Hello, Please I have this same issue now.

Could you please share a picture of the solution if you still have it?

Thanks very much.

My Setup is :

Siemens S71500 TIA V!^ Updt 4.

bekchoff Twincat 2.9 PC panel.

A_Dent

 Changing the byte order with SWAP did the trick!

They were the wrong way around, works like a charm now (read as a DWord, SWAP, and then CONV into the Real variable in the DB where we want it)

. Worth note for anyone else coming here (Beckhoff IPC was a Intel-based one, might be different with ARM ones?)

Huge thanks huggy_d1!

huggy_d1

What you are doing with the CONV block would make sense if the byte order is correct.

As an experiment, look at the ENDIAN based instructions (you can search the instructions by typing ENDIAN at the top of the instruction pane and press enter or the search down icon just below the input block).

F1-Help how the swap and such works.

Try swapping the byte order.

Of course, the right answer is to find out the order of the bytes coming from the IPC, and then match them up with the S7-1500 plc.

Also, if you simply MOVE the DWORD to a REAL rather than convert, you will have better luck if the byte order is correct.

That said, the odds are that the byte order will be wrong.

In summary, replace CONV with MOVE.  If it works, quit while you are happy.  If not, try SWAP before the MOVE and see what happens. 

A_Dent

Hi There,

I have a program that is getting a 32-bit Float as an input from a Beckhoff IPC (via an EL6331-0010 module) that I wish to treat the same way in S7 (so as a "Real").

However the input data "area" (IDXX) of 32 bits is always acting as a DWord in the S7 code: it displays as a Hex value and maths operations come out incorrect (treat the value as 0).

I have tried setting the tag to "Real" and moving to a "Real" number in a DB, and I have tried setting the Tag to DWord and running through the CONVERT block. There are no errors but it will not treat the value as a Real (see the ADD in the image).

We are using a S7-1515T, TIA Portal V15.1, and communicating via EtherCAT and ProfiNET (with EL6631-0010 as a slave on both).

Is there way to treat a 32-bit input area as a Real? Or to convert the input bits I have to a Real (assuming the format is already correct?).

Our other quick fix option is to times by 1000 and divide down (as Integers can be communicated fine), or generate the Real in code from the bits (sign bit, 8bit exponenet, 23 bit mantissa) but I'm sure I'm missing something!

Attached an annotated pic of the code at runtime to give an example of the issue to help explain what I mean.

 

 

------------------------------------------------------------------------------------------
Split from Issue Handling Real Inputs.

Closed cause question posted twice.

The second thread with the subject DP CP coupler data transfer and communication is published within the conference Decentral Peripherie.