Rockwell Automation T8030S Trusted OPC Server Package - Manual (Page 22)

Trusted
TM
OPC T8030S
Issue 23 April 10 PD-T8030S 22
The Comms ID number specifies the node address to use when communicating with the controller. It
must match the slave ID responded to by the Modbus Slave set up in the controller’s Communication
Interface modules.
The final check-box ‘Write to all IP addresses’ is included for completeness and compatibility. On dual
systems, data is read from only one of the two controller halves. However, data must be written to both
halves simultaneously. This does not apply to Trusted
TM Controllers. This check-box should be
unchecked.
Having added/deleted or changed the Controller settings, the configuration must be downloaded
to the server and the server must be restarted. To save the settings, click the disk icon (shown
left) or select File | Save. On saving a change, the config tool reminds the user that changes to
settings will not take effect until the server is restarted. Refer to section 2.3 above for the methods
available to restart the server. Note that some clients will then also need to be restarted in order to be
updated.
2.4.5. Server Timing Settings
The best settings for the OPC poll rates (Global Ungrouped Tag Update, Process Historian Update
and Sequence of Events Update), are entirely dependent on the configuration of the system. They
should be set to allow sufficient time for all communications to occur.
The OPC Server is capable of performing approximately 40 messages per second. This is the total
number of messages to all controllers, so 1 controller would be addressed 40 times, 4 controllers
would be addressed 10 times.
A single SOE or Process Historian poll takes four separate messages, so at best the OPC Server can
do 10 SOE or Process Historian polls per second.
The following information is required to determine the best poll rates.
• The number of controllers that the OPC Server connects to.
• The number of Boolean and analogue points per controller that are read directly (no SOE or
Process Historian), their communications addresses, and the update rate required. Calculate how
many messages are required to read the full map for all controllers (defined as N
O
below). A single
message can read approximately 900 Booleans or 125 16-bit analogues, provided the addresses
are contiguous.
• The number of Boolean and analogue points per controller that are polled via SOE or Process
Historian, and how often they change state. Each SOE or Process Historian poll can read a
maximum of 12 changes. If there are a large number of changes to be read regularly, the SOE or
Process Historian data would need to be read more frequently to prevent a build up of changes in
the controller. Calculate how often the event data must be polled.
Given the number of ordinary messages N
O
and the required ungrouped tag update rate for this non-
event data, calculate the messages needed per second required to keep up with the desired rate.
P
O1
= N
O
/ Ungrouped Tag Update Rate (in seconds)
Since the OPC Server can perform approximately 40 messages per second, and an event poll takes 4
messages, the messages per second remaining for event driven data are:
P
E1
= (40 – P
O1
) / 4
If C
S
is the number of controllers needing SOE polls, C
PH
is the number of controllers needing Process
Historian polls, the SOE and Process Historian minimum poll rate in milliseconds (to ensure all event
buffers get at least one poll) is:
1000 x (C
S
+ C
PH
)
P
E1
Some SCADA OPC clients can be configured to only request the points that are currently on screen.
This reduces the number of ordinary messages required, but is entirely dependent on the configuration
of the graphics layout.