Melsec H Mode Fault Question

[Agruel]

I have a Q02H PLC on a Melsec Net H network with 25 other Q series PLC's. This network has been up and running for about 2 years without many problems. The Q71LP21-25 card that is communicating on the network gets a fault and is stopping the processor from running as a critical error. The fault code that the processor is receiving is 3101. I cannot tell what the problem is on the card because when I try to do a diagnostic on it all I get is a popup message telling me that the card is faulted. I have compared the refresh settings to some of the other machines and the settings all match up. This problem only pops up about once every other day and is hard to diagnose. Does anybody have any ideas or experience with a similar problem?

[navillusi]

From the GX Developer Help: Error code list[Error code 3101] Error code(SD0) 3101 Error message LINK PARA. ERROR Common information(SD5-15) File name/ drive name Individual information(SD16-26) Parameter number LED status(RUN) Off LED status(ERROR) Flicker PLC operating state Stop Diagnostics timing At power ON/Reset/, STOP -> RUN Error description & cause (1) The link refresh range exceeded the file register capacity. (2) When the station number of the MELSECNET/H module is 0, the inter-PLC network parameter setting has been made. (3) When the station number of the MELSECNET/H module is other than 0, the remote master parameter setting has been made. (4) The network No. specified by a parameter is different from that of the actually mounted network. (5) The head I/O No. specified by a parameter is different from that of the actually mounted I/O unit. (6) The network class specified by a parameter is different from that of the actually mounted network. (7) The network refresh parameter of the MELSECNET/H, MELSECNET/10 is out of the specified area. Corrective action (1) Change the file register file for the one that enables entire range refresh. (2) Correct the type or station number of the MELSECNET/H module in the parameter to meet the used system. (3) Check the network parameters and mounting status, and if they differ, match the network parameters and mounting status. If any network parameter has been corrected, write it to the CPU module. (4) Confirm the setting of the number of extension stages of the extension base units. (5) Check the connection status of the extension base units and connectors. When a display device is connected to the main base unit or extension base unit, check the connection status. (6) If the error occurs after the above checks 1) to 3) have been made, it suggests a hardware fault. Therefore, change the faulty module. Alternatively, contact your nearest Mitsubishi representative. You say that you have checked the refresh setting of all the other PLC's, have you checked the physical connections of the cards? Next step would be to do as step 6 advises, try a new card.... It may be cheaper to move the suspected faulty card into another PLC and see if the fault follows the card or stays with the CPU

[Agruel]

Thanks for the reply. I checked the list of things on your list, but everything seems to be right. This card is on a network that has been running for about 3 years and I am having trouble finding a reason for the recent problem. Nothing has changed as far as I can tell. I also checked the forward loop and reverse loop status to see if there was a fiber break, but the whole network shows up good. So, I replaced the card thinking that it had to be the problem and it faulted out again today. I am running out of ideas of what it could be that would cause it to fault out only once a day or every other day like this after it ran for so long. Thanks again for the reply.

[navillusi]

The next thing I would suggest would either be: Completely clear the CPU memory and download the program again, possible corruption of memory areas? maybe unlikely but a low cost option Replace the backplane, if there is a fault on the backplane this could throw up a hardware fault Replace the CPU, if all else has drawn a blank, could be the only option left. BTW, have you spoken to your local Mitsubishi tech support yet? Have they had any ideas?

[panic mode]

i was about to write same thing but navillusi was faster. check the backplane, try to relocate card to a different slot of the same backplane if you don't have spare part (you could change i/o map in parameters but rest of the program would remain the same).

[Pierre]

My experience with Mitsu tells me that when all they say has check out OK then look for something else. Is there anything that could not be random? When does this happens, time of day? Beginning of shift? Was there any large conductors installed or relocated at all at this location (>200A)?