Bosch PRAESENSA, Руководство по установке

Страница: 50 / 194

50 en | From installation to configuration PRAESENSA
2019.11 | V1.00 | Installation manual Bosch Security Systems B.V.
PRAESENSA networked devices support RSTP. Connecting and disconnecting redundant loops
take some time to be processed in the network and as a result active audio connections will
be muted during that time.
The system controller is still the central point, it is the so‑called root bridge. Amplifiers can be
looped through and a ring of amplifiers can then be connected between two ports of the
system controller. In the same way a number of multifunction power supplies can be looped
through and a ring of these devices can be connected between the two other ports. Call
stations can be connected to a multifunction power supply via a double connection to two
PoE ports, or even to two different multifunction power supplies. This is the recommended
way of connecting PRAESENSA devices and mandatory for PA/VA systems that need to be
compliant to voice alarm standards.
6.3.4 Hop count
When data passes through (PRAESENSA) network devices between source and destination, a
hop occurs for every device that is passed. In PRAESENSA the relevant portion for hop
counting is the network path between the root (the system controller) and any possible end
point via the shortest path.
This is important because there is maximum to the number of hops that are allowed for proper
communication. This has to do with the latency that occurs in every hop and every connection.
There is a limit of 22 hops. Devices that are connected after the 22
nd hop are lost in the
system. Also loops, connected to the system controller, should not contain more than 22
devices. A loop with 43 devices, connected to the system controller, works fine as long as the
loop is not interrupted, because the device in the middle of the loop has a hop count of 22
towards the system controller in both directions. All other devices have a smaller hop count.
But if one link in the loop fails, this results in two branches being connected to the system
controller, and one of them will have more than 22 looped‑through devices. Then the devices
after the 22
nd
will be lost. So, always consider the worst case hop count for a device in case a
connection would fail. This must be carefully analyzed for larger systems.
The performance of a network will be better if the number of loops is smaller. The RSTP
recovery time of the network after a link failure increases when the number of loops is higher.
So, the hop count must be balanced against the number of loops.