lipton@odin.m2c.org (Gary Lipton) (05/16/88)
I am trying to run mkcp for a CMOS process. Mkcp ("Make Crystal
Parameters") generates parameters for Crystal, a UCB timing simulator.
The input is a SPICE file describing the circuit used to characterize
the transistor you are interested in for a particular (up or down)
signal transition.
I based my SPICE file on the NMOS examples distributed for mkcp. It
describes an inverter driving a MOSFET, with a capacitor between gate
and ground. There is another capacitor between the MOSFET and VDD, or
between the MOSFET and ground, depending on whether I am characterizing
an "up" or "down" transition.
When I run mkcp on the circuit, I get the following nonsense:
Running Spice ... Current C1 is 0.00010000pf
Running Spice ... Current C1 is 0.01000000pf
Output signal oscillates around 1.860000 volts.
Output signal oscillates around 2.060000 volts.
Output signal oscillates around 2.270000 volts.
Running Spice ... Current C1 is 0.03000000pf
Output signal oscillates around 1.860000 volts.
Output signal oscillates around 2.060000 volts.
Output signal oscillates around 2.270000 volts.
Running Spice ... Current C1 is 0.10000000pf
Output signal oscillates around 1.860000 volts.
And so on.
When I look at the SPICE output, I don't see any oscillations.
I just see monotonically increasing signals.
Questions:
1. Has anyone out there used mkcp for CMOS (or anything else)?
2. What are appropriate circuits to model MOSFETS?
3. What can I do about the oscillation messages?
Thanks.
Gary Lipton - Massachusetts Microelectronics Center ## ## 2 #####
lipton@m2c.m2c.org (CSNET) ***** Disclaimer: KBISFB ### ### ##
lipton%m2c.org@relay.cs.net (Internet) ************ ## # ## ##
{harvard,bu-cs,frog,ulowell}!m2c!lipton (UUCP) **** ## ## #####
Gary Lipton - Massachusetts Microelectronics Center ## ## 2 #####
lipton@m2c.m2c.org (CSNET) ***** Disclaimer: KBISFB ### ### ##
lipton%m2c.org@relay.cs.net (Internet) ************ ## # ## ##
{harvard,bu-cs,frog,ulowell}!m2c!lipton (UUCP) **** ## ## #####