Modelling of a FSM using Verilog shows X

Question

Given the following FSM:

I am trying to model it using Verilog, and here is my trial:

module fsm (u,d,clr,clk,inc,dec,c);
  
input  u,d, clr,clk;
output reg inc,dec,c ;
// the state variables
reg y,Y; // y is the present state and Y is the next state
reg S0,S1;  

  // next state and output specifications 
 
  always @ (y,u,d) begin 
    
    case (y)
      S0 : if (u==1) begin
        Y= S1;
        inc=1;
        dec=0;
        c=1;
      end
      else if (u==0) begin
        c=0;
      end
      S1 :  if (d==1) begin
         Y= S0;
         inc=0;
         dec=1;
         c=0;
      end
      else if (d==0) begin
            c=1;
      end
   endcase 
  end 
  // state update
    always @(posedge clk,clr) begin
      if (clr) 
        y=S0;
      else 
        y=Y;
    end
    endmodule

Please let me know if my code correctly describes this FSM, and if there is any enhancement in the code please let me know it.

Here is an initial test bench code:

module tb();
 reg u,d, clr,clk;
reg inc,dec,c,y ;
  
  fsm ttb (.u(u),.d(d),.clk(clk),.clr(clr),.inc(inc),.dec(dec),.c(c));
  
     initial begin 
    $dumpfile("dump.vcd");
    $dumpvars;
  end 
  
  initial begin //{
   clk=0 ; clr=0; #10
    clk=1; u=0 ; clr=0 ; d=0; #10
    clk = 0;#10
    u=1; clk=1; #10
   
    
    
    
    
    $finish;
    
    
    
    
  end //}
  
endmodule

The testbench code shows that there is something wrong. The present state variable y shows an undefined value 'X'. I don't know why. The same for the output variables inc and dec.

What happens when you try to compile and simulate code? I would remove having the clr in your sensitivity list, unless you want it to be asynchronous. — Laburtz
– Laburtz, Commented Jul 6, 2020 at 15:28
@Laburtz yes I need it to be asynchronous. Do I leave it as it is ? Also what is the difference between this written line in my code and the '@(pos clk or clr) '? — AAA
– AAA, Commented Jul 6, 2020 at 15:43

toolic · Accepted Answer · 2020-07-06 18:11:39Z

There are several causes for the X's.

In the design, you declared S0 and S1 as reg, but then you never assigned values to them. A Verilog reg is initialized to X. An FSM is conventionally coded using constant parameter values. For example, change:

reg S0,S1;

to:

parameter S0=0, S1=1;

Your next state logic needs to assign values to Y in all cases (refer to <---):

  always @ (y,u,d) begin 
    case (y)
      S0 : if (u==1) begin
        Y= S1;
        inc=1;
        dec=0;
        c=1;
      end
      else if (u==0) begin
        c=0;
        Y= S0; // <-----
      end
      S1 :  if (d==1) begin
         Y= S0;
         inc=0;
         dec=1;
         c=0;
      end
      else if (d==0) begin
            c=1;
            Y= S1; // <-----
      end
   endcase 
  end

In your testbench, you need to assert the reset (clr=1):

  initial begin
    u=0; d=0;
    clk=0 ; clr=1; #10
    clk=1; u=0 ; d=0; #10
    clk = 0;#10
    u=1; clk=1; clr=0; #10
    $finish; 
  end

Addressing the comments on your Question regarding the asynchronous clear, you should change:

always @(posedge clk,clr) begin

to:

always @(posedge clk, posegde clr) begin

Collectives™ on Stack Overflow

Modelling of a FSM using Verilog shows X

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related