IEEE Solid-State Circuits Magazine - Summer 2015 - 20

Telegraph Corporation
of pulses counted by
in 1937. The very first
the encoding counter is
PCM patent by Reeves
thus the complement of
was filed in France, but
the incoming data word.
it was immediately folThe output of the R/S
lowed by similar patflip-flop is a PWM signal
ents in Britain and the
whose analog value is the
United States, all listing
complement of the input
Reeves as the inventor
binary word. Reeves uses
[26]. These patents were
a simple low-pass filter
very comprehensive and
to recover the analog sigcovered the far-reaching
nal from the PWM output.
Figure 1: One of the earliest references to a criterion for determining the
topics of 1) general prin- minimum sampling rate required to reconstruct a signal. Extracted from
The phase inversion in the
ciples of quantization Williard M. Miner, "Multiplex Telephony," U.S. Patent 745, 734, filed
DAC is easily corrected in
and encoding; 2) the 26 February 1903, issued 1 December 1903.
either the logic or in an
choice of resolution to
amplifier further down the
suit the noise and bandsignal chain [26].
width of the transmission medium;
at the sampling instant. This pulseReeves' patents covered all the
3) transmission of signals in digital
width modulated (PWM) pulse controls
essentials of PCM: sampling, quantizformat serially, in parallel, and as
a gated oscillator, and the number of
ing, and coding the digitized samples
modulated carriers; and 4) a counpulses out of the gated oscillator repfor serial, parallel, phase-modulated,
ter-based design for the required
resents the quantized value of the
and other transmission methods. On
5-bit, 6-kS/s ADCs and DACs. Unlike
analog signal. This pulse train can be
the receiving end, Reeves proposed
the previous PCM patent by Rainey
easily converted to a binary word by
a suitable decoder to reconstruct
in 1926, Reeves took full advantage
driving a counter. In Reeves' system, a
the original analog signal. In spite
of existing vacuum tube technology
master clock of 600 kHz is used, and
of the significance of his work, it is
in his design.
a 100:1 divider generates the 6-kHz
interesting to note that after the patThe ADC and DAC developed by
sampling pulses. The system uses a
ent disclosures, Reeves shifted his
Reeves represent one of the first all5-bit counter, and 31 counts (out of the
attention to the shortwave transmiselectronic data converters on record.
100 counts between sampling pulses)
sion of speech using pulse-amplitude
The ADC technique basically uses a
therefore represent a full-scale signal.
modulation, pulse-duration modulasampling pulse to take a sample of the
The DAC uses a similar counter and
tion, and pulse-position modulation,
analog signal, set an R/S flip-flop, and
clock source. The received binary code
rather than pursuing PCM techniques.
simultaneously start a controlled ramp
is first loaded into the counter, and
A block diagram of a PCM system
voltage. The ramp voltage is comthe R/S flip-flop is reset. The counter
from the patent is shown in Figure 5.
pared with the input, and when they
is then allowed to count upward by
PCM and the Bell System:
are equal, a pulse is generated that
applying the clock pulses. When the
World War II Through 1948
resets the R/S flip-flop. The output of
counter overflows and reaches 00000,
Under a cross-licensing arrangement
the flip-flop is a pulse whose width
the clock source is disconnected, and
with the International Telephone
is proportional to the analog signal
the R/S flip-flop is set. The number
and Telegraph Corporation, Bell
Telephone Laboratories' engineers
reviewed Reeves' circuit descripInput
Output
+
tions and embarked upon their own
µ
pursuit of PCM technology. Start-
ing in about 1940 and during World
War II, studies were conducted on
a speech secrecy system that made
β
PCM techniques mandatory.
Output = µ [Input - β ⋅ Output]
By the end of the war, several groups
at Bell Labs were studying PCM; how1
Output
1
=
ever, most of the wartime results were
1
Input
β
1+
µβ
not published until several years later
because of security issues. The work
of H.S. Black, J.O. Edson, and W.M.
Figure 2: Harold Black's feedback amplifier of 1927. Harold S. Black, "Stabilized feedback
amplifiers," Bell System Technical Journal, vol. 13, no. 1, January 1934.
Goodall were published between 1947

su m m E r 2 0 15

IEEE SOLID-STATE CIRCUITS MAGAZINE

Table of Contents for the Digital Edition of IEEE Solid-State Circuits Magazine - Summer 2015