[Help-gnucap] Unstable inverter isn't unstable

[Orestes Mas]

Hello,

Last week, I gave a simulation assignment to my students. It consisted on 
simulating a simple OpAmp-based inverting amplifier. I suggested them to use a 
uA741 macromodel to simulate the OA instead of a simple high-gain VCVS. The 
macromodel was based on the PSpice one, slightly modified to get rid of 2 CCCS- 
and VCVS- POLY sources this macromodel has inside (modified model attached).

One of my students made the obvious mistake of exchanging the 2 inputs of the 
operational amplifier, thus simulating "de-facto" an Schmitt trigger circuit.
But, to my surprise, the simulation results didn't show the Operational 
Amplifier 
saturation, but linear behaviour!

The circuit is as follows:

* Inverter amplifier
.INCLUDE uA741.model

Vg 1 0 DC 1V
R1 1 2 68k
R2 2 3 204k
Vccp 4 0 DC 15V
Vccn 0 5 DC 15V
X1 2 0 4 5 3 uA741  (Inputs exchanged)

.PRINT DC V(3)
.DC Vg -2 2 0.5
.END

The simulation results are (!)
(using Gnucap 2009.12.07 RCS 26.136)
gnucap> DC vg -2 2 0.5
#           v(3)      
-2.         6.0163    
-1.5        4.5163    
-1.         3.0162    
-0.5        1.5162    
 0.         0.016187  
 0.5       -1.4838    
 1.        -2.9839    
 1.5       -4.4839    
 2.        -5.9839    

Strangely, if I extend the DC range, results are correct or, at least, it seems 
to 
behave like a Schmitt trigger:
gnucap> DC Vg -5 5 0.5
#           V(3)      
-5.        -14.613    
-4.5       -14.613    
-4.        -14.613    
-3.5       -14.613    
-3.        -14.613    
-2.5       -14.613    
-2.        -14.613    
-1.5       -14.613    
-1.        -14.613    
-0.5       -14.613    
 0.        -14.613    
 0.5       -14.613    
 1.        -14.613    
 1.5       -14.613    
 2.        -14.613    
 2.5       -14.613    
 3.        -14.613    
 3.5       -14.613    
 4.        -14.613    
 4.5       -14.613    
did not converge
 5.        -14.983    

Can you please elaborate on this behaviour? I know that, when dealing with an 
Schmitt trigger, you have to specify the initial (output) state to be abple to 
calculate things properly, but I assumed (perhaps erroneously) that GnuCap's 
solver 
should find the correct output after some iterations starting from a zero 
output. 
Am I wrong? How should I do to simulate this circuit properly and obtain the 
correct 
result?

Thanks,
Orestes Mas.


The modified PSpice macromodel:
*-------------------------------------------------------------------------------
* Library of linear IC definitions

* This is a reduced version of MicroSim's linear subcircuit library.
* You are welcome to make as many copies of it as you find convenient.
*
* The parameters in the opamp library were derived from the data sheets for
* each part.  The macromodel used is similar to the one described in:
*
*       Macromodeling of Integrated Circuit Operational Amplifiers
*         by Graeme Boyle, Barry Cohn, Donald Pederson, and James Solomon
*       IEEE Journal of SoliE-State Circuits, Vol. SC-9, no. 6, Dec. 1974
*
* Differences from the reference (above) occur in the output limiting stage
* which was modified to reduce internally generated currents associated with
* output voltage limiting, as well as short-circuit current limiting.
*
* The opamps are modelled at room temperature and do not track changes with
* temperature.  This library file contains models for nominal, not worst case,
* devices.
*
*$
*-----------------------------------------------------------------------------
* connections:   non-inverting input
*                |  inverting input
*                |  |  positive power supply
*                |  |  |  negative power supply
*                |  |  |  |  output
*                |  |  |  |  |
.subckt uA741    1  2  3  4  5
*
  c1   11 12 8.661E-12
  c2    6  7 30.00E-12
  dc    5 53 dx
  de   54  5 dx
  dlp  90 91 dx
  dln  92 90 dx
  dp    4  3 dx
* I've had to rewrite these 2 sources because pspice2 POLY fonts aren't 
supported
* in neither ng-spice nor gnucap 
* Original font, commented out
* egnd 99  0 poly(2) (3,0) (4,0) 0 .5 .5
egnd1 99 98 3 0 .5
egnd2 98 0 4 0 .5
* Original font, commented out
* fb    7 99 poly(5) vb vc ve vlp vln 0 10.61E6 -10E6 10E6 10E6 -10E6
  fb1   7 99 vb 10.61E6
  fb2   7 99 vc -10E6
  fb3   7 99 ve 10E6
  fb4   7 99 vlp 10E6
  fb5   7 99 vln -10E6
  ga    6  0 11 12 188.5E-6