Robert Watson-Watt (1892-1973)

614

   WATT AND HERD: AN INSTANTANEOUS

refer particularly to the amplifier and the oscillograph.
The amplifier should give a distortionless linear magnifica-
tion, without rectification, over a range of ± 50 volts
output, unless an oscillograph more sensitive than the
standard 300-volt pattern is used. There is room for
some very interesting work on amplifiers specially
adapted to meet these conditions when working into
such a high-impedance terminal apparatus as the
oscillograph. Such work has been begun, but need not
be discussed at present. It should, however, be noted
that the previous visual direct-reading radiogoniometer
due to Artom involved rectification, which results in
the introduction of further ambiguities in addition to
the ordinary radiogoniometric ambiguity of 180°, since
it gives bearings which are concentrated in one quadrant,
so that there is no distinction of a given bearing from
its image with respect to one of the principal axes, giving
a fourfold ambiguity when the non-discrimination of
sense is taken into account. In the present device the
only ambiguity is that of sense, and it can be resolved
by methods already used in radiogoniometry, and also
by methods peculiar to the oscillographic arrangement.

The conditions to be fulfilled by the cathode-ray
oscillograph are not, except in respect of the one point
already discussed in relation to the different distances
between screen and deflecting systems, different from
those to be fulfilled by a general-purpose oscillograph.
These are, summarily : (a) That the indicating spot shall
be small, symmetrical and sharply defined, and shall
remain so at all velocities of deflection; (b) that the
deflections shall be linearly related to the applied field
over the full range of the screen; (c) that the axes of
the two deflecting systems shall be strictly at right
angles; (d) that the deflectional sensitivity shall be as
high as possible; (e) that the fluorescent material shall
give the most brilliant response possible to the lowest
electron speeds and densities used; and (f) that the life
of the instrument shall be as long as possible.

It is a truism to say that the cathode-ray oscillograph
as a general laboratory and industrial tool is in its
infancy, since the inconveniences of the older forms have
only recently been removed. Rapid improvement in
quality and uniformity of the product may be expected
to accompany increased demand, and it is significant
that nearly half the November issue of the Journal was
devoted to this instrument. Recent work * has resulted
in the production of an oscillograph with a sensitivity
of nearly 1 cm per volt, and work initiated by the
Radio Research Board has shown that very considerable
improvements in fluorescent screens for visual work are
within immediate reach. The future of the cathode-ray
direction-finder is intimately linked with the progress
of the cathode-ray oscillograph, and must not be
judged alone on its performance with existing oscillo-
graphs, good as that performance is.

It is clearly necessary that provision should be made
for the tuning, testing and adjustment of the radio-
goniometer system and the arrangements adopted at
Ditton Park will serve as an illustration of this provision.
When the whole system is correctly tuned and adjusted,
the arrival of a sustained signal of the selected frequency
causes the indicating spot of the oscillograph to trace
a straight line which makes with the two principal
axes—representing the deflections due to the two pairs
of deflecting plates acting independently—angles which
are the angles between the direction of arrival of the
signal and the planes of the corresponding aerials. The
effect of slight mistuning is to open this straight line
into an ellipse, since the spot is now under the control
of two misphased fields. This opening to an ellipse is
a very sensitive index of mistuning, and is, in the case
of “ single ray” propagation, at least, an effective
safeguard against errors of bearing due to bad tuning.
The ellipse is quite wide before its major axis begins to
depart by a measurable amount from the correct angle.
This tuning operation may be performed directly on the
signal, or it may be performed on locally generated
oscillations of the desired frequency. A screened
calibrated oscillator is coupled to a testing instrument in
which the main loading inductances already mentioned
form the secondary windings of a crossed transformer
the loading coils belonging to one loop being coaxial and
having their common axis at right angles to the common
axis of the coils belonging to the other loop. The
primary is a coil capable of rotation about an axis in
its own plane perpendicular to these axes, and is fed
from the oscillator. The system is in fact similar to
the crossed transformer of the Bellini-Tosi radiogonio-
meter, the search coil being excited and inducing into
the field coils, instead of the inverse operation as in
directional reception. The primary may first be
coupled to one loop alone and this loop tuned up, then
the other loop may be tuned independently; finally both
may be tuned to identity as tested by the closing of the
ellipse. If, as may happen, the operation is complicated
by the presence of signals, the loops may be replaced
by dummy circuits of the same inductance and resist-
ance. This instrument also permits the testing of the
voltage amplification on each side of the system. In
the ideal case these amplifications should be identical,
but in practice circumstances may call for a ratio
differing from unity. For instance, the difference in
distance from screen to deflectors, already mentioned,
is most easily corrected by the increase of the amplifica-
tion factor on the side connected to the deflector system
nearer the screen, until the deflections corresponding to
identical inputs to the two systems are identical. It
will be seen that as the only convenient measure of the
relative amplifications is by length of line on the screen,
the operator is not required to remember anything
about this correction per se—its compensation is
automatically included in the general adjustment
amplification.

The methods of determining the need for adjustment
of amplification and for testing its amount having been
indicated, it remains to show how the adjustment is
performed. The method may of course vary according
to the typle of amplifier in use, but there is probably no
more simple and satisfactory method for general purposes
than the provision of a variable tapping of the anode
circuit resistances in one stage of resistance-capacity
amplification, so that the proportion of the whole voltage
released across this resistance which is transferred to the
oscillograph or to the next grid may be varied over
wide limits. Preliminary selection of matched triodes

* BUCHTA: Journal of the Optical Society of America, 1925, vol. 10, p. 681.