Newbie PID control question

[Guest WayOutWest]

Can anybody point me in the right direction regarding doing PID control with an AB Contrologix PLC. Never used the PID function before. Thanks, Newbie

[Guest]

sure, we'd be glad to try to help ... but PID is a pretty big subject ... can you help us out by telling us what you're trying to control? ... as many details as possible would be nice ... that way we won't have to keep asking you through numerous posts just to get a handle on things ... one big issue: we need to know how your inputs and outputs are set up (the specific hardware model numbers and the actual tagnames would be helpful) ... and how much of the project have you done already? ... specifically, can you go to the analog output tag and manually enter a value - and thus manually control the system? ... and if so, does the analog input tag change to reflect the manual control? ... major point: if you can't already control it manually - and monitor the results - then you're not really ready to set up a PID for automatic control ... also ... are there any safety aspects to this project that we need to know about? ... and what is your deadline? ... "As Soon As Possible" is not a valid answer ... we look forward to hearing from you again soon ... and finally ... don't forget to register ... best regards, Ron

[KinK 0]

dang I think that I am going to copy and paste Rons Answer it seem that that could be used as a template for answering a lot of questions.....

[Guest WayOutWest]

Thanks for the help and guidelines. I can control the process in manual, it's a very slow changing process, typically set it for a few hours and adjust every now and then. Hardware: 1756-L55 ControlLogix5555 Controller 1756-IF16 1756-OF8 Two forms of control I need to implement: Temp control TI-011 Loop Temperature 4-20mA Input TCV-411 Control Valve 4-20mA Output Differential Pressure Control (this one is trickier because it requires two pumps to operate when the first pump cannot control the pressure. I've basically set it up so that if the first pump is greater than 90% speed for 1 minute and the pressure requirement has not been met the second pump comes on line then the third and fourth. Not sure what to do with the output command from multiple PID blocks) PDIT-41 Loop Differential Pressure 4-20mA Input P01-CMD Speed Command 4-20mA Output P02-CMD Speed Command 4-20mA Output P03-CMD Speed Command 4-20mA Output P04-CMD Speed Command 4-20mA Output I've done all the scaling and interlock programming and I'm down to basically making the PID control block work but I'm not familiar with using that block. I understand the definitions of process variable, set point and control variable but I don't know the effects of the other parameters and I'm under some time constraints, I need to wrap this up tommorow or Friday at the latest. Unfortunately I don't have internet access on site and can only check in before or after work. Any help would be greatly appreciated. Thanks.

[KinK 0]

Ok sorry to do this to you but this will give you more information then I ever could and probably better explained Pid tutorial Peace and good luck

[Guest]

sorry West, but I don’t think we’re going to be able to meet your deadline ... tomorrow IS Friday where I’m located ... and if you can only check in before and after work then that only gives us about two chances to communicate ... I don’t believe we can make this happen ... basically you have two big areas to tackle ... first ... you have to set up and configure the PID blocks ... we MIGHT be able to get that done ... but I doubt it if you’ve never done it before and if we can’t communicate in something a lot closer to real-time ... second ... you’re going to have to tune the PID’s for acceptable response ... again, if you’ve never done it before, this is going to be something like trying to teach you how to drive a clutch-pedal and stick-shift car by communicating through the mail ... even if I were there with you, I’d set aside at least a day or two for this part ... I don’t know what your alternatives are but I’d strongly suggest that you start exploring them ... but if you’re truly grasping at straws, then MAYBE this will help ... (but I honestly doubt it) ... anyway ... based on what you’ve posted ... this might give you a hint on how to set up the pump PID controls ... this assumes (gosh, I hate that word) that you’re going to be able to run all of the pumps (at least all of the ones that are turned on) at the same speed ...

[Guest]

again, based on what you’ve posted ... this might give you some hints on how to configure the pump PID ... note that the Loop Update Time needs to match the 0.500 seconds of the TON that triggers the PID ...

