April 1932 QST
articles are scanned and OCRed from old editions of the ARRL's QST magazine. Here is a list
of the QST articles I have already posted. All copyrights are hereby acknowledged.
radio frequencies continued to increase during the early years of 'wireless'
development, the use of quartz crystals as a stable reference source
ran into a physical limitation because as crystal slices reached a certain
thinness, overtone and subharmonics appeared that caused problems in
circuits. A new mineral called tourmaline saved the day. With an elasticity
much greater than quartz, tourmaline is able to vibrate at higher fundamental
frequencies for a given thickness. Since that time, science has provided
the means for utilizing the much more abundant quartz crystals in overtone
modes so that tourmaline is not required. At the time, though, it was
a much-welcomed alternative.
Fundamental Crystal Control for Ultra-High
Tourmaline Oscillators for Wavelengths Down to 1.2 Meters
Fig. 1 - Lycopodium pattern of transverse oscillation of a quartz
plate cut so that vibrations starting at the center are reflected
simultaneously from the circumference. The powder thus collects
at the central nodal point.
Fig. 2 - Lycopodium pattern on the surface of a "raw" tourmaline
crystal plate that was cut perpendicular to the axis.
Fig. 3 - When the crystal of Fig. 2 was made circular the pattern
of the oscillations became more symmetrical, as shown by the concentric
rings of the lycopodium collected at the nodes.
Fig. 4 - Contrast this pattern of a disc-shaped quartz oscillator
with those of the tourmaline shown in Fig. 3.
Fig. 5 - Ii the tourmaline crystal the optic and electric axis have
the same direction.
Fig. 6 - The tourmaline controlled oscillator circuit. It is of
the same type as the familiar circuit generally used for quartz
Fig. 7 - Performance curve for an RE 134 valve with tourmaline crystals
of the same diameter but of different thicknesses to give various
wavelengths. The efficiency decreases rapidly at the shorter wavelengths.
Fig. 8 - An ultra-high frequency tourmaline crystal that was cracked
near the edges from overloading without impairing its operation.
The diameter of this plate is 8 millimeters (0.312 inch).
Fig. 9 - The tourmaline plate with respect to its three axes. The
major surfaces are perpendicular to the optic (Z) axis and parallel
to the plane of the X and Y axes.
Here is confirmation of exciting rumors that have been coming through
from Germany since last fall. Practicable fundamental crystal control
at the frequencies too high for quartz, has arrived. Tourmaline
crystal does the trick. The pleasure of presenting this article
is the keener because of the friendly cooperation of the A.R.R.L.'s
sister societies, the D.A.S.D. and the R.S.G.B., that have brought
it to us. The connection with the D.A.S.D. of Dr. Straubel, the
author, and that with the R.S.G.B. of Mr. Pilpel, the translator,
make this contribution a fine feather in the cap of amateur radio.
The method of fundamental crystal control which has been used so
far presents difficulties on frequencies higher than 6000 to 7500 kc.
because very thin quartz oscillators produce side tones on several neighboring
frequencies. This multiplicity of oscillations is noticeable even on
longer waves but can be suppressed by cutting the quartz plate in a
suitable shape. Fig. 1 shows such a plate. The surface and sides are
ground so that all oscillations starting at the center will be reflected
at the same time from the circumference of the crystal. The radius of
the curves at the corners depends on the square root of the modulus
of elasticity of quartz. If this plate is excited on its fundamental
frequency, the whole plate oscillates in all directions. Lycopodium
powder, uniformly distributed over the crystal plate, is therefore concentrated
at the middle point, which is the nodal point of the whole disc.
For considerably higher frequencies than 6000-7500 kc. crystal-controlled
short-wave transmitters nowadays do not often use this method, but employ
quartz crystals with fundamental frequencies such as 3000 to 1500 kc.,
with frequency multiplication; that is to say, one uses harmonics of
the crystal oscillator and amplifies them. An installation of this sort
becomes very cumbersome and complicated when used on the ultra short
waves because the harmonics are so weak that amplification must take
place after every frequency doubling before further frequency doubling
If it were possible to construct fundamental
oscillators for such short waves, the installation would be extraordinarily
The tourmaline crystal controlled oscillator operates at 75 megacycles
(4 meters) with the stability and other characteristics of quartz-crystal
controlled oscillators operating at much lower frequencies. Imagine
the number of frequency multiplying stages that would be necessary
to duplicate its performance with quart-crystal control!
On the above-mentioned grounds, quartz is unsuitable for short-wave
fundamental control. The author has found, however, that a tourmaline
crystal produces considerably more uniform oscillations than quartz.
It was noticed even with a "raw" crystal, a plate simply cut perpendicular
to the optical axis, that on exciting the longitudinal oscillations,
extraordinarily uniform transverse oscillations resulted, as shown in
The piezo-electric constant of tourmaline is
i.e., about 10% less than quartz.
If, in spite of the smaller
constants and the irregular natural formation of this crystal, particularly
easy oscillations are noticeable, it is highly probable that this is
because there is considerably less tendency to side-tone oscillation.
