You should consider the possibility of using C language instead of Assembler. A C compiler / linker tool set should come with a set of math library routines that support the single precision floating point format. Converting between the long integer format and the float format would be as easy as an assignment statement.

All that said it should be relatively easy to write a conversion routine. Here are some things to consider looking at to give some guidance...

- Check if you have a possibility of using a trial version of an MSP430 C compiler to prepare a long to float conversion and then look at the assembler language generated by the compiler.

- Look around for an actual MSP430 floating point math library on the web and use that. Obviously for your application you would be looking for an assembly language library.

- You could write some code in C on a PC that has long to float conversions in it. You could then inspect the binary values stored in the converted variables to get a full understanding of storage format for the numbers.

Michael Karas