[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

Re: gEDA-user: one schmatic ==> two netlist



This is an interesting conumdrum.  The SPICE netlister needs the DEVICE
attribute in order to know what type of SPICE card to spit out.  I don't
know why the Verilog DEVICE property is the instance name since I didn't
write the Verilog netlister.  OTOH, I don't know what property you would
use instead.  Changing this would require a total re-evaluation of the gEDA
attribute architecture (which maybe it needs, but is a big, big effort).

The bigger question is:  Why are you using the same schematic for analog
and digital simulation?  Most board designers (also in the professional
world) use separate schematics for this kind of thing.  THat is, if you
have a mixed signal circuit, you create a separate analog schematic with no
digital stuff on it for SPICE simulation.  You keep a main schematic for
netlisting and layout.

Ooops, I just realized that you are doing a chip.  Mentor Graphics is
currently hyping a design environment which allows you to do seamless
integration of SPICE, Verilog, and Verilog-AMS (or something like that). 
Cadence and Synopsys must have products like that too, although I don't
follow them as much.  Therefore, this type of thing is more cutting-edge
than gEDA.  If you're spinning mixed signal chips, maybe you need to buy a
professional CAD package . . . .

My thoughts/opinions only.

mit freundlichen Gruessen aus Amerika,

Stuart

Quoting Familie Kaiser <Die.Kaisers@t-online.de>:

> 
> Hi,
> 
> does anybody had the problem to create two different netlists from one 
> schmatic and run into problems with the device property?
> 
> 
> Concrete problem:
> I am designing a mixed signal ASIC. Digital simulation is done with
> Verilog, 
> analog simulation and LVS with Spice.
> 
> The device property with Verilog is the package instance and with Spice
> the 
> device property tells the netlister what spice primitve to use.
> 
> 
> Has anybody a solution to this problem?
> 
> 
> Peter
> 
> 
> 
> -- 
> 
> ----------------------------------------
> die.kaisers@t-online.de
>                                             oder
> dh0may@darc.de
>                                             oder
> http://www.darc.de/c/12
> ---------------------------------------
>