Converted signed 16 Bit to unsigned?

[Steffen Koepf]

Hello,

I have a AD Converter that delivers a 16 Bit signed value 8000 - 0 - 7FFF

I want to convert this to unsigned 16 Bit from 0 - 8000 - FFFF

by

signal RecRegA : std_logic_vector (15 downto 0);

signal RecRegA1 : std_logic_vector (15 downto 0);


RecRegA1 <= std_logic_vector(unsigned(RecRegA) + 32768);


This works in the lower 3/4 range. But as soon as the result exceeds 7FFF
the result is unpredictable. Why?

Thanks in advance,

Steffen

 

[Gabor]

Steffen Koepf wrote:

Hello,

I have a AD Converter that delivers a 16 Bit signed value 8000 - 0 - 7FFF

I want to convert this to unsigned 16 Bit from 0 - 8000 - FFFF

by

signal RecRegA : std_logic_vector (15 downto 0);

signal RecRegA1 : std_logic_vector (15 downto 0);


RecRegA1 <= std_logic_vector(unsigned(RecRegA) + 32768);


This works in the lower 3/4 range. But as soon as the result exceeds 7FFF
the result is unpredictable. Why?

Thanks in advance,

Steffen


That's a lot of conversions. All you really need to do is invert the

MSB of the first vector. What libraries are you using?

-- Gabor