PSpice Switch Convergence

[engr4fun]

Hi,

I'm running a long overnight simulation, and I set up some circuitry
that would cause the circuit to have a convergence failure if a certain
condition was met so that I can save computation time on the system. I
used an S switch and configured it so that if the condition occurs the
switch would just toggle on and off and then it would hit the iteration
limit. However, PSpice finds a solution for this. I was wondering if
anyone knows the convergence routine for this switch. I've pasted a
simplified version below where I get a constant 1.2 volts at node 2
which when I look at the equations in their manual for this component,
I just can't see how that number is reached. Any insights? Thanks,
Alex.

*
R1 2 0 1
S1 2 1 2 0 TEST
V1 1 0 12
*
..MODEL TEST VSWITCH (ROFF=1E9 RON=1 VOFF=2 VON=1)
*
..TEMP 27
..TRAN 2e-008 1u 0 UIC
..PRINT TRAN V(1) V(2)
..PLOT TRAN V(1) V(2)
..PROBE
..END

 

[Helmut Sennewald]

"engr4fun" <engr4fun@yahoo.com> schrieb im Newsbeitrag
news:1124736499.803361.5970@g43g2000cwa.googlegroups.com...

Hi,

I'm running a long overnight simulation, and I set up some circuitry
that would cause the circuit to have a convergence failure if a certain
condition was met so that I can save computation time on the system. I
used an S switch and configured it so that if the condition occurs the
switch would just toggle on and off and then it would hit the iteration
limit. However, PSpice finds a solution for this. I was wondering if
anyone knows the convergence routine for this switch. I've pasted a
simplified version below where I get a constant 1.2 volts at node 2
which when I look at the equations in their manual for this component,
I just can't see how that number is reached. Any insights? Thanks,
Alex.

Hello Alex,
you shouldn't believe all the formulas.
I am sure now that there is a small but important mistake in PSPCREF.pdf.

They defined Vd regardless whether Von is smaller or greater than Voff.

Vd=Von-Voff

In your case Von is lower than Voff. So you get a negative number.
I tried with the opposite formula for your case.

Vd=Voff-Von

Now you have also a positive number in this case.
The result for Rs from the big formula is then in accordance
with the simulation result.

I have used LTspice to figure out this PSPICE documentation error.

Best regards,
Helmut

*
R1 2 0 1
S1 2 1 2 0 TEST
V1 1 0 12
*
.MODEL TEST VSWITCH (ROFF=1E9 RON=1 VOFF=2 VON=1)
*
.TEMP 27
.TRAN 2e-008 1u 0 UIC
.PRINT TRAN V(1) V(2)
.PLOT TRAN V(1) V(2)
.PROBE
.END

 

[engr4fun]

Thanks Helmut,

It looks like you're right. Once I make that polarity change, I get
the expected results. Thanks for your help.

Alex