[Guest]

again, based on what you’ve posted ... this might give you some hints on how to set up the scaling for your pump PID ... this assumes (gosh, I hate that word) that you’ve set up your analog input module to scale the 4-20 mA input signal to read out as 0 to 150 psi ... and that the analog output module is set up to take a 0 to 100% value and put out a 4-20 mA signal ... naturally you’ll have to adjust this screen if your scaling is different ...

[Guest]

again, based on what you’ve posted ... this might give you some hints on how to tune the pump PID ... NOTICE !!! the Tuning Constant numbers entered here are just WILD GUESSES ... but maybe they’ll help get you started ... hint: I strongly suggest that you leave the Derivative Rate entry at 0.0 until you’ve gained some more experience ... you’ll need to IN-crease the Proportional Gain entry to get more Proportional action ... that makes sense ... BUT ... with the configuration (Dependent PID equation) that I’ve recommended ... you’ll need to DE-crease the Reset Time entry to get more Integral action ... that seems backwards to most people ... I’ve suggested the Dependent PID equation since that’s the way most books do it ... there’s no sense in adding in an extra layer of complication at this point ... I wish I could be more help ... but your deadline and communications limitations are a killer ... good luck ... Ron

[Chris Elston 78]

dang Ron....I need to start awarding metals of honor for excellent explainations.... Another "pat on the back!"

[Guest WayOutWest]

Ron, Thank you very much for your help. I didn't realize it was already Thursday when I posted. It is really ironic you mentioned learning to drive stick over the mail/phone. When I was 16 I got stuck out in San Diego with a drunk friend who couldn't drive us home, I'm from L.A. He had a stick shift Nissan truck and everyone but me was way too drunk to drive. I didn't drink at the time so I just figured it out on the way to the freeway on-ramp. Same thing happend today. Downloaded some document titled "You don't need a PhD to do PID". Figure that was right up my alley. Make a long story short, figure out how to implement PID control for the various loops on the system. Tunning the loop took longer than figuring out how to use a PID block. Each loop took about 30 to 45 minutes to tune. Still having trouble with on loop because the temp control valve is more like a butterfly valve, I get very little response until I get within the range of 40 to 60 percent, then the temp drops like a rock. Need to figure out a way around that one. Again thank you for the help. Other than PID control I'm pretty familiar with most AB products so I'll try and contribute here in the future. Thanks, Mike

[Guest]

Greetings Mike, congratulations ... and I'm totally impressed ... but I can't help asking: when you first signed on as "Newbie" you meant "new to this forum" and not "new to PLC stuff" right? ... I don't know many (any?) people who could have pulled this one off the first time out of the gate ... thanks for letting us know how it turned out ... and please drop in again ... best regards, Ron PS ... how about a link to that "You don't need a PhD to do PID" document? ... that sounds like a good one ... I tried Google but didn't get any hits on that title ...

[Peter Nachtwey 15]

Here is the article http://www.embedded.com/2000/0010/0010feat3.htm Tim Wescott often posts on the news groups like sci.engr.control. He also has a web page for his articles.

[Guest]

thanks, Peter ... Ron

[Macgyver BR 0]

Very good, tutor. Peter and Ron, as everybody appreciate teacher and learner .That such fear him "PID for internal control" Always, applies PID for external control. I know that can be utilized, in the internal control of the software, estatics, anticipate values, can simplify or replace calculs Fair, I know that is possible, is little exploited is option Ps: It new about, open new post i see.

[Guest WayOutWest]

Ron, I'm a newbie to the forum. I've done PLC work on and off for a years but I'm not a process guy. I typically work from a spec. I do alot of the HMI as well, some stuff off the shelf some stuff custom VB. Lots of client server apps to make sense of tons of logged data. I see someone already posted the PID/PhD link. I don't have an internet connection on site so I can't check in very often. Thanks again, Mike