From: Winfield Hill
Subject: Re: how to master electronics, use of transistor, op amp...?
Date: 2 Nov 2002 06:53:17 -0800
Organization: Rowland Institute
References: <3DC282F0.firstname.lastname@example.org> <3DC2E327.email@example.com> <3DC3B8B8.firstname.lastname@example.org>
X-Newsreader: Direct Read News 4.00
> Well, this sounds like an interesting little circuit. I actually
> don't usually approach a design from this prospective either.
> Again, I'm self taught, and although it has obvious advantages to
> me, it typically means that my approach is to solve a particular
> need. What this normally requires is some understanding of the
> purpose of the circuit, not in turms of specs on paper, but in
> the real world application. Something like: here is a signal
> that need to get through a 50FT. cable with no more than so much
> loss. The signal can't see an impedance of less than x, and must
> arrive at the other end with as little distortion as possible,
> etc, etc.
> You see, I simply skip most of the calculations and just put
> together something that would probably have at least fair
> performance, then tweak the thing as needed. In the case of your
> circuit requirements, I can see that simply using a standard
> breadboard would screw up the whole thing. Stray capacitances
> alone would see to that.
No, my circuit is (or at least will be) constructed on a breadboard
with miniDIP ICs and standard 1/4-watt sized 1% resistors, etc., but
using slightly-unusual yet standard techniques where low capacitance
is required, such as spacing, shields and air wiring. We only need
one or two, so we won't create a PCB for it at this time. So, Wafer,
if you're correct in your assertions, you should be able to use your
intuitive methods, or whatever, to accomplish the goal. But I think
you are flat wrong about it, calculations cannot be skipped.
OK, you say you need to understanding of the purpose of the circuit,
not as a paper spec, but the real-world application? OK, fair enough.
One purpose of the circuit is to bench test a fast (1-us) switchable
low-current (0.5pA), low-capacitance (0.8pF) current source. I am
especially concerned with quick settling after the shutoff, to about
0.5% within say 1us. The current sources are for precisely canceling
ionic currents through single nanometer-sized pores in a membrane,
thereby eliminating the reset glitches that high currents can create
in a capacitance-feedback transresistance patch-clamp amplifier as its
integrator fills up, without adding noise or any more capacitance than
necessary to the amplifier's summing junction.