M
Max Hauser
Guest
At dinner recently with an experienced analog designer (Paul Brokaw, bit of
a mentor to me since 1977) over adequate restaurant Reuben sandwiches [1],
discussing experiences old and new, and circuit-design training techniques
[2], it became obvious that some people are, maybe not circuit designers
from birth, but sooner or later stuck with the habit, like it or not. A
clear example is an engineer I know who considered other careers, spending
time as a monk in France. He found himself involuntarily sketching
transistor circuits; this and other factors convinced him his vocation lay
elsewhere. (The Carthusians had very reasonably assigned him duties
repairing their electrical systems.) I briefly worked in the interesting
and nearer vocation of communication systems and still found myself
sketching transistor circuits (one of which by the way got a patent much
later, by someone else -- a case alluded to in a tirade in the current
Plagiarism thread on comp.dsp). I've known a number of people with such
involuntary circuit habits. One or two I knew in college were not studying
engineering but could not break the habit of building things or trying
design ideas to solve problems. Incorrigibles.
Supplemental notes
[1] Even adequate Reubens are still sad, because a good Reuben sandwich is
not only a serious pleasure but easy to make. Good ingredients and good
"Russian Dressing" are the main things. Even though the Reuben is young for
a classic sandwich (1956 is the likely origin), still it has managed to turn
bland. One culprit is "Thousand Island Dressing" which traditionally (De
Gouy, 1948) is a bland version of Russian Dressing with lots of mayonnaise,
with whipped cream added to suppress any surviving flavor. (By the way, the
Reuben manages to violate all major religious dietary codes of the world and
all common or fashionable dietary restrictions, including low
carbohydrates.) Sinful.
[2] One example was the "beta schools." Student engineers at some schools
have been taught about bipolar transistors in basic amplifier stages
emphasizing "beta" or common-emitter current gain. In this view collector
current is controlled most fundamentally by base current, and if you come to
think of beta as a central and solid parameter, you can even set up a
common-emitter amplfier by forcing a fixed current into the base, and then
designing for an expected resulting collector current. As far as I know, no
expert transistor-level designer does this (anyway I have not seen it in 30+
years of knowing many such people) but it is how the subject is sometimes
taught. (A practical impracticality is that "beta" is among the least
predictable and stable params in all of solid-state electronics. Expert
designers typically organize their designs to be as unaffected by beta as
possible.) The "-1" factor in the tidy "Ebers-Moll" models is an even more
misleading abstraction, but that's enough tirade for now.
Max Hauser
(Copyright 2004)
a mentor to me since 1977) over adequate restaurant Reuben sandwiches [1],
discussing experiences old and new, and circuit-design training techniques
[2], it became obvious that some people are, maybe not circuit designers
from birth, but sooner or later stuck with the habit, like it or not. A
clear example is an engineer I know who considered other careers, spending
time as a monk in France. He found himself involuntarily sketching
transistor circuits; this and other factors convinced him his vocation lay
elsewhere. (The Carthusians had very reasonably assigned him duties
repairing their electrical systems.) I briefly worked in the interesting
and nearer vocation of communication systems and still found myself
sketching transistor circuits (one of which by the way got a patent much
later, by someone else -- a case alluded to in a tirade in the current
Plagiarism thread on comp.dsp). I've known a number of people with such
involuntary circuit habits. One or two I knew in college were not studying
engineering but could not break the habit of building things or trying
design ideas to solve problems. Incorrigibles.
Supplemental notes
[1] Even adequate Reubens are still sad, because a good Reuben sandwich is
not only a serious pleasure but easy to make. Good ingredients and good
"Russian Dressing" are the main things. Even though the Reuben is young for
a classic sandwich (1956 is the likely origin), still it has managed to turn
bland. One culprit is "Thousand Island Dressing" which traditionally (De
Gouy, 1948) is a bland version of Russian Dressing with lots of mayonnaise,
with whipped cream added to suppress any surviving flavor. (By the way, the
Reuben manages to violate all major religious dietary codes of the world and
all common or fashionable dietary restrictions, including low
carbohydrates.) Sinful.
[2] One example was the "beta schools." Student engineers at some schools
have been taught about bipolar transistors in basic amplifier stages
emphasizing "beta" or common-emitter current gain. In this view collector
current is controlled most fundamentally by base current, and if you come to
think of beta as a central and solid parameter, you can even set up a
common-emitter amplfier by forcing a fixed current into the base, and then
designing for an expected resulting collector current. As far as I know, no
expert transistor-level designer does this (anyway I have not seen it in 30+
years of knowing many such people) but it is how the subject is sometimes
taught. (A practical impracticality is that "beta" is among the least
predictable and stable params in all of solid-state electronics. Expert
designers typically organize their designs to be as unaffected by beta as
possible.) The "-1" factor in the tidy "Ebers-Moll" models is an even more
misleading abstraction, but that's enough tirade for now.
Max Hauser
(Copyright 2004)
480)460-2350 | |