Hello,

I have a system of the CPU 1513-1 PN and CM PtP RS422 / 485 BA. 

Send_P2P I use to send data to the slave, but the data is not sent to the CM module output (TX led does not light up and listen to the program (com sniffer) data "does not see".

Why Send_P2P does not work?

TIA Portal V13 SP1

protocol: freeport (ascii)

p.s. translated by google)

Hello,

as I cant look to your project (don't have needed update now), can you pls explain what Slave device type is? What communication frames / type you need prepare?

Slave - remote I / O module with RS-485 interface protocol Modbus RTU / Modbus ASCII. I want to implement the Modbus protocol by freeport.

I suppose that if I set a REQ bit, then Send_P2P instruction should be executed and the Tx LED should blink, or Send_P2P should have the status of an error code. But Send_P2P always has the status of 7000.

@Duris_I,

Here is a PDF of the code if it will help you.

I am missing the configuration blocks.
Is it needed if the configuration is done in hardware?

w

skygrabber

I tried using these instructions, but the situation has not changed. If I'm not mistaken, port_config, send_config and receive_config used when you want to dynamically change the communication parameters. 

What you say makes sense.

I have had a look a sample code published on the forum using S7-1200 linking to HyperTerminal. There also the port configuration is not done. Unluckily those have a definite split between send and receive.

What I do see is from the help in the software:
"If several instructions need to read or write data records simultaneously on a CPU, the user program may need to create a time offset between the calls of each instruction."

What I can tell you is that your code do have an overlap between the send and the receive. There should be a small gap between those but not an overlap. While EN is triggered to the receive, the port is actively listening to what is said...

Receive is active while EN is active. That EN should be off while Send is requesting: between positive edge of REQ and DONE.

I hope it makes sense and is helpful.
w

Hello friends,

thnx William for pdf. Quickly have looked to pdf, in my opinion better will be for first check SEND functionality without wired connection to partner (unplug cable to Slave), simply also drive that communication with bit memory, also in SEND instruction, not use clock memory now. So on REQ define your new bit memory %MX.X as you need. Simply via Watch table you can turn it ON and after success TX light on (watch carefuly for that), reset that REQ memory (also you can implement this RESET programmatically after succes DONE). I will be not use that SEND instruction - DONE memory "send_done" as start bit for RECV instruction. Simply I'll be also define memory bit for this EN_R input and leave it ON permanently as soon as SEND DONE will be true. Simply after success send, receive must be initiated as soon as possible. This RECV step can be tested with wired connection to partner, but after success test of SEND. Pls skygrabber look carefuly for this, your stauses like DONE, NDR are available with it's val;ues only for one cycle. Make tests and give us info.

If we can talk about needed Port cfg, it can be done via HW cfg. In some cases it can be change programatically, as you preffer.

Update: I just see little difference between Picture and pdf, you've change clock memory from M110.7 to M0.7, did you find some problems with functionality?

Hello,

probably I found problem, on SEND instruction is buffer defined like:

"out".Static_1[0]

I think this is problem, better can be point to whole buffer like

"out".Static_1

This can be solution for right send with defined bytes length. also on RECV is needed probably this change like on SEND

"in".Static_1