Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
14 years ago

Very simple Verilog code, but not working on DE2

I am running this simple code, pushing KEY[3] works fine (decrement), but when I push KEY[1] the 7-seg changes with both pushing and releasing and display some random number on the 7-seg.

hex_7seg is just a module to display count on the 7-seg. It works fine for other projects.

Looks like the problem is with the if statement. Since when I test KEY[3] instead of KEY[1], the problem will be when I push KEY[3]

Any help is appreciated

module d1_7seg(KEY, HEX0);

input [3:0] KEY;

output [6:0] HEX0;

reg [3:0] count;

always @ ( negedge KEY[3] or negedge KEY[1])

begin

if(KEY[1] == 0)

begin

count <= count + 1;

end

else

begin

count <= count - 1;

end

end

hex_7seg (count, HEX0);

endmodule

3 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    14 years ago

    As a simple fact, you can't have two different edge sensitive triggers for a single counter.

    The code actually defines an edge sensitive (synchronous) trigger for KEY[3] and an asynchronous trigger for KEY[1], which doesn't work for a counter. The combination of negedge KEY[1] and if (KEY[1] == 0) is a specific Verilog syntax, that's e.g. used for asynchronous resets. You typically should get a serious synthesis warning like "count[x] doesn't hold it's state between clock edges" indicating, that the counter doesn't work.

    Facing this fact, you obviously need to use a different method to detect keypress events. The common method is synchronous edge detection utilizing a system clock and comparing present and previous state of a signal.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    14 years ago

    First, thanks. You are right. It works now.

    However, I thought that

    always @ (negedge KEY[3] or negedge KEY[1] )

    if (KEY[1] == 0)

    is similar to

    always @ (posedge clock or negedge RST)

    if (RST == 0)

    which works fine

    I guess I was wrong ...

    Thanks again

    --Mokhtar
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    14 years ago

    The RST example creates asynchronous logic, as said. You'll never be able to e.g count reset events under the (RST == 0) condition.