LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Real-Time -> Function-Blocks -> Control -> PID vs. Control design & Simulation -> PID

Hi,

 

I'm developing a RT-Application at cRIO-9074 target (using LV 2012). I'd like to control a simple engine throgh PID-controller.

 

Now I found two possible PID vi's:

 

1. Real-Time -> Function-Blocks -> Control -> PID

PID1.png

2. Control design & Simulation -> PID

PID2.png

 

What is the difference between both?

 

Is it possible, that the first PID-Controller might cause problems when it is running a long time (longer than 12 hours)? I placed the sub-VI with the controller into a timed-loop. After more than 12 hours runtime, the timed-loop didn't excecute. The other loops (not timed) for TCP-IP to the Front-panel did work fine...

 

I don't know if my problem is caused by the PID.

 

The other possible reason might be the timing source. Are there any known problems with timed loops and sync to scan engine as timing source? Is there maybe a reinitialisation of the scan engine, that causes a problem?

 

My long-term test was with compiled startup-application, so I can't state more Debug info.

 

I'm looking forward to your answers

 

Fabian

0 Kudos
Message 1 of 4
(3,750 Views)

Hi FJohn,

 

I have looked around. Seems there is no difference between them (I could be wrong). Have you installed PID driver in your cRIO? There should be an option to install PID in NI RIO drivers. 

 

Have you tried the second PID? Does it has the same effect as the first? 

 

Something for you to look at as well:
https://decibel.ni.com/content/docs/DOC-4134

 

Warmest regards,

Lennard.C

Learning new things everyday...
0 Kudos
Message 2 of 4
(3,709 Views)

Hi Fabian,

 

could you please post the VI which you use for your test?

0 Kudos
Message 3 of 4
(3,695 Views)

Hello,

 

thanks for your answer.

 

First time I tried the first PID.

Now I changed to the second. After this I needed to install PID & Fuzzy for RT-Systems (the first controller run without the Toolkit, that is one difference).

 

In my timed loops I changed the timing source to the 1kHz internal clock (before I had the NI-Scan-Engine as timing source).

The second change I made at the timed loops was to give them different priorities (100 and 101, before was 100 and 100).

 

I reapeted my long-term-test and it worked for 21hours (until now) without any error.

 

For now everything seems working. If not I´ll post my timed loop sources (don´t have them now at my local PC...).

 

Thank you for your suggestion with different priorities

 

Fabian

0 Kudos
Message 4 of 4
(3,689 Views)