I'm not an expert on that, and I haven't found the time to try it.
However, I think you have to first obtain or determine the design rules
for your technology.
If your technology uses the FreePDK45 design rules, it looks like you
could get them at:
http://www.eda.ncsu.edu/wiki/FreePDK45:RuleDevel
https://www.eda.ncsu.edu/wiki/FreePDK45:Contents
If your technology uses the Microwind design rules, it looks like you
could get them at:
http://srv-wwwperso.insa-toulouse.fr/~sicard/microwind/index.html (e.g.,
from cmos45n.RUL when viewed in a text editor)
http://hdl.handle.net/10603/5393 (Chapter 4 of the dissertation titled
"Design and analysis of low power phase locked loop with multiple output
using VLSI technology" by Ujwala Belorkar)
If your technology uses your own laboratory equipment (photolithography,
deposition, etch, or other process systems), then you likely have to get
the design rules from your equipment specialist(s).
Then, I think you should be able to enter the design rule parameters by
hand into Electric's Technology Creation Wizard [
http://www.staticfreesoft.com/jmanual/mchap08-11.html ]. Once you have
completed the technology creation in the wizard, it looks like there is
a "Write XML" button that you could use to create an Electric .xml file
that you could share with others so that they could import the
technology into Electric and use it too. And/or Design Rules
Preferences may need to be used [
http://www.staticfreesoft.com/jmanual/mchap09-02-03.html ].
If your good a programming, you could probably write your own program in
your preferred programming language (whether that be Python, JAVA, perl,
C++, or something else) for example to convert the design rules in the
cmos45n.RUL to the Electric's xml file format [
http://www.staticfreesoft.com/jmanual/mchap08-10.html ].
The above is for the layout design rules.
As others mentioned, you also would need SPICE files (netlist, and maybe
model files) for your technology to be able to simulate the circuit.
For example, if your technology uses the fictitious 45nm ASU Predictive
model, I think the netlist is:
http://ptm.asu.edu/modelcard/LP/45nm_LP.pm
Then, you would have to check if the SPICE program you are using can
read and interrupt that netlist file (e.g. 45nm_LP.pm). If it cannot,
then you may have to convert the netlist file by hand into a format that
your SPICE program can use unless you can find a file converter online
that can do it for you.
If the components used in the netlist file are not part of the library
API of the SPICE program (e.g., for LTSPICE library API, refer to
http://ltwiki.org/index.php5?title=LTspice_Library_API ), then you
likely need to create a new spice model for the component(s) [
http://www.zen22142.zen.co.uk/ltspice/ltindex.htm ]. I'm not familiar
with how to do that, but the model for the new component might have to
be defined with .SUBCKT blocks [
http://www.simonbramble.co.uk/lt_spice/ltspice_lt_spice_tutorial_4.htm ]
using fundamental components (e.g., resistors, inductors, and
capacitors) or it may be that the SPICE source code has to be modified
if new fundamental component is needed (an example of this might be a
Memristor [ https://en.wikipedia.org/wiki/Memristor ]). If you contact
the developer of the SPICE program you are using, I would expect that
they would be able to better help you with that. So the models and
param.inc file at http://ptm.asu.edu (on Latest Models page) might need
to be converted into the file format for the model that the SPICE
program can use (e.g., might need to be a .lib file for LTSPICE [
http://denethor.wlu.ca/ltspice/#models ]).
Once you have produced the netlist and model file (.lib) for your
technology, you could combine it with the .xml to make it into your own
Electric PDK that you could share with others.
Hopefully that can help and good luck.
how can I make technology descritpion for use with Electric DRC?
--
You received this message because you are subscribed to the Google Groups "Electric
VLSI Editor" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
