IQ to Float conversion in HVACI_Scalar(f28335)

Andrey.B.

Good day everyone

I want to convert HVACI_Scalar project from fixed poin to float and met with the error "expression must have integral type" on SWGENMF_MACRO.

It looks like the problem is in bit shifting " >>"

I tried to use algarithm described in SPRC990 (IQmath user`s guide) But it doesn`t work.

        svgen_mf1.Gain = (int16)(32768.0L*vhz1.VoltOut);
        svgen_mf1.Freq = (int16)(32768.0L*SpeedRef);
         SVGENMF_MACRO(svgen_mf1);

Thank you for any help

Andrey

0 Manish Bhardwaj over 10 years ago

Andrey,

We use left and right shift as for fixed point IQmath this is the most optimal way to do multiply/divide by 2.

In IQmath library v15.c or 160 we have a new function which you cna use instead of the shift

v.Ta=_IQdiv2(_IQ(1.0)-v.dx-v.dy);

instead of the shift and it should compile fine..

I will report an enhancement to the existing solution,

0 Andrey.B. over 10 years ago in reply to Manish Bhardwaj

Hello Manish ,
Thank you very mach for your help.

C2000 microcontrollers forum