March 1956 Radio-Electronics
[Table of Contents]
Wax nostalgic about and learn from the history of early electronics.
See articles from Radio-Electronics,
published 1930-1988. All copyrights hereby acknowledged.
If you look up "bit" on
Wikipedia, is claims the word is
a portmanteau of binary + digit, but according to the "Electronic Counter Tube"
article in the March 1956 issue of Radio−Electronics magazine, it comes
from binary + unit. The cited reference is a 1980 book. That's a minor point,
but I wonder if the belief has changed over time. Going further down the metaphorical
rabbit hole, per the
same source, "The term
coined by Werner Buchholz in June 1956, during the early design phase for the IBM
Stretch computer." Now, back to the topic at hand. The
Counter Tube" was produced by Mullard,
located in the UK. It had an ingenious design which directed an internal electron
beam onto fluorescent numerical segments according to the number of pulses supplied
to the increment grid (D1 in Fig. 1). A holding voltage was not needed to maintain
the position of the last pulse. Pulse shape was critical, and was supplied by a
special-purpose tube, the Mullard E90CC.
Electronic Counter Tube
A special type of C-R tube, the E1T can
count in tens
By J. R. Davies
One of the most fascinating branches of electronics is that of counters and computers.
Devices of this type, with their phenomenal speed and accuracy, appear to the layman
to be somewhat of a near-miracle. To describe some of the more advanced computers
as "electronic brains" is not too far from the truth - such equipment, in their
special use, simulate functions of the human brain very closely.
The basic "cell" of most conventional counters and computers is the vacuum tube,
usually working in what are known as binary units, or "bits." The reason for this
choice of units is that a single tube can be relied upon only to indicate one of
two states: when it is conductive (passing cathode current); when it is nonconductive
(beyond grid cutoff). It would be impracticable to expect a tube to cope reliably
with three states such as: off (past grid cutoff); half-on (low cathode current);
fully on (high cathode current). Tube spread and deterioration could cause the half-on
and fully on states to lie too close together after a period of use, and there would
be a risk of misinterpretation.
Fig. 1 - Internal construction of E1T.
Fig. 2 - Schematic of a typical circuit in which the E1T
may be used.
A tube capable of counting in 10's or decades would overcome this trouble. This
type of tube exists, a typical one being the E1T made by Mullard in England. This
tube may be used in cascade with others with the result that when one tube has reached
what would otherwise be "10" it triggers the next tube in the circuit and slips
back itself to zero. Thus, a group of three tubes can be used to count to 999: the
first tube counts digits and triggers the second tube at every decade; the second
tube counts in 10's and similarly triggers the third tube which counts in 100's.
To add to its versatility, each of the E1T tubes not only counts but also indicates
its number on an easily read built-in fluorescent screen. The photo shows the tube
and its indicating screen. The top counting speed is 30 kc.
Construction and circuitry
The internal construction of the E1T is shown in Fig. 1. Fig. 2 illustrates
a typical circuit.
In Fig. 1 the cathode emits an electron stream which passes through grid
G1 and the beam-forming electrodes B1, B2. The stream next passes through G2 and,
being at B-plus potential, is accelerated. Leaving G2 the electron stream is in
the form of a thin ribbon-shaped beam, the major width of the ribbon being at right
angles to the surface of the paper.
The beam now travels through deflector plates D1, D2 to G4. However, G4 is perforated
and the narrow beam is not wide enough to pass through more than one of its 10 apertures
before it reaches segmented plate A2. This plate has apertures also and the beam
has to pass through one of these before it finally strikes fluorescent conductive
target T. The apertures in A2 are staggered alternately, thus allowing the fluorescent
indications on the target to be similarly staggered for well spaced out readings.
In addition, A2 is sectionalized, preventing the beam from spreading too widely
after it has passed through the particular aperture it has chosen in G4.
Due to the narrowness of the beam, once it has settled at a particular aperture
in G4, it remains locked there until some external force is applied to move it elsewhere.
Deflector plates D1 and D2 are arranged so that they provide this beam-moving force
in a horizontal plane relative to the surface of the paper in Fig. 1. Plate
D2 has a positive potential higher than that on D1. Thus the beam is attracted,
when the tube is first switched on, to the furthermost right-hand aperture of G4.
This position corresponds to 0 on the fluorescent target.
When a positive counting pulse of the correct shape (provided by a suitable input
filter) is injected into the circuit of Fig. 2 at A, D1 becomes momentarily
more positive than D2 causing the beam to move to the left. However, the shape of
the pulse is such that the beam moves only one step (or one aperture in G4) whereupon
it then stops and becomes locked in its new position. Successive pulses cause the
beam to move to successive apertures. On the arrival of the tenth pulse, the beam
passes from the tenth aperture of G4 and strikes A1. Current is drawn by this plate,
causing a negative pulse to appear at point B (Fig. 2). This negative pulse
triggers an external circuit which does two things: it applies a reset pulse to
the tube so that it may return to zero; it passes a counting pulse to the next decade-counting
tube along the line.
The reset pulse consists of a negative pulse injected at C and causes G1 to go
negative, whereupon the tube cuts off altogether. When the negative pulse at G1
ceases, the counting cycle starts up all over again with the electron beam of the
tube passing through the furthermost right-hand aperture in G4.
The tube is then ready to count another decade.
Posted October 17, 2022