PIDE Block Not outputing CV

[jhunter1993]

Hello,

Long time listener to these pages first time caller.

I have a facility with 2 of my company's units there. They have been running for about 4 years. They are compressors so there are 2 PID's one that controls a recycle valve to control discharge pressure/ flow and another one that controls motor speed.

Everything has been working well for about 4 years then all the sudden one of the units started acting weird (the site says) and now the other one. The speed control is suppost to start the units an minimum speed which would be 4mA output or 950 rpm and has more gas flow is needed the motor speeds up. 1 of the units motors runs at 1200rpm and the other at 1800rpm all the time now, come to find it its because the PIDE isn't outputing a CV and the VFD's has a setpoint to bring the motors to a base speed if signal was lost (they just happened to be different).

Anyways, I cannot figure out why the PID isnt sending out a CV I checked and its status is OK and I have no errors. When the unit is running it is calculating the correct E it just doesnt do anything about it. The other PIDE in the program is having no issues and seems to be set up pretty much the exact same way (a few differences obviosly).

I've called some old buddies and we couldn't come up with anything, especially since it worked for almost 4 years.

I attached the L5k file and a picture of the block while the unit was running. Any ideas will be greatly apprieciated.

 

[Steve Bailey]

In the screen capture your SP is 375 and your PV is 390. Zero seems a reasonable number for the CV under those conditions.
How does the CV get from the PLC to the VFD, hardwired analog signal or over a network connection? If hardwired, does "Loss of signal" mean a 0 mA or 4mA signal level?
In situations like this the first question you should ask is, "what's different now from the last time it worked"? Chance are the program hasn't changed.

 

[jhunter1993]

Hey Steve,

Just some background as well I'm just a field service "engineer". And the guy who build alot of my company's is gone and I've just learned how to deal with how he did things, I've learned that he did things strangely but they've always worked pretty well.

So for the other PIDE in the program if it has a difference of .5 or more the CV changes either way it needs to. At an error of 15 its doing nothing.

The signal to the VFD is via a hardwired 4-20mA. The VFD's backup parameter is for when a signal less than 2mA is seens but un this case it is seeing 0mA because the PID isn't outputing a CV.

I've been asking myself that, and nothing that me or the site can come up with. I'm the only person that goes online with these units (I went directly online with it this morning with the file I had from last time) so nobody has changed it and it didn't default an older file in anyway.

But I can't find anything that would disable the PIDE from outputting. It's enabled and has an OK status. I've had issues in the past were errors occur or it gets disabled because of some weird reason.

 

[drbitboy]

So for the other PIDE in the program if it has a difference of .5 or more the CV changes either way it needs to. At an error of 15 its doing nothing.

The error has a magnitude of 15, but its value is -15, which means the PV is higher than SP (E=SP-PV; Control_Action is 0 i.e. False).

Since E is negative, PID will seek to decrease the CV, initially via Proportional Action, and then decrementing CV more over time via Integral Action (assuming the Integral constant is not 0), and it has done so down to the CV lower limit of 0, as @Steve Bailey has written.

Presumably that CV of 0 is scaled to some value (INT?) so the output card will output a current of 4mA; you should be able to check that online, if you know what type of card is being used.

Assuming it is sending the correct 4mA-equivalent INT to the output channel, the question I would ask next is why 4ma of current is not "flowing" in the cabling between the PLC's output channel terminals and the VFD terminals. Broken wire? Bent pin? Short? Bad channel?

Whatever it is, it seems to me it is most likely an electrical issue, because that PID, and probably the rest of the PLC software, are doing exactly what they have been told to do.

 

[Steve Bailey]

I would have expected the analog output channel to have a value of 4 mA for commanded value of zero. That's one of the key advantages of 4 - 20 mA over 0 - 10 volts, a value less than 4 mA indicates a broken wire or a failed channel. Have you tried using a different channel on the analog output module? Does the 4 - 20 mA signal for both of the VFDs that aren't working correctly come from the same analog ouput module? Presumably the analog output module requires an external 24 VDC supply. Have you checked that?

 

[drbitboy]

The speed control is supposed to start the units an minimum speed which would be 4mA output or 950 rpm and as more gas flow is needed the motor speeds up.

1 of the units motors runs at 1200rpm [i.e. > 950rpm] and the other at 1800rpm [also > 950rpm] aall the time now, come to find it its because the PIDE isn't outputting a CV and the VFDs have a setpoint to bring the motors to a base [backup] speed if signal was lost (they just happened to be different).

The signal to the VFD is via a hardwired 4-20mA. The VFD's backup [base] parameter is for when a signal less than 2mA is seen but in this case it is seeing 0mA because the PID isn't outputting a CV.

So the VFD backup logic detects an analog electrical signal that is below 2mA, which detection means that electrical signal is lost, and the VFD response to the lost signal is to run the motors at the base [backup] speeds of 1200rpm and 1800rpm.

And at those base speeds, both of which are over the 4mA speed startup speed of 950rpm, one of the compressor systems is supporting a pressure (PID PV) of 390+, which is 15+ over the PID SP of 375.

If the PLC output channel were sending a 4mA electrical signal (because software PID CV was making it all the way to the electrical signal into the VFD), the speed would be 950rpm, and the pressure would be lower than it is at the higher base speed at which it is now running.

What happens if the backup [base] speeds are dropped by, say 25rpm, to 1175rpm and 1775rpm? Does the pressure (PID PV) decrease below the current 390+?

But I can't find anything that would disable the PIDE from outputting. It's enabled and has an OK status. I've had issues in the past were errors occur or it gets disabled because of some weird reason.

We can ignore the PIDE because it is doing exactly what it exactly what it has been told to do i.e. in this case drive its CV output to its minimum to lower the speed of the compressor.

A feedback loop only works if the loop is not broken (PV => PID => CV =>^broken? Process => PV). The problem is that the PIDE CV output is not being translated to a 4-20mA signal delivered to the VFD; that problem is happening somewhere after the PIDE instruction.