PV+ Data arrays, question again

Hi all. I searched and was unable to find what I was looking for. We are discussing and speccing a large retrofit, and talking about PV+ screen updates. On one hand we have a guy who is suggesting improved performance by locating data read by each screen in DINT registers that are addressed directly. On the other hand, we have a guy who says we should just direct address the objects to whatever tag we desire in CLX. His arguement is that the array formation requires additional developement time, and that because the direct addressed tags are only called when the screen is active, that this extra work of creating the array is not going to yield significantly better response. What's the general response here?

Russ

because the direct addressed tags are only called when the screen is active

Click to expand...

As long as we're talking about an EtherNet/IP or ControlNet network, the pushbutton and indicator "response time" is going to be roughly equivalent for individual tags as it is for array elements.

The big difference in performance that I see is when you switch between active displays on the PanelView Plus. If each display has a set of different individual tags, the tags have to be "brought on scan", which requires RSLinx Enterprise and the ControlLogix to negotiate an optimized data block. This can take a noticeable amount of time when you change screens.

When you are using optimized arrays, the PV+ and Logix don't have to negotiate that optimized data block, either because the tags on scan don't change, or because the array and data block are very simple to correlate.

I see the merit of the development argument, but from a troubleshooting standpoint I always prefer to have my operator interface data tables (both for reading and writing) segregated from the operation variables in my control code. It takes some time during development, but pays back dividends during startup and in troubleshooting over the life of the system.