Prev: Pipelining
Next: fixed point precision in matlab
From: kungcoccos on 24 Jun 2008 04:42
Hi,

I need some information of how to convert fixed point numbers to floating
point. Is there any standard way to do this kind of operation? I will
implement it on a FPGA.. so, please do not refer to any functions in matlab
etc..

Regards

Carl H.


From: glen herrmannsfeldt on 24 Jun 2008 08:06
kungcoccos wrote:

> I need some information of how to convert fixed point numbers to floating
> point. Is there any standard way to do this kind of operation? I will
> implement it on a FPGA.. so, please do not refer to any functions in matlab
> etc..

I don't know about standard, but you need to know the number of high
order zeros to set the exponent and shift the data. For FPGA that
usually means a barrel shifter after you count the zeros with a
priority encoder. You need to know the floating point format
you want to generate.

-- glen

From: Vladimir Vassilevsky on 24 Jun 2008 10:10


kungcoccos wrote:
> Hi,
>
> I need some information of how to convert fixed point numbers to floating
> point. Is there any standard way to do this kind of operation? I will
> implement it on a FPGA.. so, please do not refer to any functions in matlab
> etc..

STFW "IEEE-754"


Vladimir Vassilevsky
DSP and Mixed Signal Design Consultant
http://www.abvolt.com
 | 
Pages: 1
Prev: Pipelining
Next: fixed point precision in matlab