[sdiy] Tips on network theory??

David G Dixon dixon at mail.ubc.ca
Fri Oct 19 17:41:42 CEST 2012


> So what stuff SHOULD i learn of network theory?

***** WARNING: POTENTIALLY CONTROVERSIAL OPINIONS BELOW *****

It seems to me that, in order to be able to derive the required expressions
for the two-port impedances used in network theory, you've got to conduct a
complete nodal current balance around the circuit at hand anyway.  Hence,
why not just conduct the nodal current balance and be done with it?  It
gives the same answer, and does so in any more intuitively satisfying way.
As a case in point, have a look here:

http://en.wikipedia.org/wiki/Two-port_network

Scroll down to the two-port network analysis of the bipolar current mirror
with emitter degeneration (Figure 3).  Perhaps someone could tell me what is
to be gained by thinking about the circuit in Figure 3 in terms of Figure 5,
when it must be analyzed in terms of Figure 4 in either case?  Perhaps, if
every functional block in a larger circuit were reduced to two-ports, then
it might make the linear algebra a bit more manageable, but at this point
you'd probably want to use a SPICE simulator anyway.

That's my take on it.  Others may have different opinions.  The two books I
spent the most time with for learning the basics were "The Art of
Electronics" by Horowitz & Hill and "Electronic Principles" by Malvino (3rd
ed.).  Neither book even touches on network theory.  I find that most
telling.  I also find it telling that most EEs are forced to learn network
theory in their undergraduate courses, but most of them don't acquire enough
skill to design even the most basic analog circuits.





More information about the Synth-diy mailing list