floating point calculations on fixed point dsp

[mehboob_iiui]

I am learning to use my c5416 dsp, however i found out that it is a fixed point dsp, so how do i perform floating point calculations on it....

 

[vadkudr]

Only by emulation.
If you use C, you can check compile properties - there is usually fixed point emulation on.
However it significantly slows down processing.
If using floating point is critical for your task, you have to choose another chip.

 

[brmadhukar]

you may have to write your own floating point libraries. That is quite easy but takes time.

 

[naresh850]

example code is available with processor's devlopment as all the processor have.

I m using ADI fixed point processor and that routine is availabe with VDSP++

 

[Mr.MEB]

Your time critical dsp calculations should be in fixed point (interupt, sampling), so you need convert your floating point parameters in fixed point before.
You can make floating code calculations in your non-time critica code sections.l