Typical ASI components: All components are new except Link 20E Master. 2 Safety Monitors, 6-8 D-I/O modules type 3RK1205 and 3RK1405, 10-13 E-Stops using ASI modules 3SF5500-0BA, 4-8 Pull cable devices model 3SF2 140-1BF00-0BA1.

Topology: The ASI bus contains 25 slave devices and 2 Safety Monitors. Length 150 meters. The Power supply and Link 20E are located in a cabinet near the center of the network. Two main branches of ASI cable leave the cabinet, one to the left, the other to the right. On the longer ASI cable an Extension Plug is installed.

Communication:

• Using ASI Interface Analyser (version 1.3.4.0) to verify communication: 3-4 components have errors detected. Consecutive error log shows nearly all errors are of “1x” consecutive frequency. Advanced statistics show error rates 1% to 30%.
• Disconnecting a section of the ASI bus will eliminate most communications errors. (i.e.) disconnect the ASI cable routed to left of the power supply that has about 50% of the ASI slaves or disconnect the ASI cable routed to the right with the other 50% slaves. Either section of the ASI bus can communicate error free by itself.
• Installing a repeater and new power supply to create two ASI bus segments eliminates most of the errors. However one device, node address 3 on segment 1 of the repeater, continued to have an error rate of 20%.
• Node address 3 was on segment 1 of the repeater; node address 2 was on segment 2 of the repeater. Switching the node addresses eliminated all communication errors. Note the physical device was not relocated; only the logical address was changed.

Reconfiguring ASI Node Addresses:

• The final solution noted above to eliminate errors on node 3 led our team to further investigate the benefit of modifying node addresses to reduce communication errors.
• On a similar ASI system as described above the following was tested.
• A repeater was installed but still 3-4 components continued to have 5 - 20% error rate. The repeater was removed.
• Replacing and/or relocating components were tested but did not resolve communication errors.
• By strategically swapping the node addresses for components with errors we were able to eliminate all communication errors. It appears that having contiguous logical addresses assigned to components physically located within close proximity to each other has some benefit for communication. During the process of reconfiguring logical addresses we encountered some change in error rate with slave devices directly beside the device being modified. Eventually eliminating all communication errors.

Questions:

Link 20E Master:

• The Link 20E manages communication: the master broadcasts a message across the ASI bus seeking a response from a specific node (i.e.) node 3; node 3 receives and responds with a status telegram. If telegrams are broadcast across the entire ASI bus why would re-assigning logical addresses improve or degrade communication?