Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von YouseKalack (Guest)

2019-11-14 12:20

Attached files:

ffff.PNG
37 KB

Rate this post

•	▲ useful ▼ not useful

Hello i've got a matter with my simulation, the result of the sub 
operation is not the result expected.

My code :

I) The dataflow code to describe cell.

library ieee;
use ieee.std_logic_1164.all;
use ieee.std_logic_unsigned.all;
use ieee.numeric_std.all;
entity cell is 
port(Ai,Bi,Ci : in std_logic;
     S : in std_logic_vector (1 downto 0);
     Co,Go : out std_logic);
end cell;
architecture dflow of cell is
signal oper : std_logic_vector(1 downto 0);
signal tabc : std_logic_vector(2 downto 0);
with tabc select oper <=
                          '0' & Ai + Bi  when "001",
                          '0' & Ai + Bi when "010" ,
                          '0' & Ai - '1' when "011" ,
                    '0' & Ai + '1' when "100" ,
                          '0' & Ai - Bi when "101" ,
                          '0' & Ai + Bi + '1' when "110",
                          '0' & Ai when others;
Co <= oper(1);
Go <= oper(0);
tabc <= Ci&S;
end architecture;


II) The structural code to link 4 cells to implement the arithmetic 
operator

library ieee;
use ieee.std_logic_1164.all,ieee.std_logic_unsigned.all;
entity operator is
port(A,B : in std_logic_vector(3 downto 0);
     S : in std_logic_vector(1 downto 0);
     cin : in std_logic;
     G : out std_logic_vector(3 downto 0);
     cout : out std_logic);
end entity;
architecture rtl of operator is
component cell is 
port(Ai,Bi,Ci : in std_logic;
     S : in std_logic_vector (1 downto 0);
     Co,Go : out std_logic);
end component;
signal C : std_logic_vector(3 downto 1);
cell0 : cell port map(Ai=>A(0),Bi=>B(0),Ci=>cin,S=>S,Co =>C(1),Go => G(0));
cell1 : cell port map(Ai=>A(1),Bi=>B(1),Ci=>C(1),S=>S,Co=>C(2),Go => G(1));
cell2 : cell port map(Ai=>A(2),Bi=>B(2),Ci=>C(2),S=>S,Co=>C(3),Go => G(2));
cell3 : cell port map(Ai=>A(3),Bi=>B(3),Ci=>C(3),S=>S,Co =>cout,Go =>G(3));
end architecture;

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

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Lothar M. (Company: Titel) (lkmiller) (Moderator)

2019-11-14 13:12

Rate this post

•	▲ useful ▼ not useful

YouseKalack wrote:
> the result of the sub operation is not the result expected.
What do you expect and wht do you get instead?

> My code
Pls use the [vhdl] tags further on to wrap your code!
See the "Formatting options" above each edit box...

> use ieee.std_logic_unsigned.all;
> use ieee.numeric_std.all;
Never ever both together!
Otherwise you will encounter strange problems due to multiple type 
definitions.

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von YouseKalack (Guest)

2019-11-14 13:32

Attached files:

sumuuu.PNG
1.8 KB

Rate this post

•	▲ useful ▼ not useful

The result expect would be "8" when S = "11" but the result is "2" as 
you can see in the simulation ....

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Yousekalack (Guest)

2019-11-14 13:43

Rate this post

•	▲ useful ▼ not useful

My code :

I) The dataflow code to describe cell.

library ieee;
use ieee.std_logic_1164.all;
use ieee.std_logic_unsigned.all;
use ieee.numeric_std.all;
entity cell is 
port(Ai,Bi,Ci : in std_logic;
     S : in std_logic_vector (1 downto 0);
     Co,Go : out std_logic);
end cell;
architecture dflow of cell is
signal oper : std_logic_vector(1 downto 0);
signal tabc : std_logic_vector(2 downto 0);
with tabc select oper <=
                          '0' & Ai + Bi  when "001",
                          '0' & Ai + Bi when "010" ,
                          '0' & Ai - '1' when "011" ,
                    '0' & Ai + '1' when "100" ,
                          '0' & Ai - Bi when "101" ,
                          '0' & Ai + Bi + '1' when "110",
                          '0' & Ai when others;
Co <= oper(1);
Go <= oper(0);
tabc <= Ci&S;
end architecture;

