How to add two signals of type std_logic

Ashok M. (a72) OP (Company: DefineView Consulting) 2021-10-08 19:01 #6841914

Rate this post

•	▲ useful ▼ not useful

Hello,

I understand that you cannot add two std_logic_vector types because a 
std_logic_vector does not have a numeric value. You have to convert 
std_logic_vector to signed/unsigned; do the math operation and convert 
result back to std_logic_vector.

But how do you add two signals of type std_logic? A std_logic type 
cannot be converted to signed/unsigned type because that requires an 
array type.

I tried the following

signal D, E, F : std_logic;
signal G : std_logic_vector (0 to 0);

begin
  Monitor: process is

  begin
   D <= '1'; E <= '0';
   --F <= D + E; --Compile Error - D and E are not numeric values.
   --F <= unsigned(D); --Compile Error - D must be an array type.
   --F <= unsigned(G); --NO
   --G <= '1'; --NO

Any suggestions?

Thanks very much.

Reply Quote Edit Move Thread sperren Anmeldepflicht aktivieren Anpinnen Delete topic Thread mit anderem zusammenführen Report post

Lothar M. (lkmiller) Moderator (Company: Titel) 2021-10-08 19:41 #6841939

Rate this post

•	▲ useful ▼ not useful

Ashok M. wrote:
> A std_logic type cannot be converted to signed/unsigned type because
> that requires an array type.
Why digging in the mud and defining own "rules"?

The definition of a signed/unsigned ist very similar to that of a 
std_logic_vector, you can check that out the free available code of 
those packages:
https://www.csee.umbc.edu/portal/help/VHDL/packages/numeric_std.vhd

Package numeric_std is
--==================================================
-- Numeric array type definitions
--==================================================
type UNSIGNED is array ( NATURAL range <> ) of STD_LOGIC;
type SIGNED is array ( NATURAL range <> ) of STD_LOGIC;



And:
https://www.csee.umbc.edu/portal/help/VHDL/packages/std_logic_1164.vhd

-------------------------------------------------------------------    
-- unconstrained array of std_logic for use in declaring signal arrays
-------------------------------------------------------------------    
TYPE std_logic_vector IS ARRAY ( NATURAL RANGE <>) OF std_logic;


> Any suggestions?
VHDL does not give room to laziness and implicity. Everything thats 
written must be written down explicitly.
Simply use the numeric_std package to cayst between std_logic_vector and 
signed/unsinged. And use the vesy same package to convert from integer 
to signed/unsinged and vice versa:
http://www.lothar-miller.de/s9y/categories/16-Numeric_Std

The same picture also exists in english mutations.

> std_logic_vector (0 to 0);
You are taking all tricky steps at once: 1. a one element array 2. with 
very strange bit order...

Reply Quote Edit Delete Report post

FPGA NOTFALLSEELSORGE Guest 2021-10-08 22:02 #6842047

Rate this post

•	▲ useful ▼ not useful

Ashok M. wrote:
> F <= D + E; --Compile Error - D and E are not numeric values.

Sure! But what do you expect? D and E are one Bit each and F is also 
only one Bit. So how should addition work? What is '1' + '1'? If you say 
2 this is "10" and will not fit in the single Bit of F.

Ashok M. wrote:
> G <= '1'; --NO

G <= "1";

Reply Quote Edit Delete Report post

FPGA NOTFALLSEELSORGE Guest 2021-10-08 22:31 #6842062

Rate this post

•	▲ useful ▼ not useful

But ... it can be done, add two std_logic and get one std_logic_vector(1 
downto 0).

library ieee; 
use ieee.std_logic_1164.all; 
use ieee.numeric_std.all;
entity tb_add_std_logic is
end tb_add_std_logic;
architecture tb of tb_add_std_logic is
constant dummy : unsigned(1 downto 0):="00";
signal D : std_logic:='0';
signal E : std_logic:='0';
signal F : std_logic_vector(1 downto 0):="00";
D <= not D after 10 ns;
E <= not E after 13 ns;
F <= std_logic_vector(dummy + (""&D) + (""&E));

Attached files:

add_std_logic.png (3.2 KB)

Reply Quote Edit Delete Report post

Ashok M. (a72) OP (Company: DefineView Consulting) 2021-10-09 19:41 #6842780

Rate this post

•	▲ useful ▼ not useful

Thanks very much for the solution. It worked. I come from Verilog 
background, and it is straight forward (loosely typed language) to add 
two single bits and assign result to a single bit (discarding the msb).

In line with your solution, I tried the following and that worked as 
well.

  variable G : std_logic_vector (1 downto 0);
  variable D, E : std_logic;
  variable uS1 : unsigned (1 downto 0);

  uS1 := unsigned'('0' & D) + unsigned'('0' & E);
  G := std_logic_vector(uS1);

Thanks.

FPGA NOTFALLSEELSORGE wrote:
> But ... it can be done, add two std_logic and get one std_logic_vector(1
> downto 0).
>
>

library ieee;
> use ieee.std_logic_1164.all;
> use ieee.numeric_std.all;
> entity tb_add_std_logic is
> end tb_add_std_logic;
> architecture tb of tb_add_std_logic is
> constant dummy : unsigned(1 downto 0):="00";
> signal D : std_logic:='0';
> signal E : std_logic:='0';
> signal F : std_logic_vector(1 downto 0):="00";
> D <= not D after 10 ns;
> E <= not E after 13 ns;
> F <= std_logic_vector(dummy + (""&D) + (""&E));

Reply Quote Edit Delete Report post

FPGA NOTFALLSEELSORGE Guest 2021-10-12 20:31 #6845804

Rate this post

•	▲ useful ▼ not useful

You can write:

F <= std_logic_vector(unsigned'('0' & D) + unsigned'('0' & E));

Reply Quote Edit Delete Report post