Hello
we are currently using a CPU X20CP3687X with connected a slave with installed an AnyBus CompactCom 40 module and we are experencing a problem when the master (CPU) is restarted.
It seems the slave does not reconnect again anymore to the master and it necessary to restart also the board.
I have not seen this HW before. I’m also not sure if anyone has actually used this. But if the producer claims that it supports Powerlink, maybe you can try asking the producer?
I agree with Jaroslav, and I’m pretty sure it’s not a specific behavior of the Powerlink protocol.
Powerlink includes hot-plug and network re-setup mechanisms.
According to the Powerlink specification, a CN (controlled node) “… shall change it’s state to NMT_CS_PRE_OPERATIONAL_1 …” after detecting a higher number then a predefined threshold value of losses of SoC (Start of Cyclic) or SoA (Start of Acyclic) frames.
This losses are exactly what happens, when the master is rebooted (or, for example, when the network connection gets lost).
The state NMT_CS_PRE_OPERATIONAL_1 is the state inside the CN’s internal state machine, where the CN then has to wait for the ident request of the master and where the “plug&play” initialization and parameter setup process starts, to integrate the CN into the MNs network.
Since this is always the same process to go through, independent if the whole network is powered on, or if a new CN is hot-plugged, or if the MN restarts, and because of two of those three use-cases are working (as I understood), I can’t imagine that it’s a common protocol based behavior.
Sure, that’s not the solution for your issue, I’m aware of that
The information only should clarify, that it’s not a “normal protocol based behavior”, and that the manufacturer (and maybe also our support team, if the manufacturer needs assistance) has to dig deeper into.