Losing PM 866 Controller redundancy due to Modulebus connectivity issue.
Configuration : Red. PM 866 Controllers with 6 Clusters in Modulebus (TB840A)
Note: The PM 866 controller are replaced with the new ones after the recent field alert.
The controllers are getting out of redundancy very frequently. On checking the controller log, found out that the one of the cluster is not visible in the backup controller. The customer has brought the optic fibre technician and he checked the signal, re polished the cables and put it back in service. It worked for 2 -3 days. forgot to say that this is a marine project (vessel). Now again the problem started, I have attached the controller logs of primary and backup just after resetting the backup controller (which was in fault state). My question is
1. Is the any possible reason for this other than cable issue. The customer is challenging my finding, saying that the cable continuity and signal strength has been tested and values were good enough.
2. Is there any filter time for controller to wait for the signal to be back.
Asked 5 months ago
Answers
1
I agree that the lower CPU does seem to lose contact with a cluster. But it is possible that this is just the first detection of the whole module bus connection to the Lower CPU failing.
1. Yes. Lots. Pretty much the entire hardware chain could be suspect. Without a drawing of exactly what is installed its not possible to say.
2. No. There is a setting for Module Bus scan rate which "might" help but could instead make things worse. Its not possible to say if you can change this without knowing the details of your application software. In any case - it's just hiding the problem. The modulebus should be stable and the communications to the cluster should not stop.
You say you've checked the Fibre connections, but what about the pre-made POF cables ? Modems etc ?
1. Yes. Lots. Pretty much the entire hardware chain could be suspect. Without a drawing of exactly what is installed its not possible to say.
2. No. There is a setting for Module Bus scan rate which "might" help but could instead make things worse. Its not possible to say if you can change this without knowing the details of your application software. In any case - it's just hiding the problem. The modulebus should be stable and the communications to the cluster should not stop.
You say you've checked the Fibre connections, but what about the pre-made POF cables ? Modems etc ?
Answered 5 months ago
Add new comment