This is a collection of some of my non-internet publications. Those of you who
know me well are familiar with my obsessive and compulsive interest in precise
time and frequency measurement. Many of the following items reflect that
interest. (Much more evidence of my interest in this hobby can be found on
the No-Frills Technical Links
Page. I assume you are familiar
with QST,
TV Technology,
EDN, and
Broadcast Engineering.
New! My article
about Channel
37 was published in TV Technology, September
7, 2005, and was soon quoted here in
its entirety. But you can also read it here and
see the footnotes, references, illustrations and updates.
The article, as published in TV Technology, contained one serious typographical
error: Instead of "10 to 15 watts", it should
say "10–15 watts", which would be .001 femtowatt.
QST, November 1983: This was a
letter to the editor of QST (a ham radio magazine) which showed up in the Technical
Correspondence section. It is a critique of an article which appeared in the
August 1983 issue of QST, in which the
author (NØAJY) described
a frequency standard he had designed, which used the color subcarrier in a television
receiver as a frequency standard. This was a good idea for a few years in the
1970's, because the TV networks used atomic standards (rubidium clocks) for control of
the subcarrier frequency. However a technological advance — of which most people
were unaware — negated the usefulness of this technique, and that advance was the
widespread use of frame synchronizers at local TV stations. Note that the term
"cb" in this case is an abbreviation for "color burst", not "citizens band".
EDN, July 1985: This article
describes a BASIC program I wrote as an improvement upon a program which appeared
in an article a few months earlier. This program quickly calculates the two best
standard one-percent resistor values to use in a simple voltage divider. The
other guy's program went through every possible combination of values and kept track
of the best results, but since there
are 96
standard one-percent resistor values, his program had to make the same
calculations 96×95 times. My program
took a different approach to solve the problem more quickly. The 96 standard values
are evenly spaced powers of the 96th root of 10, so it is quicker to
use the voltage ratios to arrive at the correct resistor values.
EDN, 1986: This is another Design Idea
article co-authored with Richard Kihn, who is now the Chief Engineer of
KFDM-TV (not KSDM) in Beaumont, TX. This project began in the late 1970's
when inexpensive digital clocks first became available. Unlike 21st-century
digital clocks which usually have their own crystal oscillators, the first digital clocks
used the 60-Hz power line frequency as a time base. Unfortunately, Beaumont
was (and is) a heavily industrial area with frequent thunderstorms and the quality
of the AC power was often far from ideal, so the digital clock in use at
KFDM-TV drifted unacceptably. (Sidebar discussion: One clever way around
this problem came from the invention of the
MM5369 which
was a crystal-controlled source of clean 60-Hz square waves.) Anyway, Richard
wanted to use the station's very
stable sync generator as a timebase for the digital clock. Simple
arithmetic revealed that the 59.94 Hz "Vertical Drive" signal
in the station's sync generator needed a correction of exactly 0.1%, which
was accomplished by adding one extra pulse to the clock input after
every thousand Vertical Drive pulses.
EDN, 3/31/1987: This
article in the Design Ideas section describes a circuit which interprets
the quadrature outputs of an optical encoder into "Up" and "Down" pulses
without backlash. This is accomplished by examining the state of
one encoder output at the moment there is a transition on the other
output. Simple logic chips sort it all out.
EDN, 10/27/1988: This Design Idea was one of
the simplest and one of the most popular. Several people have mentioned seeing this article, which is
quite gratifying. The circuit is little more than a fifteen-bit pseudorandom sequencer and a solid state
relay. The "chattering" relay makes an incandescent lamp flicker, which is useful in theatrical effects
or outdoor advertising. The schematic diagram published in EDN included the following errors which were
not in the drawings I submitted: The "0.68 pF" capacitor is of course 0.68 µF. The
pin on the 555 with no connection (NC) is pin 5. And the Exclusive-OR gate is a 74HC86, not
a 74HCB6.
EDN, 8/3/1989: One of my favorite designs is
this low-frequency rumble generator. Random noise (from a noisy diode) is converted into a logic level
producing a random bit stream. This is used to randomly increment or decrement an eight-bit counter which
is connected in such a way that it cannot overflow or underflow. The counter staggers up and down between
0 and 255 (decimal) and this number is sent to a digital-to-analog converter (DAC) which produces a voltage
that fluctuates between 0 and 2.55 volts. The result is wide bandwidth noise that is rich in very low
frequencies. Applications for this include theatrical effects and random motion generation for "shaker
tables" and other testing devices. This design started out as an invention that would gently rock my
infant son to sleep, since he didn't like going to bed but he always went right to sleep in the car.
Broadcast Engineering,
December 1989: This article, entitled
"Wait Just a Second", may have been -- at that time -- the only
magazine article ever written about
leap seconds.
EDN, 2/15/1990: This
Design Idea article describes a simple circuit to extract the horizontal
sync pulse from Field 1, Line 10 in the
NTSC television system. Line 10 was used in the late 1970's and
early 1980's as a means of precisely comparing clocks which are miles apart
but both within range of a television station. For example, the
Line 10 method was used to "transfer" the time of day from the
NBS (now NIST) laboratories
in Boulder to
the WWV transmitter
site in Fort Collins by measuring the timing on a Denver TV station. My
article was used many years later as a reference
in U.S. Patent #5619275.
|
|
