Communication Protocols · Free Tool

Modbus Float Converter (IEEE 754, Two Registers)

Turn two Modbus registers into a 32 bit floating point value, and see the result for every byte and word order at once, so you can spot the right one immediately.

IEEE 754ABCD, CDABWord swapFloat to hex

First register (hex)

Second register (hex)

Float value

Byte order of your device

Float value123.456

As unsigned 32 bit1123477881

As signed 32 bit1123477881

How this result was calculated

Bytes in order ABCD are assembled into a big endian 32 bit word, then read as IEEE 754 single precision.

Same registers in every byte order

Order	Float	Unsigned 32 bit
ABCD	123.456	1123477881
CDAB	−1.883367 × 10^+25	3917038326
BADC	−9.861115 × 10^+32	4131551721
DCBA	1.518500 × 10^+35	2045376066

How a Float Fits in Two Registers

32 bit IEEE 754 float = 4 bytes A B C D (sign, 8 bit exponent, 23 bit mantissa)
ABCD: register 1 = A B, register 2 = C D · CDAB: register 1 = C D, register 2 = A B

Modbus defines 16 bit registers only, so each manufacturer chooses how to place the 4 bytes of a float. Word swapped (CDAB) is very common in PLCs and meters.

Worked example

123.456 is 0x42F6E979. In ABCD order register 1 = 0x42F6 and register 2 = 0xE979; a CDAB device sends 0xE979 first.

Key insight: if a value reads as a huge number or a tiny fraction such as 1.2 × 10^−38, the byte order is wrong; the table shows all four so the sensible one stands out.