topic Re: long int to float in 8-bit Microcontrollers

long int to float

juancarlossuare — Wed, 06 Feb 2013 03:52:27 GMT

Hello.

¿How to convert from long int to float?

Re: long int to float

juancarlossuare — Thu, 14 Feb 2013 22:19:47 GMT

We can use type cast to convert long int to float. For example:

volatile long int l_val=5;

volatile float f_result,f_val=1.23;

f_result=(float)l_val/f_val;

In above code " l_val " computes as float.

Re: long int to float

bigmac — Fri, 15 Feb 2013 09:44:44 GMT

Hello,

You might want to see if you get approximately the correct value when I_val is initialised with a value that exceeds the 16-bit numeric range, e.g. long int I_val = 123456L; Of course, the conversion to float will usually result in some loss of precision, maybe with accuracy limited to about 5 digits.

Regards,

Mac

Re: long int to float

kef — Fri, 15 Feb 2013 09:59:38 GMT

Mac,

conversion from int to float is exact provided integer can fit FP mantissa. For single precision float, with its 24bit mantissa, no precision is lost converting long int's in the range -16777215 to +16777215.

Re: long int to float

Monica — Mon, 18 Feb 2013 16:07:43 GMT

Juan, I invite you to mark the best answer as correct, this way other people experiencing the same problem can identify the solution.

If the answers here have not helped, let us know!

Regards!

Re: long int to float

juancarlossuare — Mon, 18 Feb 2013 23:06:34 GMT

Hello Monica.

This answer was given in a service request of freescale.

We can use type cast to convert long int to float. For example:

volatile long int l_val=5;

volatile float f_result,f_val=1.23;

f_result=(float)l_val/f_val;

In above code " l_val " computes as float.