Dan McMahill wrote:
Larry Doolittle wrote:
Guys -
On Tue, Jan 24, 2006 at 07:00:16PM -0800, Steve Meier wrote:
I once was listening to a EE Professor who was giving a lecture on the
"Ideal" opamp. As I recall, he stated that the "Ideal" opamp should
have
a large imput impedence and infinite gain..... Which he went on to
further define as "approximetly 400."
Putting on my "engineering theorist" hat ...
There are conceptual problems with an amplifier that has
infinite gain at finite frequencies, or even finite gain
at all frequencies.
An ideal op-amp has large, finite gain-bandwidth-product,
which I could further define as "approximately 1 GHz".
Note that this treatment does involve infinite gain at DC,
which is OK.
so just for fun, now that we're on the topic of silly op-amp tricks....
Here's a circuit which I came across inside some commercially sold
equipment. The unit was not cheap (in terms of cost) and was not for
the consumer market.
This is a little circuit with only an op-amp and 3 passives. Can't
get more simple than that right? Turns out that the designers of it
had some troubles understanding it.
The answers (to an accurate enough degree for engineering
calculations) can be had in about 3 lines of hand analysis but I think
it illustrates nicely how one must be careful with those "ideal"
op-amps. It's also a good case of where one shouldn't substitute a
simulator (with a model which may not capture important details) for
thinking!
Have fun.
-Dan
p.s. hint: The pin on the schematic is the only connection (other
than power and ground) to the rest of the circuit that this little gem
was found in.
Dan,
Great teaser! I haven't seen this trick for a few years. We used to
actually use circuits like this in products we designed. Care to
divulge where you ran across it?
Another interesting, related, and possibly more useful circuit is
obtained when you exchange R1 and C1 while making R2 quite large.
"Duality" at work.
Joe T
