Real Data Type limits

RatherBeFishing

Member
R
Join Date
Feb 2017
Location
Anniston, AL
Posts
24
Am I understanding correct that the "real" data type in an RSLogix 5000 is a 32-bit (double integer), but the "rules" change as needed?

Say I had a negative number, would it limit it to: -32767.9999847412109375 or somewhere there abouts?

I had a compactlogix limit me to: 65535.0 (approximately, I didn't write it down when it happened).

It appeared to limit the number at the MAX for an unsigned integer. Why wasn't the decimal portion at something other than a zero? Why would the limit be for an UNSIGNED integer? I was incriminating it with a small number (somewhere in the neighborhood of .00218666). o_O

The number just stopped btw. Kind of figured I would have had a minor fault showing..
 
The REAL is completely different from a DINT. It's an IEEE 754 single-precision floating point encoded value.

In general a REAL value in a ControlLogix can represent a value up to about 3.402823 × 10^38, but only to about six significant digits.

The problem I think you're encountering is that you're doing an accumulator using a small increment and at some point, the relative difference between the accumulated value and the increment value is zero, so the ADD instruction simply has no effect on the REAL value.

When the increment is 1, this happens at about 7 million.

Ideally, use an integer accumulator or some kind of rollover method when you need to do large accumulations of small increments.

There are numerous threads on this Forum and elsewhere on the Internet discussing this issue in computer science.
 
Ken Roach said:
The REAL is completely different from a DINT. It's an IEEE 754 single-precision floating point encoded value.

In general a REAL value in a ControlLogix can represent a value up to about 3.402823 × 10^38, but only to about six significant digits.

The problem I think you're encountering is that you're doing an accumulator using a small increment and at some point, the relative difference between the accumulated value and the increment value is zero, so the ADD instruction simply has no effect on the REAL value.

When the increment is 1, this happens at about 7 million.

Ideally, use an integer accumulator or some kind of rollover method when you need to do large accumulations of small increments.

There are numerous threads on this Forum and elsewhere on the Internet discussing this issue in computer science.
Click to expand...


I know it has been a while since I posted this, but I thought I would come back and say thanks! Your responses scattered throughout the web have helped me a lot in the past few months!
 

Similar Topics

I
LREAL Data Type in Control Logix L71
Is there a way to use LREAL Data type on L71. I am getting 64 bit Double (IEEE754 Double precision 64-bit) data from a modbus device onto PLC via...
Replies
6
Views
952
drbitboy
drbitboy
E
Can't send REAL data from cimplicity scada to AB plc-modbus TCP SLAVE type
Hi All, I'm using AB plc 5370(Modbus TCP slave progm type) with GE cimplicity scada V8.1. For DI,DO,AI communication is working good. But, I...
Replies
10
Views
2,551
thingstodo
T
R
s7 1200 indirect addressing using real data type
Hi, Im new to SCL, but decided to start learning it - if only to be able to do some simple jobs such as indirect addressing. got some simple...
Replies
3
Views
6,901
Rob.Hill
R
W
controllogix data type REAL's Struecture
Hi all, As you know, when reading from modbus floating point data (32 bits) by prosoft MCM card, we use real data type in controllogix to...
Replies
2
Views
7,048
kamenges
K
L
Read real data from hmi to plc
Hi Iam using monitouch hmi(V9 soft) with omron plc cj2m (CX programmer). In this I want to read a data from hmi to plc. The data was like...
Replies
0
Views
94
Lee sung
L
Back
Top Bottom