From: Tony Williams
Subject: Re: Historical question: negative feedback and the op amp
Date: Thu, 14 Nov 2002 09:28:05 +0000 (GMT)
References: <email@example.com> <3DD24CD4.ECD230B9@webaccess.net>
NNTP-Posting-Date: Thu, 14 Nov 2002 09:29:38 +0000 (UTC)
User-Agent: Pluto/1.14i (RISC-OS/3.60)
In article <3DD24CD4.ECD230B9@webaccess.net>,
Chuck Simmons wrote:
> There might well have been some of that in WWII. However, a lot of
> analog computing devices used then were electromechanical and used AC
> carrier control.
I've done some searchings.....
The earliest analogue computing element that I can
think of is the Miller Integrator, used initially
as a linear sawtooth generator, but known as a
mathematical integrator from day.1. F.E Terman
credits the application of the Miller Effect to the
integrator application to A.D Blumlein. That was
via Blumlein's colleague F.C Williams, because
Blumlein himself was killed in a wartime plane crash.
I don't have any date-reference for Blumlein's work,
mainly because (by 1939) the UK was into WWII and I
believe that Blumlein was already working at a high
security establishment (TRE, Malvern). However,
Terman himself (apparently) described Blumlein's
Integrator in his 1943 Radio Engineers' Handbook,
based on a pentode with a feedback capacitance from
anode-grid. He notes that Cin = Cf*(1+A), where A= gm*Rl.
Note the "apparently". I don't have Terman's 1943
book but have a 1950 Terman book, where he references
back to his 1943 book in the footnotes.
Some time in the 40's, by about 1946/47, that feedback
concept had developed into a more generalised form.
Ein +---[Z1]---+---[Zf]---+---+ Eout = Ein * (-Zf/Z1)
|A >---+ A= >> 1.
That generalised mathematical model above was probably
known by 1946/47, where the term (-Zf/Z1) was already
being called the mathematical operator (p). Even the
term "virtual ground" might have been in use by then.
Operational feedback theory was developed far ahead of
the available hardware, with early work being done with
operational amplifiers based on a single pentode.
I don't have access to old papers, but I suspect that
a useful search would be to look through Journals of
the I.E.E from about 1946 onwards, looking for papers
which include the author, F.C Williams. In particular;
Williams, F.C., F.J.U Ritson, and T. Kilburn.
Journal I.E.E, Vol 93, pp 1275, (1946).
Title not known.