Hi,

Can someone explain better the purpose and usage of the AT identifier. S7-SCL manual is not very helpful. And of course, it is very hard to search the forum for the keyword - "AT". 

The only thing given is something that is not understandable:

Simple use case - I have 4 byte vars and I want them to be a double word.  

MyBytes : ARRAY[1..4] OF BYTE;

MyDword AT MyBytes : DWORD;

code:

MyBytes[1] := SomeByte1;

MyBytes[2] := SomeByte2;

MyBytes[3] := SomeByte3;

MyBytes[4] := SomeByte4;

SomeOtherDword := MyDword AND 16#89ABCDEF;

Another one: my dumb Modbus device sends me 2 integers claiming it is a REAL. (actually in case of Modbus there are some registers, some of them 16 bit and for some cryptic reason they are used in pairs to store a REAL data type numbers).  

MyInts : ARRAY[1..2] OF INT;

MyReal AT MyInts : REAL;

code:

MyInts[1] := SomeInt1;

MyInts[2] := SomeInt2;

OtherReal := MyReal + 1.0;

In general AT let a programmer to change the definition of some specific memory area.  

Mr. Jacek already published a good answer by the time. My version is not much different but there is the mention of how data is seen from outside the code block.

qwazee

Are your doubts resolved?

Thanks, I got some clues after all this. 

However, I didn't recognize how AT is used  in your example. Can you clarify, please? 

If I understand correctly, any data can be seen as any other type of data using AT, as long as viewed data is smaller than declared data. Like, DWORD can be seen as series of 4 bytes, two successive integers in array can be treated as one full REAL.

What will happen if we are not careful enough, and data don't match? For example, if we use two bytes, and try to see it as a DWORD? 

E.g.

MyByte : Array of Bytes [0..1]
MyDword AT MyByte : DWORD

(don't panic if my syntax is not perfec) :)

banconi

qwazee

Are your doubts resolved?

Thanks, I got some clues after all this. 

However, I didn't recognize how AT is used  in your example. Can you clarify, please? 

If I understand correctly, any data can be seen as any other type of data using AT, as long as viewed data is smaller than declared data. Like, DWORD can be seen as series of 4 bytes, two successive integers in array can be treated as one full REAL.

What will happen if we are not careful enough, and data don't match? For example, if we use two bytes, and try to see it as a DWORD? 

E.g.

MyByte : Array of Bytes [0..1]
MyDword AT MyByte : DWORD

(don't panic if my syntax is not perfec) :)

Hello.

The example from Mr Jacek is superior. You are well on your way if you understand what he was saying.

Having an overlay that is smaller than the original tag is possible and allowed. Having an overlay that is bigger should give a compiling issue. It will be good for you to make a test program and run it through the simulator to get a feel for what can be done. 

The S7-1200 and S7-1500 can also have an AT-overlay definition, non-optimized data structure is required. Different than the older generation, the S7-1200 and S7-1500 allows AT in LAD / STL / .. also.

A simulation in TIA Portal might be a little more friendly to do. Try it.

~

banconi

What will happen if we are not careful enough, and data don't match? For example, if we use two bytes, and try to see it as a DWORD? 

E.g.

MyByte : Array of Bytes [0..1]
MyDword AT MyByte : DWORD

(don't panic if my syntax is not perfec) :)

banconi

If I understand correctly, any data can be seen as any other type of data using AT, as long as viewed data is smaller than declared data.

Actually, as you already noticed, you can't. The compiler keeps you from looking under the skirt. 

But opposite is possible. So, in your case (after changing the order) you will be able to see only 2 bytes of of 4 .