• Previous message: cfk: "Re: mosis cell library"
• In reply to: Jeff Sondeen: "Re: mosis cell library"
• Next in thread: cfk: "Re: mosis cell library"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On the good news side, I was able to compile tclspice yesterday as described
in the README in the /home/cad/tclspice/src/tcl directory. The ./configure
script needs to arguments :
"./configure --enable-tcl --enable-experimental --enable-xspice --disable-sh
ared", then a "make tcl", followed by a "make tcl_install" (yes that one is
a little different from magic and irsim).

At this point, I can starup magic and I get a blank BLT plot along with the
magic window from "magic -w". Magic is reading /usr/lib/spice/pkgIndex.tcl
and runnings its LoadSpice procedure. Next, from the magic prompt, an
invocation ";load /usr/lib/libspice" loads the complete package. At this
point, it is possible to say

;spice::version
;spice::source test.cir
;spice::run
;spice::reset
;spice::asciiplot out
or even
;spice::shell kedit test.cir

all from the magic prompt.

I can run mesa.cir from tclspice/tests/mesa and get a transfer function
curve. The plot doesnt seem to work and tclspice is prone to segfault if it
gets confused with a command such as ";spice::help", but at least it is very
interesting to run this from inside magic.

This is all preparation on my part for the day when I can load an inverter,
nand, nor, xor or dff from a library (perhaps the mosis library, at least I
am hoping so), run irsim for the switch level simulator (that one allready
works fine with the library from http://www.rsl.ukans.edu/~mlinhart/magic ,
with those files, I can load invf100.mag do a:

;startirsim
;watchnode A1
;watchnode O
;graphnode A1
;graphnode O 1
;l A1
;s
;h A1
;s

and see the switch level simulator switch just fine. It also works with the
nanf201 nand gate and the nor gate, I havent gotten to the flip flops yet.
Unfortunately, the .spice file created by ;exttospice even when surrounded
by a test.cir with a power supply and a capacitive load dont spice yet, they
irsim, they just dont spice.

One of the reasons I was wanting to use the mosis library was to be able to
have a standard of comparision for irsim and spice from within magic that
could be independently corroborated from a known functioning set of cells
where the only variable is now the setup of magic.

Charles

• Next message: cfk: "Re: mosis cell library"
• Previous message: cfk: "Re: mosis cell library"
• In reply to: Jeff Sondeen: "Re: mosis cell library"
• Next in thread: cfk: "Re: mosis cell library"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]