Hello All,

I have similar problem with Modbus

I use CPU 1516 + ET200SP with RS485 Module on Porfinet. Modbus works OK, but if there is an interruption somewhere in Profinet network, Modbus hangs up and I can't find possibility to reset it, but hardware reset to CPU. The whole ET200SP keeps working, only Modbus hangs up.

With a SCALANCE router, I've created the star-topology for Profinet network and have put ET200SP on separate "star ray", but the problem keeps persisting, even if there's a Profinet interruption on other "star ray".

If someone could advise how to resolve this it would be of great help.

------------------------------------------------------------------------------------------
Split from Busy output stays on if resetting communication.

Hello Pavel,

maybe you can try make test, in case of interruption of Profinet and again established connection (you can initiate this sitiuation manually), try make MB_COMM_LOAD init, prepare some prior bool conditions to this test and also put COMM_LOAD to some code outside of OB100. You need probably know when Profinet is ok and reinit all instruction sequence, especially that MB_COMM_LOAD. I expect you MB_COMM_LOAD in OB100, therefore probably only init of PLC can help you.

Hi Duris,

thank you for your reply.

I have already tried what you say, this doesn't help. MB_COMMS_LOAD is in OB1 (actually some other cyclic OB) and I provided the re-init bit to reinit at any time. This doesn't help. 

Sorry, I've noted this topic and added my question. Next week I gonna be on site then I analyze one more time what happens in details and put additional info here.  

Hello

I'm facing the same problem with S7 1518 F Cpu and ET 200SP PTP CM

looking closer to MB Master DB I can see that there are some data on the RX Buffer

Is there any recommendation?

I'm testing the communication these days, i'll be back for more informations soon

Thank you

Hello.

Did You solved the problem ?

Is'n not best idea to post another message under Your own.

But...

It seems that its's an Siemens module issue.

I use the 1512SP - 1 PN CPU with ET200SP CM_PTP.

I have changed the modbus cables in the panell.

Since then I didn't get any problems about the communication.

It seems that the module frezzes with the electromagnetical noise comming from wrong wiring or wires shielding.

Update.

I organized my program in that way that the all the calls of modbus master were done in function block called in the main OB.

As in the network there are 7 ATV320 inverters the cycle of communication with data exchange was around 3 seconds, which is quite bad for the network with baud rate 38400.

I tried different solutions to speed up the communication so i found an idea that maybe the cycle of the main program is the problem, now the averange cycle time is around 40ms.

Nothing helped so i found idea, i created cyclic ob31 with cyclic time of 10000 ms, the communication cycle to exchange data with all 7 inverters was less than 1 second, but it worked only couple of hours. After the modbus master got frozen, i changed the cyclic time of the ob to 20000ms, the cycle of the communication changed to around 1,5 seconds, wchich was aceptable.|
But again one of the modbus master got frozen with buisy status. To reastart it i had to restart CPU, nothing else helped.

Now i returned to fc in the main ob. 

It worked for couple of weeks without problem.

But now what is the problem ?

How to incerase the speed of communication , and what's the problem.

I read couple of topics about modbus master freezing, but there's no solution, no reason.
Any ideas ?

Hello.

I have rebiuild the software to be sure about the sequence.

Now i have 14 calls of modbus_master fuction.

The instance db is the same, but each call is using different data.

The read of first node is triggered by bool "read 1" when modbus master status "done" or "error" is set , the "read 1" is set to 0 and the "write 1" is triggering the send operation, when send is finnished with done or error , next call of reading operation is triggered.

Is such a mode i dont understand which timer can be issue.

The trigger for each modbus_master remains on until the function will return error or done.

Is it possible that the error is caused by changing the data which should be sent ?
Till now my program was writing the data from the DB with speed and other parameters to the data_PTR db each cycle, so in this mode it was possible that sometimes the data was changed during the sending, of corse if the sending operation takes more time than the cycle of the PLC.

Sliwka

Hello.

I have rebiuild the software to be sure about the sequence.

Now i have 14 calls of modbus_master fuction.

The instance db is the same, but each call is using different data.

The read of first node is triggered by bool "read 1" when modbus master status "done" or "error" is set , the "read 1" is set to 0 and the "write 1" is triggering the send operation, when send is finnished with done or error , next call of reading operation is triggered.

Is such a mode i dont understand which timer can be issue.

The trigger for each modbus_master remains on until the function will return error or done.

Is it possible that the error is caused by changing the data which should be sent ?
Till now my program was writing the data from the DB with speed and other parameters to the data_PTR db each cycle, so in this mode it was possible that sometimes the data was changed during the sending, of corse if the sending operation takes more time than the cycle of the PLC.

 Hello Sliwka, Did you solved the problem? I have exaclty the same problem now. And the solution seems to be to slow the transmits rate (not speed, but calls to SEND_PTP per second).