The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design Languages >> Verilog-AMS >> Error message from divider model without HiddenSta
https://designers-guide.org/forum/YaBB.pl?num=1128094710

Message started by tumeda on Sep 30^th, 2005, 8:38am

Title: Error message from divider model without HiddenSta
Post by tumeda on Sep 30^th, 2005, 8:38am

Dear Ken Kundert

I reallly appreciate your answer for my previous questions. Now I tried to directly use the behavioral model codes of divider without hidden state. Unfortunately, there is an error message.
###
Error found by spectre during SpectreHDL compile.
…...
line 40: Error: (Archaic Syntax) Does not comply with the Accellera
Verilog-AMS 2.0 Standard and beyond. To comply with the current
standard, delete the null statement. Null statements (also known as
empty statements and coded as just `;') can be used only in case or
event control statements.
###
After then, I deleted the semicolon ( ; ) in the line 40 and tried again. But there is still new error message:
###
Error found by spectre during SpectreHDL compile.
……
line 41: Error: analog operator `ddt' embedded in conditionally
executed statement or expression.
###

Now, I am confused about that, what's wrong?
My icfb version is 5.1.0 02/01/2005.

The following are only some lines of your codes.

Code:

....
analog begin
vmax = max(abs(vdd),abs(vss));
g = 5*ln(n);
@(cross(V(pin,nin) - thresh, 0));

// hold capacitor with capacitance tt
I(vg,hold) <+ tt*ddt(V(vg,hold));

// clamp hold capacitor during dc analysis
if (analysis("static"))
I(vg,hold) <+ V(vg,hold);
...

Title: Re: Error message from divider model without Hidde
Post by Andrew Beckett on Oct 2^nd, 2005, 12:38am

A whole bunch of extra checks were put into the verilog-a parser around the time of the version you're using - but some of these were reporting things that aren't really problems. These were fixed around March 2005 - so in IC5141 and MMSIM60 versions of spectre after that, the messages are no longer given.

I just tried the models from the hidden state paper in IC5141 USR2 and MMSIM60 ISR 10, and the warning is not given.

So use a later subversion, and the problem should go away.

Regards,

Andrew.