Automation Direct PID Loops

[dash]

OK, this is the first time I have had a need to use a PID loop with an Automation Direct PLC.


Hardware : D4-450 with updated firmware V3.3 (started with V1.905 main CPU and Gate Array V2.463, date code 0298)

Software: DirectSoft 32, latest version, no known updates

System : Foam injection press for making lost foam patterns. It is a vertical parting line mold cavity with the platten moving horizontally. The mold closes and then steam is injected to fuse the foam beads. The steam pressure in the chest (mold cavity) is being attempted to be controlled by the PID loop in the D4-450 PLC processor. The steam pressure is measured by a WIKA 4-20mA loop powered pressure transducer. The control valve is a Shubert&Salzer sliding gate valve with digital positioner.

(See http://www.schubert-salzer.com/index.php?id=139&L=1)



Input module : F4-04ADS (12 bit input resolution)
Output Module: F4-04DAS-1 (16 bit output resolution)

Problem:

Using independent (if you are not familiar to A-D, this does not refer to independent vs. dependent gains, it is referring to the the PV/SP being a different resolution than the output) format with the input set to 12-bit and the output set to 16-bit.

First of all we can not get the auto-tuning with the new software to work at all. The output just sits there like a dead duck and does not funtion at all like the manual states.

Also, had the same result trying to use the velocity form of the PID. The only way I could get the PID to attempt to adjust the CV was to use the positional form of the PID.

So we tried just setting the loop in auto-mode, making a step change in the SP and looking at the result. The really funky thing was that say the set-point was 250 counts, the PID loop would manipulate the CV in such a way that the PV would oscillate around a PV value of 80 counts (very tightly +/- 10 counts).

Does anyone know if this is a known issue with the A-D PID loop using the independent format?

Does anyone have some suggestions on the tuning of an A-D PID loop?

Thanks,

Darren

 

[dash]

Oh, yes before it comes up we have verified the outputs to the valves and the positioner accuracy. Also have verified the inputs from the pressure transducers.

Darren

 

[jprindle]

I struggled a little with a PID loop in an 06 a while ago. I don't remember exactly what the problem was, but you might want to monitor the bits for the loop in the ladder. I think that is where my problem was.

 

[dash]

Yes, we assumed initially that we had programming errors since this was the first time we have used an A-D controller for a PID loop. The nice thing about the new software is that you can go to the PID screen and control modes, gains, etc from there. So we diabled all our ladder logic and just used the PID view to try to adjust the loop.

We then found that we had to update the firmware just to get the new PID screen to operate at all. That took a couple of hours to understand that it just was not working right at all. It at least now looks like it is functional, it just does not seem to be working like we expect.

Darren

 

[dash]

After struggling with the Automation-Direct PID loops it looks like they are functioning now. The major issue turned out to be not realizing how painfully slow the PID graphs update in the new software. I have gotten spoiled by some of the current motion & drives products we normally use that actually buffer the data to aid in loop tuning in the device so that you get much closer to realtime on the graphs. The DirectSoft32 software does no such thing.

We finally resulted to some swags based on the steam pressure and the valve response info from the manufacturer and then tweaked it by watching an analog pressure gauge that is mounted in line with the mold.

Seems to be working so far. Yesterday a valve on the vacuum line failed so the mold would not pressurize properly, but it ran for around 6 hours with no problem prior to the failure. Hopefully after the actuator is replaced it will go back to running like a top.

If you are adding PID loops to an older D4-450 I would strongly recommend updating the firmware before you end up with much frustration. Also if you are having problems with your serial connection dropping to the PLC sporadically, according to the tech notes this is improved with the later firmware. It seems to have improved the situation with our PLC. Whenever we would e-stop the equipment or reset the e-stop the serial connection would fail to our programming laptop. We remove power from the analog output modules under e-stop condition so I do not know if this was causing the PLC heartburn or what.

Darren