II) The structural code to link 4 cells to implement the arithmetic
operator

library ieee;
use ieee.std_logic_1164.all,ieee.std_logic_unsigned.all;
entity operator is
port(A,B : in std_logic_vector(3 downto 0);
     S : in std_logic_vector(1 downto 0);
     cin : in std_logic;
     G : out std_logic_vector(3 downto 0);
     cout : out std_logic);
end entity;
architecture rtl of operator is
component cell is 
port(Ai,Bi,Ci : in std_logic;
     S : in std_logic_vector (1 downto 0);
     Co,Go : out std_logic);
end component;
signal C : std_logic_vector(3 downto 1);
cell0 : cell port map(Ai=>A(0),Bi=>B(0),Ci=>cin,S=>S,Co =>C(1),Go => G(0));
cell1 : cell port map(Ai=>A(1),Bi=>B(1),Ci=>C(1),S=>S,Co=>C(2),Go => G(1));
cell2 : cell port map(Ai=>A(2),Bi=>B(2),Ci=>C(2),S=>S,Co=>C(3),Go => G(2));
cell3 : cell port map(Ai=>A(3),Bi=>B(3),Ci=>C(3),S=>S,Co =>cout,Go =>G(3));
end architecture;

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Lothar M. (Company: Titel) (lkmiller) (Moderator)

2019-11-14 13:46

Rate this post

•	▲ useful ▼ not useful

YouseKalack wrote:
> as you can see in the simulation ....
So setup a second test bench for your "cell" and have a look wether its 
working as desired.

> I) The dataflow code to describe cell.
By looking at the picture bottom left corner a thought pops up in my 
mind: shouldn't that "cell" also be described as a logic/structural 
expression instead as a functional data flow description? And shouldn't 
the CI be used for any operation in the "cell"?

To clear the view: the top left hand table is not related to the "cell" 
driectly. That table shows the expected behaviour of the top level 
module "operator".

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Achim S. (Guest)

2019-11-14 13:54

Rate this post

•	▲ useful ▼ not useful

in your description of "cell" you never set the Carry-output. Oper(1) is 
set to 0 for all possible combinations of input signals. Your arithmetic 
won't work without correct Carry-Signals.

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Yousekalack (Guest)

2019-11-14 14:00

Rate this post

•	▲ useful ▼ not useful

Yes the test work for one cell but only sub is not working good.. i 
think the problem is when A = '0' and A - 1 = ???

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Achim S. (Guest)

2019-11-14 14:04

Rate this post

•	▲ useful ▼ not useful

Yousekalack wrote:
> Yes the test work for one cell but only sub is not working good.. i
> think the problem is when A = '0' and A - 1 = ???

No, for a single cell your code is also not working correctly.

Perform a simulation for a single cell. What are the correct outputs for 
A=0, Cin=0 and opeartion A-1? What does your simulation show instead?

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Huynh J. (yousekalack)

2019-11-14 14:17

Rate this post

•	▲ useful ▼ not useful

This is the test for one cell with Ai = 0, Ci = 0 and S = '11'

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Huynh J. (yousekalack)

2019-11-14 14:18

Attached files:

simu2.PNG
2.3 KB

Rate this post

•	▲ useful ▼ not useful

Oper(1) is the carry out from all operations, i put'0' & but the '0' 
will be replace if there is a carry

Report post Edit Delete Quote selected text Reply Reply with quote

Re: Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded

von Achim S. (Guest)

2019-11-14 16:00

Rate this post

•	▲ useful ▼ not useful

Huynh J. wrote:
> Oper(1) is the carry out from all operations, i put'0' & but the '0'
> will be replace if there is a carry

ok you're right, my mistake. "&" and "-" have identical priority and are 
evaluted from left to right. So

   '0' & Ai - '1'

turns into

  ('0' & Ai) - '1'

and then  the higer bit (i.e. the carry) can indeed be set after the 
operation

Report post Edit Delete Quote selected text Reply Reply with quote

EmbDev.net

Forum: FPGA, VHDL & Verilog Arithmetic operator (Sub/Add/+1/-1) N bits Cascaded