User`s manual

CR-SEB
When a master CR-SEB is powered on, it becomes active immediately, even if no slave CR-SEBs are
detected. It will attempt to send packets addressed to devices that it thinks are not on its local network.
When a slave CR-SEB is powered on, it listens for a master CR-SEB and attempts to register with the
radio in the master. This detection and registration process typically takes 2 seconds. During this time, no
packets will be sent or received over the RF link by the slave.
If the application is such that more than one point-to-point link needs to located in the same area, each
master/slave pair must be assigned different network numbers. This will allow the CR-SEBs to identify
the appropriate other CR-SEB to which they should communicate. Because different network numbers
have different hopping sequences, this also allows various pairs to operate in the same area without
interfering with each other. Refer to the section on radio commands for details on setting network
numbers.
Point-to-Multipoint Mode
In point-to-multipoint mode multiple slave CR-SEBs link with a single master CR-SEB. Similar to the
point-to-point mode, each slave must register with the master, a process that typically takes up to 2
seconds. All of the radio addressing and registration occurs automatically and is transparent to the
application. A maximum of 15 slave CR-SEBs can be connected to a master CR-SEB. The amount of
data a slave can transmit depends on the number of slaves connected and the hop duration of the radio
network. Because packets from one slave that are intended for another slave must pass through the master
CR-SEB, the master CR-SEB should always be allocated 50% of the data throughput, even at the expense
of some throughput for the slaves. This is the factory default configuration.
In frequency hopping technology, the following applies… the shorter the hop duration, the lower the data
latency but also the lower the throughput. This is because the overhead required is the same regardless of
the hop duration. Thus at shorter hop durations, the overhead is a larger percentage of the hop time.
Longer hop durations provide more throughput but have a higher data latency. If data from a slave
appears just after the slave’s designated transmit time, the slave will have to wait one hop duration before
it can transmit the data. As the hop durations of the CR-SEB products are very short, ranging from 10ms
to 17ms, the better approach is to pick the hop duration to maximize data throughput without
unnecessarily penalizing latency.
The factory configuration is set up that the master will have one-half of the bandwidth reserved for it. The
slave devices split the remaining time equally and transmit as much as the can each hop. The amount of
time remaining will depend on the hop duration. The table below gives data throughputs for multiple
slaves based on the default hop duration for a CR-SEB.
Slaves Hop Duration Slave Throughput
Master
Throughput
Aggregate
Throughput
1 10ms 208Kbps 208Kbps
416Kbps
2 10ms 118.6Kbps 208Kbps 445.2Kbps
3 10ms 69.5Kbps 208Kbps 416.5Kbps
4 10ms 44.8Kbps 208Kbps 387.2Kbps
If the application has more slave devices, or if it is desirable to increase the throughputs of the slaves at
the expense of the master CR-SEB, contact Cirronet technical support for more details.
Co-located Networks
The radio architecture in the CR-SEB allows multiple networks to be located in the same area without
interfering with each other. Depending on the scenario, either the network and/or the sync commands are
 2000- 2004 Cirronet Inc 16 M-2410-0025 Rev A