EmbDev.net

Forum: FPGA, VHDL & Verilog multiplier a matrice

von Mohammad M. (mohammadmother)

2014-10-25 08:35

Rate this post

0	▲ useful ▼ not useful

Hi every one
my working is to implement a circuit in FPGA ,i get the admittance 
matric of circuit that be multiply it into current vector(value of 
current vector changing with control signal state)
when the clock rise(posedge), elements of inverse admittance matric 
(G11,G12,G21,G22) multiply into vector of current and when the clock 
down(negedge), the vector of current be updated.
the code has been written in verilog.
the problems that ISE gives is that:

Line 29: r1 is not a constant
Line 30: y11 is not a constant
Line 31: y22 is not a constant
Line 21: Module <scircuit> ignored due to previous errors.

`timescale 1ns / 1ps
//////////////////////////////////////////////////////////////////////////////////
// Company: 
// Engineer: 
// 
// Create Date:    17:21:30 10/24/2014 
// Design Name: 
// Module Name:    scircuit 
// Project Name: 
// Target Devices: 
// Tool versions: 
// Description: 
//
// Dependencies: 
//
// Revision: 
// Revision 0.01 - File Created
// Additional Comments: 
//
//////////////////////////////////////////////////////////////////////////////////
module scircuit(
    input control,
    input clk,
    input reset,
    output reg [15:0] v1,v2
    );

reg  [15:0] r1=1'b1, G1=1'b1 , R=4'b1010, vdc=8'b100100;
reg signed [15:0] y11=(1/r1)+G1 ,y12=-G1 ,y21=-G1 ,y22=R+G1;
reg signed [15:0] yy=(1/((y11*y22)-(y12*y21)));
reg signed [15:0] G11=(y22/yy) ,G12=(-y12/yy) ,G21=(-y21/yy) ,G22=(y11/yy);
reg signed [15:0] j1=0000000000000000;
reg signed [15:0] j2=0000000000000000;
reg signed [15:0] v11,v22,i1,i2;
always @(posedge clk)
  if (reset == 1)
       begin     
          v1 <= 16'b0000000000000000;
    v2 <= 16'b0000000000000000;
       end  //end if(reset ==1)
   else
       begin
        i1 <=(vdc/r1)+j1;
        i2 <= -j1;
        v11 <= (G11*i1) + (G12*i2) ;
        v22 <= (G21*i2) + (G22*i2) ;
      v1 <= v11;
      v2 <= v22;
      end
      always @(negedge clk)
          case(control)
           1'b1:
            j1 <=-G1*(v1-v2);
         1'b0:
               j1 <= G1*(v1-v2);
                
   endcase

endmodule

2014-10-25 22:34: Edited by User

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

Re: multiplier a matrice

von Duke Scarring (Guest)

2014-10-27 08:05

Rate this post

0	▲ useful ▼ not useful

Mohammad Mothermohammad wrote:
> i1 <=(vdc/r1)+j1;
> Line 29: r1 is not a constant

ISE (and I belive other synthesis tools too) support only division by 
2**n.

You can change your divisior to fullfill this constrain or use IP core 
for division or write your own code/FSM for division.

Duke

Report post Edit Delete Quote selected text Reply Reply with quote

Re: multiplier a matrice

von Mohammad M. (mohammadmother)

2014-10-27 13:46

Rate this post

0	▲ useful ▼ not useful

thanks Duke
i change my code i want to define this parameter ,as input .
when program be run the values(G11,G12,G21,G22) get take from memory
the problem is how write code to access elements from memory
 the new code is:


`timescale 1ns / 1ps
//////////////////////////////////////////////////////////////////////////////////
// Company: 
// Engineer: 
// 
// Create Date:    17:21:30 10/24/2014 
// Design Name: 
// Module Name:    scircuit 
// Project Name: 
// Target Devices: 
// Tool versions: 
// Description: 
//
// Dependencies: 
//
// Revision: 
// Revision 0.01 - File Created
// Additional Comments: 
//
//////////////////////////////////////////////////////////////////////////////////
module scircuit(
    input control,
    input clk,
    input reset,
    input G11,G12,G21,G22;
    output reg [15:0] v1,v2
    );

reg signed [15:0]  j1,j2;
reg signed [15:0] v11,v22,i1,i2;
always @(posedge clk)
  if (reset == 1)
       begin     
     v1 <= 16'b0000000000000000;
    v2 <= 16'b0000000000000000;
     end  //end if(reset ==1)
   else
       begin
        i1 <=(vdc/r1)+j1;
        i2 <= -j1;
        v11 <= (G11*i1) + (G12*i2) ;
        v22 <= (G21*i2) + (G22*i2) ;
      v1 <= v11;
      v2 <= v22;
      end
      always @(negedge clk)
          case(control)
           1'b1:
            j1 <=-G1*(v11-v22);
         1'b0:
              j1 <= G1*(v11-v22);
                
   endcase

endmodule

Report post Edit Delete Quote selected text Reply Reply with quote

Re: multiplier a matrice

von Duke Scarring (Guest)

2014-10-27 14:43

Rate this post

1	▲ useful ▼ not useful

Mohammad Mothermohammad wrote:
> the problem is how write code to access elements from memory
Take a look in the synthesis guide of your tool.

For Xilinx you can use the XST user guide (UG627). There you can find 
a lot of (working) coding examples.

Duke

Report post Edit Delete Quote selected text Reply Reply with quote

Watch this topic |

Disable multi-page view

Reply

Entering an e-mail address is optional. If you want to receive reply notifications by e-mail, please log in.

Rules — please read before posting

Post long source code as attachment, not in the text
Posting advertisements is forbidden.

Formatting options

[c]C code[/c]
[avrasm]AVR assembler code[/avrasm]
[vhdl]VHDL code[/vhdl]
[code]code in other languages, ASCII drawings[/code]
[math]formula (LaTeX syntax)[/math]

Contact – Data privacy information – Advertising on EmbDev.net