The registers using the "32-bit floating-point" data format are encoded using the "binary32" encoding defined in IEEE 754 (also known as the "single precision floating-point format").
The lower 16 bits of floating-point numbers are placed in the Modbus registers listed in the table, while the highest 16 bits are placed in the register with number/address+1, as specified in the Open Modbus TCP specification version 1.0. This is also known as the "small end" or "Modicon" character sequence.
Despite this specification, some Modbus master stations may expect a "big end" word sequence (with the most significant word first). In this case, you must select the "Word Swap" floating-point format for the Modbus registers of the device in the Modbus master.
Occasionally, when reading measurement data registers using incorrect floating-point format settings, seemingly correct values may be obtained. It is strongly recommended to first try reading register number 7938 (floating point test) to see if you have correctly configured the floating point format on the Modbus host system.