PLC data topic hub
A field reference for signed and unsigned 32-bit values, byte and word order, Modbus register offsets, and IEEE 754 REAL conversion.
A PLC value is a bit pattern plus an interpretation contract. The same 32 bits can be displayed as a signed DINT, an unsigned DWORD, a hexadecimal value, four bytes, two Modbus registers, or an IEEE 754 REAL. Converting the representation without confirming that contract can produce a plausible-looking but incorrect number.
Signedness changes the numeric meaning without changing the bits. For example, 0xFFFFFFFF is -1 as a signed 32-bit two's-complement integer and 4,294,967,295 as an unsigned integer. Byte order and 16-bit word order introduce another layer when the value crosses a protocol or device boundary.
Modbus documentation may label the first holding register 40001 while a client uses zero-based offset 0. Record the published address, function code, client offset, register width, signedness, byte order, and word order before changing a mapping.
Original public reference
Use this as a screening reference. Confirm the installed equipment, configuration, and site requirements before making a field change.
IEEE 754 example: 12.5 = 0x41480000
| Layout label | Byte sequence | 16-bit registers | First check |
|---|---|---|---|
| ABCD | 41 48 00 00 | 4148 / 0000 | Documented high word first |
| CDAB | 00 00 41 48 | 0000 / 4148 | Swap 16-bit words |
| BADC | 48 41 00 00 | 4841 / 0000 | Swap bytes in each word |
| DCBA | 00 00 48 41 | 0000 / 4841 | Reverse words and bytes |
Preserve the original hexadecimal pattern before attempting a conversion.
Treat byte order and 16-bit register order as separate decisions.
Confirm whether a Modbus document uses human-readable 4xxxx notation or protocol offsets.
Validate REAL conversion with a known value such as 12.5 = 0x41480000 in standard IEEE 754 byte order.
Supporting guides
Recognize when the same 32 bits represent a negative DINT, a large unsigned quantity, a bit mask, or protocol data before converting or comparing the value.
Separate byte swaps from 16-bit word swaps when PLCs, HMIs, gateways, historians, and Modbus devices exchange 32-bit integers or floating-point values.
Resolve 40001 notation, zero-based protocol offsets, function codes, signed values, and multi-register data before blaming communication quality.
Decode a 32-bit single-precision floating-point value only after proving the source bytes, word order, special values, and expected engineering range.
A signed 32-bit DINT uses the highest bit as the sign bit. When bit 31 is set, the same pattern is interpreted as a negative two's-complement value even though an unsigned display would show a large positive number.
A 32-bit value spans two 16-bit registers. Some devices transmit the high word first and others the low word first. That word-order decision is separate from byte order inside each register.
Use the device register map first, then preserve both registers in hexadecimal and test the documented word and byte order. A correct IEEE 754 interpretation should match a known operating value and remain sensible across more than one sample.