Siemens SIMATIC S7-200 System Manual page 251

For some applications, however, reducing
the number of masters on the network is
not an option. When there are several
masters, you must manage the token
rotation time and ensure that the network
does not exceed the target token rotation
time. (The token rotation time is the amount
of time that elapses from when a master
passes the token until that master receives
the token again.)
If the time required for the token to return to the master is greater than a target token rotation time,
then the master is not allowed to issue a request. The master can issue a request only when the
actual token rotation time is less than the target token rotation time.
The highest station address (HSA) and the baud rate settings for the S7-200 determine the target
token rotation time. Table 7-14 lists target rotation times.
For the slower baud rates, such as 9.6 kbaud and 19.2 kbaud, the master waits for the response
to its request before passing the token. Because processing the request/response cycle can take
a relatively long time in terms of the scan time, there is a high probability that every master on the
network can have a request ready to transmit every time it holds the token. The actual token
rotation time would then increase, and some masters might not be able to process any requests.
In some situations, a master might only rarely be allowed to process requests.
For example: Consider a network of 10 masters that transmit 1 byte at 9.6 kbaud that is
configured with an HSA of 15. For this example, each of the masters always has a message ready
to send. As shown in Table 7-14, the target rotation time for this network is 0.613 s. However,
based on the performance data listed in Table 7-12, the actual token rotation time required for this
network is 1.48 s. Because the actual token rotation time is greater than the target token rotation
time, some of the masters will not be allowed to transmit a message until some later rotation of the
token.
You have two basic options for improving a situation where the actual token rotation time is
greater than the target token rotation time:
You can reduce actual token rotation time by reducing the number of masters on your
-
network. Depending on your application, this might not be a feasible solution.
You can increase the target token rotation time by increasing the HSA for all of the master
-
devices on the network.
Increasing the HSA can cause a different problem for your network by affecting the amount of time
that it takes for a S7-200 to switch to master mode and enter the network. If you use a timer to
ensure that the Network Read or Network Write instruction completes its execution within a
specified time, the delay in initializing master mode and adding the S7-200 as a master on the
network can cause the instruction to time out. You can minimize the delay in adding masters by
reducing the Gap Update Factor (GUF) for all masters on the network.
Because of the manner in which requests are posted to and left at the slave for 187.5 kbaud, you
should allow extra time when selecting the target token rotation time. For 187.5 kbaud, the actual
token rotation time should be approximately half of the target token rotation time.
To determine the token rotation time, use the performance data in Table 7-12 to determine the
time required for completing the Network Read and Network Write operations. To calculate the
time required for HMI devices (such as the TD 200), use the performance data for transferring
16 bytes. Calculate the token rotation time by adding the time for each device on the network.
Adding all of the times together describes a worst-case scenario where all devices want to
process a request during the same token rotation. This defines the maximum token rotation time
required for the network.
Communicating over a Network
Table 7-14 HSA and Target Token Rotation
Time
HSA 9.6 kbaud
HSA=15 0.613 s
HSA=31 1.040 s
HSA=63 1.890 s
HSA=126 3.570 s
Chapter 7
19.2 kbaud 187.5 kbaud
0.307 s 31 ms
0.520 s 53 ms
0.950 s 97 ms
1.790 s 183 ms
237