Re: slew filter arguments

owner-verilog-ams@server.eda.org wrote on 21-03-2008 19:01:08:

> Xavier Bestel wrote:
>
> >Hi,
> >
> >just a little remark about the slew filter: when absent, the negative
> >slope is said to be the "inverse" of the positive slope, whereas it
> >should probably be the "opposite".
> >
> >
>
> Hi
>
> I have another remark to add to the slew description. In the 2.3 draft
> LRM it says
>
> "If the rate of change of expr is less than the specified maximum slew
> rates, slew() returns the value of expr."
>
> My literal interpretation of this is that if the signal is slewed, then
> as soon as the input expression slope falls below the maximum slew rate,
> the slewed output will jump from its slewed value to that of the imput
> expression.

That can't happen as such a jump would mean a derivative with a value
higher than the maximum slew rate.

> To me, this isn't what I'd expect from slew. I would expect that the
> output of slew(), once it is slewed, to continue rising/falling at the
> slew rate until it catches up with the input expression (either because
> the input flattens out, or because the input drops/rises to rejoin the
> slewed output). Thus for a slewed signal, it is the delta between the
> slewed output and the input expression that counts.

As the jumps cannot be present, this is exactly what happens. As the slew
rate clamps the absolute value of the first derivative to the maximum slew
rate, the resulting behavior will be exactly as you describe.

Cheers,
Marq
-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.