Hi all,

I am deep in trouble with the OPC interface of RS Linx and Iconics SCADA as OPC Client.

Iconics V7.x is running on an PIV 2.4 GHz-256 MB Ram HP/Compaq PC with the XP Proffesional installed. Installed Software is Iconics V7, RS Linx 2.41, RS Logix V6.
Communicating Hardware is 1784-PKTX. Single Channel. PCI Card. PLC Network: DH485, 11 Nodes, 19200 Baud.
Main screen has got appr. 120-130 Points polled from all 11 nodes. When an additional client adds on and polls for additional data communication drops step by step to all nodes. This additional client could also be the RSWho window in RS Linx. Here I can follow how the nodes go offline.

I have installed all patches and drives I could found in the Rockwell knowledge base and everything received from the Local Rockwell service. Is there anybody who can tell me what the problem could be? We are using OPC more and more as we believed in that technology. Iconics and Rockwell are OPC Foundation members and developer. How come this incompatibilty?

 

Anonymous---

Good to hear from you again!

You don't say how often you are polling for the data points you mention, but I suspect you are challenging the bandwidth of DH485. If you ask for data more oftn than it can be provided, you will get a lot of failures, which you should be able to track via the diagnostics in RSLinx.

You really need to discuss this with the tech support folks at Rockwell. Call them at 440-646-5800.

Hope this helps!

Larry Lawver
Rexel / Central Florida

 

Are you polling 120-130 points total from all nodes, or that many from each node? How fast are you polling that data?

It doesn't sound like an OPC problem at all, but rather that you're simply running out of bandwidth on a DH485 network.

Even in the simplest analysis, 19200 baud shared by 11 nodes means roughly 1700 bits per second available for each node to transmit. I think you're simply asking for too much data for the network to handle without experiencing slowdowns that are so severe that the software applications relying on the network eventually time out.

What are the devices on this network? Maybe you could get things running on a higher-speed network like Ethernet or DeviceNet.

 

It looks like the OPC Server is not agregating the requests from more than one OPC client. This itself should not cause any other problem than a decreasing performance. Please contact be Rockwell to discuss the matter.

Until the opc server is improved you could try to lead all communications through ICONICS DataWorX32. DWX32 will agregate all OPC connections from all connected OPC Clients and make a single connection to the destination OPC Server. You do not have to configure DWX32 itself, you will only have to instruct the OPC clients to utilize DataWorX32

Example:

imagine you currently have a tag called
yourOPCServer\yourTag
to use the power of DataWorX32 you change your tag to
ICONICS.DataWorX32\yourOPCServer\yourTag

if DataWorX and your OPC Server resides on a networked computer you would use
\\pcName\ICONICS.DataWorX32\yourOPCServer\yourTag

Regards,
Raymond van der Tas

 

hi,

Sorry for the information leak. I am polling appr. 10 Words/node. I have divided the 11 nodes in two groups and set the polling rate 7 seconds and 9 seconds for the two groups. That's what I have done on RS Linx side. On Iconics side I have set each page to 15 seconds. Meanwhile I have upgraded to RS Linx 2.41 and upgraded from OEM to Pro due to some suggestions that OEM would support only "one way" comm. No idea why and what that is. We are talking about a "server" which is "serves" to CLIENTS and not one client. But nothing changed.