Page 217 - Troubleshooting Analog Circuits

204 G. More on Spice

boards, but I like them because they often offer a greater insight and understanding as
to what’s really going on, so if you survive their problems, you are smart enough to
keep out of other kinds of trouble. But, that’s just a bunch of philosophical stuff.
The thing that makes me nervous about SPICE is that it was largely designed by a
group of grad students (Laurence Nagel and others at Berkeley) back in ‘73. Now,
when you find a problem, a discrepancy, a glitch, a flaw, an error that seems to be
built into SPICE, can you go back to the people who designed it? Hardly. There is no
continuity. There are some people who claim to “support” SPICE, but I don’t always
agree with their statements.
My biggest gripe with SPICE is its lack of convergence. The ordinary SPICE 2G6
has all sorts of problems, even if you don’t use FETs. (We find that FETs usually
make the convergence situation really unhappy.) For example, one time I had a mod-
erate-sized circuit with about 33 bipolar transistors, and it didn’t converge well. Then
all of a sudden, one day it started to converge beautifully and quickly. I was so im-
pressed, I backed up to find the “scene of the crime.” I tried to duplicate all the
changes I had made since I last had problems. It finally turned out that I had an un-
used resistor and an un-used capacitor each tied from one point to ground. Nothing
else was connected to that point. Originally they were “commented out” by an as-
terisk. But, at one point, I deleted the asterisk, and the useless R and C were dropped
into the circuit-and they just happened to make the convergence a lot better. When I
removed the R and C, things got worse again.
This lead me to appreciate two things: That the convergence is a lot more fragile
than we suspect, and that we may be able to randomly throw useless resistors into a
circuit, and sometimes they could improve the convergence. In other words, if you
have a circuit that shows bad convergence, the computer might have a subroutine to
randomly sprinkle a few resistors into the circuit and see if that helps-a kind of
“autoconverge” scheme. At present we are still working on this, but it may be a
useful approach. This concept is not too surprising if you have ever heard that the
convergence of a circuit may be improved or degraded depending on the names and
numbers you call the nodes of the circuit. If you swap a couple nodes’ numbers, and
things get better (or worse)doesn’t that make you nervous?? Or, at least,
suspicious?
Another serious problem I had with SPICE was when I ran some simple transient
tests-mangle waves-on the collector of a transistor. I ramped the V, up and down
from +5 V to + 15 V, back and forth, and ran several tests. Then I added some compli-
cating factors, so I wanted to look at the circuit for the first 202 ps of a 10-kHz mangle
wave. Namely, after the first two cycles of the triangle wave, I decided to look at the
collector current ( I = C X dV/dt) of the transistor at t = 201 ps. I got my answer
printed and plotted, and it did not make any sense. I studied the whole circuit, and I
used every troubleshooting technique I could think of, and it did not make any sense.
The current through the 1 pF of C, was not 0.2 PA, but 5 FA. How could that be?
After several hours, I finally decided to look at the incoming waveform. I had
commanded it to go back and forth from 5 V to 15 V, at the rate of 50 ps per each
ramp, so I knew what it had to be doing. But when I looked at t = 201 ps, the dV/dt
had suddenly increased from 0.2 V/ps to 5 V/ps. It turned out that because I had
commanded the PLOT mode to stop at 202 ps, the transient generator had decided to
go from 15 V to 5 V, not in the time from 200 to 250 ps, but in the span from 200 to
202 ps. The dV/dt speeded up by a factor of 25, without being asked to, for a com-
pletely unexpected reason. Nothing in anything I had ever seen about SPICE, nothing
my friends had ever heard, would lead you to expect this. In fact, SPICE sort of en-
courages you to look at the waveforms any time you want-it offers a sort of “in-