PID setup

Hi All, temps on out heater jumping around a lot 20 some a degrees up and down and it should stay within 10-degree window in production mode. I pulled up PID and found output %100 and it needs to fluctuate. Any suggestions?? Rslogix 5000. Thsnks

for starters your picture did not come out try resizing it.
if the PID output is at 100% then it looks like the heaters aren't large enough to do the job
it sound like the you need to go back to the heat engineers and have them recalculate the rquirmnts

First of all, was this working correctly before or is it a new installation.
The chances are that if it did work before and it is running at 100% you probably need to look at the control cct. it will depend on what the loop is controlling i.e. steam modulating valve is there a heat exchanger, what type of medium is it heating, etc.
The most common faults on heating systems are heat exchanger blocked, valve positioner not set up correctly, low steam pressure the list goes on.

I swapped input and output cards with working heaters and this made it to change and now seem like PID tuning needs to be done due to large heat fluctuations. I believe it started acting up two weeks ago and no one reported any chabges recenly that were made to it. Thanks

You stated that the output for the PID was stuck at 100%, then if you suspect the PID parameters I suggest unless someone has changed them this is not the problem. Did you replace the analog output card ? this might need setting up if you do try to alter the PID parameters then make a copy of them first.

I have alread tried swapping ouu the isolated analog output module with one next to it that's known to work. I don't believe anyone made any changes within PID constants. Usually changing these constants performing PID tuning does not make heaters to fluctuate as much. The temps fluctuate like crazy even when machine in a stopped or paused state. I would have to look more into it. I will look further into share my findings. I open to any suggestions as well.

Don't overlook the possibility of noise on the analog input signal. Are the fluctuations in temperature believable? By that I mean is it possible that the PID output swinging between zero and 100% could cause the temperature change reported by the analog input channel?
Try putting the SP, CV, and PV signals on a trend chart and observing them over time.

Is the PID control in the main task ?
in the Logix 5000 the PID controls must be in parodic task recommended 10ms or longer
the main task runs to fast for it to work the output will go full scale either full on 100% or full off 0% with nothing in between
I have run into that myself before

GaryS said:

Is the PID control in the main task ?
in the Logix 5000 the PID controls must be in parodic task recommended 10ms or longer
the main task runs to fast for it to work the output will go full scale either full on 100% or full off 0% with nothing in between
I have run into that myself before

Click to expand...

Also, the PID instruction has a parameter "xxx.IPD" that needs to be set to the same time period (in mS) as the PID instruction gets executed.

The PID instruction is a calculation based on time elapsed since the last calculation, so it needs to be executed at a fixed time interval, and it needs to be told what that interval is.

Some people do this with a self-resetting timer in the continuous task, some people do it by putting the PID instructions into a Periodic Task.



EDIT : In my experience, most PIDs, unless they need to be fast acting, dont need to be executed at a rate faster than 100's of mS, and many that I have successfully programmed and tuned 1 or 2 seconds interval. Also, if you are controlling VFDs, remove any "ramping" from the drives, this interferes with the feedback parameters you get.


I prefer to put PID instructions into periodic tasks, so that I can programmatically read the Periodic Task execution interval, and programmatically load those same values into the corresponding "xxxx.UPD" parameters. That way, if someone changes the periodic task execution rate, the PID instructions will automatically compensate. Also periodic task execution interval is more deterministic than a self resetting timer. On large programs your "interval" will not be totally constant.

So long as each PID instruction is told the rate at which it is being executed, retuning is rarely necessary if that rate changes.

You only need the actual PID instructions in the Periodic Task(s), there is no need to put any of the "control" logic, such as auto/manual controls, or bunpless transfer data swaps.

I have done many jobs with many PIDs in (30+), getting them set up so they know their execution rate will mean you rarely have to revisit them, and if they are not controlling as well as they used to, suspect the external hardware first, I/O seoond, and code last.

Also remember you do not need to do any Analog Input/Output module configuration if you swap them out. the controller downloads the "C" (configuration) tag when it makes its "connection" to the module. This makes I/O modules essentially "plug and play".

The high shut off limit in logic was set to 800 degrees
same as set point in PID. Once this limit was set to 850 degrees. The temps working normal now.