11.6 Conversion functions (Conversion of data type)
11.6.3 Conversion functions for floating-point numbers
The conversion function CONV converts fixed-point numbers into floating-point numbers and vice versa. For conversion from floating-point numbers into fixed- point numbers, there are additional features that differ in rounding behavior (Table 11.3). If an error occurs during conversion, no conversion takes place.
Fig. 11.14 Conversion function CONV, representation and function CONV
DT1 to DT2
EN EN
OUT
OUT ENO
ENO Conversion function CONV
LAD FBD
IN IN
Declaration Name
EN ENO IN OUT
– – INPUT OUTPUT
BOOL BOOL Data type 1 Data type 2
Description Data type
CONV DT1 to DT2
Function:
The following data type combinations are possible:
Data type conversion for from Data type DT1 to Data type DT2 BYTE, WORD, DWORD, BYTE, WORD, DWORD USINT, UINT, UDINT,
SINT, INT, DINT USINT, UINT, UDINT, USINT, UINT, UDINT, SINT, INT, DINT,
SINT, INT, DINT BYTE, WORD, DWORD,
REAL, LREAL
INT BCD16
DINT BCD32
BCD16 INT
BCD32 DINT
USINT, UINT, UDINT,
REAL, LREAL SINT, INT, DINT,
REAL, LREAL SCL OUT := DT1_TO_DT2(IN);
Bit sequences
Fixed-point numbers
BCD numbers
Floating-point numbers
Enabling input Enabling output Input tag Output tag
The conversion function CONV is used to convert data types. The content of the tag with data type DT1 at the IN parameter is converted and transferred to the tag with data type DT2 at the OUT parameter.
Conversion of fixed-point numbers into floating-point numbers with the CONV function
The conversion function CONV converts a fixed-point number (SINT, INT, DINT, US-INT, UINT, UDINT) at the IN parameter into a floating-point number (REAL, LREAL) and outputs the result at the OUT parameter. If the accuracy suffers during the conversion of DINT or UDINT into REAL, ENO is set to “0”.
Conversion of floating-point numbers into fixed-point numbers with the CONV function
The conversion function CONV converts a floating-point number at the IN parame- ter (REAL, LREAL) into a fixed-point number (SINT, INT, DINT, USINT, UDINT) and outputs the result at the OUT parameter. During conversion figures are rounded to the next integer. ENO is set to “0” if during conversion the allowed number range is exceeded or an invalid floating-point number is specified.
Conversion of REAL into LREAL and vice versa with the CONV function The conversion function CONV converts a floating-point number at the IN parame- ter into a floating-point number with a different data type (REAL to LREAL or LREAL to REAL) and outputs the result at the OUT parameter. ENO is set to “0” if the per- missible numerical range is violated during the conversion or if an invalid floating- point number is specified.
Conversion of REAL into fixed-point number with the CEIL function
The conversion function CEIL converts the data type REAL or LREAL of the tag at the IN parameter into a fixed-point data type, and outputs the result at the OUT pa- rameter. During the conversion, CEIL rounds-off to the next largest integer. ENO is set to “0” if the permissible numerical range is violated during the conversion or if an invalid floating-point number is specified.
Table 11.3 Conversion of floating-point numbers Function Conversion
CONV Conversion from USINT, UINT, UDINT, SINT, INT or DINT to REAL or LREAL Conversion from REAL or LREAL to USINT, UINT, UDINT, SINT, INT or DINT CONV With rounding to the next integer (as ROUND)
CEIL With rounding to the next higher integer FLOOR With rounding to the next smaller integer ROUND With rounding to the next integer
TRUNC Without rounding
Conversion of REAL into fixed-point number with the FLOOR function The conversion function FLOOR converts the data type REAL or LREAL of the tag at the IN parameter into a fixed-point data type, and outputs the result at the OUT pa- rameter. During the conversion, FLOOR rounds-off to the next smallest integer.
ENO is set to “0” if the permissible numerical range is violated during the conver- sion or if an invalid floating-point number is specified.
Conversion of REAL into fixed-point number with the ROUND function The conversion function ROUND converts the data types REAL and LREAL of the tags at parameter IN to a fixed-point data type and outputs the result at parameter OUT.
During conversion, ROUND rounds to the nearest integer. If the result is exactly between even and odd numbers, the even number is selected: ROUND(0.5) = 0, ROUND(1.5) = 2. ENO is set to “0” if during conversion the allowed number range is exceeded or an invalid floating-point number is specified.
Conversion of REAL into fixed-point number with the TRUNC function The conversion function TRUNC converts the data type REAL or LREAL of the tag at the IN parameter into a fixed-point data type, and outputs the result at the OUT pa- rameter. TRUNC returns the whole number part of the number to be converted;
the fractional part (the decimal places) is “truncated”. ENO is set to “0” if the per- missible numerical range is violated during the conversion or if an invalid floating- point number is specified.
Fig. 11.15 Conversion functions for floating-point numbers, representation and function DT1to DT2
EN EN
OUT
OUT ENO
ENO
Conversion functions with rounding-off for floating-point numbers
IN IN
Function:
ROUND CEIL FLOOR TRUNC
Data type 1 (DT1): REAL, LREAL
USINT, UINT, UDINT, SINT, INT, DINT, REAL, LREAL Declaration Name
EN ENO IN OUT
– – INPUT OUTPUT
BOOL BOOL Data type 1 Data type 2
Enabling input Enabling output Description Data type
DT1to DT2
SCL OUT := Function_DT2(IN);
Input tag Output tag Function
LAD FBD Function
The ROUND, CEIL, FLOOR and TRUNC functions convert a floating-point number at the IN para- meter into a number at the OUT parameter, taking into account three types of rounding-off:
Conversion without rounding-off
Conversion with rounding-off to the next integer Conversion with rounding-off to the next largest integer Conversion with rounding-off to the next smallest integer
Data type 2 (DT2):
Overview of rounding-off when converting floating-point into fixed-point Table 11.4 shows the different effects of rounding when converting floating point to fixed point using the example of the REAL data type in the range between -1 and +1.