Do you ever need to derive system transfer functions of analog circuits or filters? Once you get beyond a few components, the algebra can become quite tedious. Why not use the symbolic toolbox to handle this for you?
Let's assume that you have a netlist of resistors, capacitors, inductors and even opamps.
The netlist might look like this:
====================
# example op amp circuit
#
r1 vi e1
r2 e1 e2
c2 e2 vo
opamp1 0 e1 vo
c3 e1 vo
====================
Note that there are no component values; only the topology is presented. The netlist file needs to have a 'vi' and 'vo' node.
This pair of .m files will read in the file, and provide the transfer function in analytic form, allowing you to substitute real component values and even run your own sensitivity analyses. I find this helpful in looking at what resistors or capacitors combination affects which poles and zeros.
The output might look like this:
The solutions are:
e1 = 0
e2 = (c2*r2)/(r1*(c2 + c3 + c2*c3*r2*s))
vi = 1
vo = (c2*r2*s + 1)/(r1*s*(c2 + c3 + c2*c3*r2*s))
Or if you want to make it look nice, you can run:
>> pretty(xsoln(end))
c2 r2 s + 1
 
r1 s (c2 + c3 + c2 c3 r2 s)
Enjoy!
Raj
