RSLogix 5000 REAL to DINT

[awaege]

I've always been a bit confused as to how RSLogix 5000 converts between data types of floating point to whole numbers. For the application I am working on I want to use a "floor" type function to simply round down a real value to the nearest whole number. Will simply passing the real output of the computation truncate the decimal? How is this handled?

Thanks for any insight.

 

[widelto]

just move real to dint

 

[OkiePC]

I believe that it will round, not truncate when you MOVe a REAL to a DINT:

http://www.plctalk.net/qanda/showthread.php?t=97762

 

[ASF]

Logix controllers use a "Dutch rounding" method (also known as convergent rounding, statistician's rounding, Gaussian rounding, or bankers' rounding) when converting from REAL to DINT. This means that it rounds to the nearest integer. When the REAL is exactly halfway between integers (e.g. 1.5), it rounds to the nearest even integer. So:


1.499999 >> 1
1.5 >> 2
2.5 >> 2
2.500001 >> 3
3.5 >> 4
4.5 >> 4


and so on.

 

[Mickey]

5000 converts between data types of floating point to For the application I am working on I want to use a "floor" type function to simply round down a real value to the nearest whole number.
Thanks for any insight.

If you want it to round down then subtract .5 ,then move it.

 

[ASF]

Or just use the TRN (truncate) instruction. You can also use the TRN operator inside a CPT instruction.

1.00001 >> 1
1.99999 >> 1

Obviously, this only applies if you always want to round down, as opposed to rounding to nearest. Given your description of a "floor" function, I expect you want the former.