Logix 5000 PLCs and Wonderware Intouch question

What type of "controller tag" do you use with wonderware's DA server and the Logix 5000processor?
We like the "User Defined" data type but we are worried about having the tags stacked in a nice
consecutive order. Would a consecutive array of Integers be a better choice or doesn't it matter.
When using the "User defined" data type the tags can have a nicer defined name to use in wonderware.

I use all of the above...I don't limit my tag development to WW. WW will support both arrays (eg MyArrayTag[0]), UDTs (eg MyTag.Pump) or a combo of both (MyArrayTag[0].Pump). The older I/O servers I think were much more constrained to having consecutive memory location allocation (think Data tables in PLC5 and SLC) in regards to performance, but I've never seen an issue with this using the DAServer.

Now with OPC servers and their tag name capabilities, try not to limit your creativity to satisfy a HMI or SCADA system. If they don't support native tags in ANY form then perhaps that is not the right driver. UDTs are nice for the PLC and the organization of data, I say that whatever you are most familiar with will most likely be the best fit for you. Data is Data, it really does not matter how you get it, unless your requirements are for fast or "real time" data.

bit

As a standard, we map all our HMI data to dedicated arrays ... one for Integers (and binary data), and one for Reals.

Within our controls, we uses AOI's and UDT's all the time ... then for the data going to the HMI, map it to the HMI dedicated arrays.

Just curious. Have you seen some large Logix systems deployed with wonderware using UDTags. I have one with several AB700 VFDS, 3 logix processors and 6 Wonderware computers on the same network with managed switches. The network seems to distribute data just fine. I am thinking the same as you are that with managed switches and Ethernet it doesn't matter anymore if the tags are grouped tight together.

Managed switches don't have anything to do with how data is arranged in the PLC or the WW PC. Switches just pass data packets from one port to the other, they really have no idea what the data means or how it will be used. All a switch is concerned with is where the packet should go.

If you group your tags in arrays in the ControlLogix, you will save some processing overhead for both the ControlLogix and the WW PC driver. That is where packets are formed and redistributed.

With more processing power available in ControlLogix and PCs, this is less of a concern.

Older systems like the PLC5 with RSLinx used a very simple method of forming packets. If all your tags were in file N7, it would simply make a packet for N7:0 to 99, and another for N7:100-199, and so on. If you put all your tags in N7:0 - 99, you got 1 packet. If you had one tag each at N7:13, N7:133, N7:249, then you got 3 packets. So grouping definitely made fewer packets.

The ControlLogix is much more complicated than that and makes packets as needed. It packs different tags into the same packet, this reduces the number of packets, but the more tags it has to keep track of in each packet, the more work it does. It is much easier to read/write a single dint array of 120, than to read/write 120 separate dint tags when sending or receiving a packet.

However, it does a pretty good job of it, and as long as you aren't having issues, I wouldn't worry too much about it. If you are planning to move massive amounts of data, then making it as easy as possible for the hardware may be a good thing.