converting real number to single precision floating point numbe

[gauravbhand]

hello,
i want to convert a real number to single precision floating point number without using any float_pkg provided by David Bishop. The package has some errors when i synthesize which has not been resolved yet. Please suggest me another better way to do the job

Thanks in advance,
Gaurav Bhand
email: gauravbhand@gmail.com

 

[gopi_vlsi]

use IEE 754 The IEEE Standard for Floating-Point Arithmetic (IEEE 754) is a technical standard for floating-point computation

 

[gauravbhand]

use IEE 754 The IEEE Standard for Floating-Point Arithmetic (IEEE 754) is a technical standard for floating-point computation

Hello gopi_vlsi,
Thanks for your reply. I know the IEEE 754 standard for FP arithmetic, but i want some help regarding the vhdl code for converting real to floating point

Thanks in advance,
Gaurav Bhand

 

[FvM]

"convert a real number to single precision floating point" only applies to compile time constant generation and hasn't to do with synthesis. It's achieved by the VHDL float package without any problems, I think. Which errors do you have in mind?

For synthesis of float arithmetic, you'll usually refer to vendor IP.

 

[gauravbhand]

"convert a real number to single precision floating point" only applies to compile time constant generation and hasn't to do with synthesis. It's achieved by the VHDL float package without any problems, I think. Which errors do you have in mind?

For synthesis of float arithmetic, you'll usually refer to vendor IP.

Hi FvM
thanks for your reply. I am trying to use fixed_pkg_c provided in David Bishop website, i am using Xilinx 9.2 ise and i followed the process mentioned in the documentation. Yet, i get the following error

Error Xst:2683 - Unexpected error found while building hierarchy.

I dont understand where the problem is?
Can you suggest a proper step by step procedure to convert real number (input given as binary for ex. 9.2 is 1001.1 etc) and output in single precision floating point format. That is basically i want to write a code of a real number given as binary input to a floating point converter.

Thanks in advance
Gaurav Bhand