PLC5 to Logix - ie are the algorithms the same or similar enough under the hood that we can use the same tuning constants and expect similar results? We're using the basic ladder based PID instruction in the new code (not the FBD PIDE), and I'll confirm scan times/execution rates are same. It sounds like maybe due to the -4096 to 4096 scaling stuff, there might need to be some scaling factor applied to the constants when bringing them across?


Anyway, if anyone has had any luck with this please let me know...or if folks have tried and it didn't work and they just re-tuned, I'd appreciate any info like that as well. We're on a tight shutdown schedule, there is time for tuning built in, but we have some jacketed tank temp. loops which can take forever to tune, so even if we have something that gets us some good starting values for the constants based on the old PLC5 ones, that would be a huge help.