Floating Point Registers + Allen Bradley

I am having trouble using floating point numbers. When I use the DIV function and specify an F8 register location for the output it rounds the number. Additionally I thought that foating point registers were 32 bit it looks like to me that the F8 registers are 16 bits.

What am I doing wrong?

What is your actual divide?
What are the operands?

INT / INT still yields an int.

If you are trying to divide a number by a constant (or a constant by a number), make the constant a float (123.0 instead of 123).

And in most AB platforms (all I work with) a float is indeed a standard 32bit approximation. To copy one, you need to copy two words (16 bit values), not just one.

floating points (ex: F8:such-and-such) are indeed 32-bit structures ... but they're not the same as a 32-bit integer ... page F-6 (way at the back) of this manual might help you understand them ...

if you can give us some specific examples of what values you're trying to work with, we might be able to suggest some "work-arounds" that would help ... we also need to know what platform you're working with ... some processors support different formats than others ...

Here is what I am trying to do with an AB simulator.

When I run this the value in F8:2 is 0 instead of the decimal value.

F8:0 is 2 trying to do 2/3

Trying to take an integer and divide it by an integer which will yield a floating point decimal

Place 3.0, not 3, as source B in the DIV instruction.

Just a question?

sgsims said:

Here is what I am trying to do with an AB simulator.

When I run this the value in F8:2 is 0 instead of the decimal value.

F8:0 is 2 trying to do 2/3

Click to expand...

In your example wouldn't the empty rung 002 cause an error that stops the program running? I haven't used the simulators but to download to my 5/03 or 5/05 I have to verify the program first and an empty rung is not allowed. Hope this might help.