Communication issue Siemens S7-300 / B&R x20

[nicocharlebois]

Hello everyone!

I'm currently working on a project which requires communication between an S7-319 PLC (programed in TIA Portal V14) and an x20 B&R. The Siemens S7-319 acts as the server and communicates via ISO-on-TCP throught it's CP-343 card. It uses AG_SEND and AG_RCV blocs to manage the communication and calls them every 50ms (send then receive).

The issue I have is that the S7-300 can receive all 10 bytes the B&R is sending right away, but the 10 bytes the S7-300 is sending shows up with a 45-60 SECONDS delay... If I shut down the power to the S7-300, the B&R will still show up the 10 bytes for about 45-60 seconds.

Have you ever encountered a similar problem? I'm open to any suggestions you may have.

Thanks for your time,

Nick

 

[seth350]

What is the scan rate of the task set to in the B&R?

 

[nicocharlebois]

The longest scan time in the B&R was around 20ms.

We messed around a bit with the parameters in both the B&R and the S7-319. We found out that if we increase the allocated space for the reception of the data in the B&R it works. Also if I send less stuff it helps too. It looks like the AG_SEND bloc sends all the data at once and the B&R can't manage that

So we increased the B&R receiving buffer from 10 bytes to 50 bytes and I reduced the size of what I'm sending from 10 bytes to 1 byte (I only need to send 1 byte anyway, the rest is spare space).

We don't really know why, but it seems to do the trick for the moment.

 

[seth350]

Hmmm...

You were sending 10 bytes of data, so did you set up the buffers for exactly 10 bytes?

Unless the PLCs take care of the Ethernet frame, padding, ipv4, etc and are just worried about the payload, you may need to take into account the size of a stripped down “no data” packet. Some reading mentions that an empty packet could be 64 bytes and that there is a minimum packet size. Which is where padding comes in to fatten the packet.

Your issue is probably not at all that complicated, but I would be interested to see how it works if you increase the buffer to 74 bytes. Just for grins...