Both clock edges

[Steve]

I'm designing a simple VHDL system, which is driven by a clock. Is it bad
practice to use both clock edges in my system.

The system I'm working on needs to output data, which must be valid while
the clock is high. So I was thinking I could use the falling edge to set
everything up. It must also read data, which is also valid while the clock
is high. To do this, I could use the rising edge.

For example:

p1 : process(CLK)
begin
if rising_edge(CLK) then
blah blah blah (read data)
end if;
end process p1;


p2 : process(CLK)
begin
if falling_edge(CLK) then
blah blah blah (write data)
end if;
end process p2;


This would make my job easier, but is it a good way to do things? Most
examples I have seen tend to use one edge of a clock.

 

[Pieter Hulshoff]

I'm designing a simple VHDL system, which is driven by a clock. Is it bad
practice to use both clock edges in my system.

Unless it's truly needed, it's best to avoid it. It usually creates more
problems than it's solving. For your problem: just precalculate the needed
values in the previous clock cycle (pipelining your design).

Regards,

Pieter Hulshoff

 

[rickman]

I'm designing a simple VHDL system, which is driven by a clock. Is it bad
practice to use both clock edges in my system.

The system I'm working on needs to output data, which must be valid while
the clock is high. So I was thinking I could use the falling edge to set
everything up. It must also read data, which is also valid while the clock
is high. To do this, I could use the rising edge.

For example:

p1 : process(CLK)
begin
if rising_edge(CLK) then
blah blah blah (read data)
end if;
end process p1;

p2 : process(CLK)
begin
if falling_edge(CLK) then
blah blah blah (write data)
end if;
end process p2;

This would make my job easier, but is it a good way to do things? Most
examples I have seen tend to use one edge of a clock.

Using both edges of the clock is not necessarily a bad practice. It is
definitely bad for the same signal or variable since it will not
synthesize. But the way you are using it should be ok... except for one
thing. The way you have defined your IO signals, you should be clocking
both input and output on the falling edge. If your input data is is
valid while the clock is high, then you would have *no* setup time
clocking on the rising edge. Typically FPGAs have a zero hold time so
it would be ok to clock on the falling edge.

I vaguely remember working on a system spec'd by the government this
way. I thought it odd since it was very unclear how they really wanted
you to do the job. In the end I took a look at existing systems and
determined that outputs were clocked on one edge and inputs were clocked
on the other edge. This clocking compensated for a few ns of skew in
the cable. The spec said the inputs would be valid while the clock was
high, but really I needed to analyze the situation. I suggest that you
do the same.

--

Rick "rickman" Collins

(e-mail address removed)
Ignore the reply address. To email me use the above address with the XY
removed.

Arius - A Signal Processing Solutions Company
Specializing in DSP and FPGA design URL http://www.arius.com
4 King Ave 301-682-7772 Voice
Frederick, MD 21701-3110 301-682-7666 FAX