Requirements

• Cascadable system (a lot of TRBs have to be connected)
• Node-to-Node length at least 100m
• Low latency (< 1us)
• Small cable, optical links or short cables
• One network for the trigger distribution and the IPU data
• Flexible for any future extension
• No data loss on the network layer, error detection and correction
• System should recover from any problem

Concept overview

* Overview:

There is also the option to keep the old CTU and build an old-to-new converter (test board for the new TRB).

The stack of this concept consists of 3 simple layers:

• The physical/link layer: Physical transportation of the serial bitstream, serdes chips, clock recovery: NewTriggerBusMedia
• The network layer: How the HUBs interact with each other, handshake protocol, address and data transportation: NewTriggerBusNetwork
• The transport layer: What is inside the data paket, e.g. how trigger tag and code, status (busy) information: NewTriggerBusAPI

Have a look to the NewTriggerBusToDo