Actually, in the natural crystal only a few symmetrical side tones could
be detected and the frequencies of these were far removed from the fundamental.
From these results it can be assumed that the oscillating properties
even of such a thin crystal will remain constant, as they must do if
it is to be used for fundamental control of ultra short-wave transmitters.
A further advantage for the production of very high frequencies lies
in the high speed of sound, necessitated by the extraordinarily large
elasticity modulus of 1,600,000 kilograms per square centimeter in the
direction of the optical axis. Tourmaline, therefore, supplies a frequency
35% higher than quartz of the same thickness, the constant being 80
meters wavelength per millimeter thickness.1
Circuit and Tubes
A circuit for the production
of the oscillations is shown in Fig. 6, while the photograph shows the
actual transmitter for use on 7 meters and employing 5 watts. The simplicity
of construction and absence of any type of screening are noticeable.
For waves down to 2 meters (150 mc.) ordinary detector or power
valves were used (such as Telefunken RE 084 or RE 134), but for shorter
waves a special short-wave valve was employed, the Valvo S 0401. For
longer waves of 5 meters and upwards, a larger valve was found suitable,
the RS 241, which is similar to the American Type '10 and the Philips
TB 04/10. It was noticed that, as is usual with all self-excited transmitters,
the higher the frequency the lower the efficiency, oscillation eventually
ceasing altogether. Fig. 7 shows the performance of an RE 134 valve.
The reason for this is partly the large self-capacity of the crystal
which, for a given valve capacity, upsets the phase of the reaction
on short waves to such an extent that the crystal no longer oscillates.
The crystal then works only as short-circuit capacity while the valve
either oscillates unstabilized or ceases to oscillate altogether. This
can be avoided by reducing the capacity of the crystal, which can be
done only by decreasing the diameter. An increase in the diameter does
not enable the crystal to stand a greater load on these ultra short
waves, however.Suitable Mountings
electrodes next presented certain difficulties. Silvering,2
such as used in thick oscillators, must be omitted as this affects the
fundamental frequency of the thin plate quite considerably and a great
load will cause peeling, although the oscillator remains undamaged.
Therefore, silvering was dispensed with and perfectly level electrodes
were used. Even slight unevenness of the electrodes caused certain parts
of the crystal to melt and become disintegrated. Splintering, such as
occurs with quartz, was never experienced.
By using perfectly
plane electrodes the load on the crystal could be considerably increased.
Working temperatures of more than 100° Centigrade did not affect the
operation in any way. Great overloading (crystal 8 mm. diameter, RS
241 Valve with 350 volts plate potential) once caused the edge of the
crystal to crack without affecting the middle portion or causing the
efficiency to decrease materially. (Fig. 8)
When the crystal
was in a horizontal position, no extra weight was placed upon the electrodes.
Although an additional weight reduced the crystal's controlling properties,
it was possible, in the cases of crystals of larger diameter (12 mm.),
to subject them to pressures up to 500 grams per square centimeter before
they stopped oscillating. In order to render them safe from shocks and
vibrations, a light spring was always used to supply pressure.
With the 75-mc. transmitter
depicted experiments were carried out to determine the actual degree
of frequency stabilization obtainable. The modulated transmissions were
received on a simple detector with regeneration over a distance of 1
kilometer although no aerial was used for either transmitting or receiving.
As far as the transmitter was concerned, no hand-capacity effects were
noticed; but when the hand was placed near the inductance, reception
was louder owing to the body acting as a capacity-coupled aerial. Only
when the coil was actually touched did oscillation cease, to recommence
immediately on the same frequency when the hand was removed. This was
particularly noticeable when the transmitter was keyed in the plate
circuit. The heterodyne note as heard in an oscillating receiver was
absolutely pure on the frequency of 75 mc. (4 meters) and showed d no
frequency change whatsoever, better than could be obtained when using
a 7500-kc. quartz crystal. The well stabilized transmitted frequency
in every case could be received on the detector with regeneration and
without the need for the broad resonance curve of a super-regenerative
By varying the condenser C (Fig. 6) no jumping in frequency
could be noticed. Naturally, the frequency of the tourmaline, just as
that of quartz, was affected by varying the tuning of the plate circuit,
but frequency jumping never took place. To obtain this condition, however,
perfectly plane parallel crystal surfaces are necessary.
temperature coefficient of tourmaline oscillators is about 10% greater
than the average of quartz and is negative. Repeated measurements in
the region of 20° to 60° Centigrade showed this coefficient to be 46.6
parts in a million per degree Centigrade. For an accurately constant
frequency one must use a thermostat and heater, the cost of which is
inconsiderable in comparison with the amount saved by the elimination
of frequency-doubling and amplifying stages, and the simplification
of the installation. For simple transmitters one can easily dispense
with temperature control. The frequency change of tourmaline is always
in proportion to the temperature variation.
Fig. 9 shows how
the crystal plate is cut from the actual crystal. The position of the
different axis is the same as in Fig. 5. Most suitable tourmaline crystals
are found in Brazil and South Africa.3
I wish to offer my thanks to the firm of Carl Zeiss, Jena, Germany,
for help in my experiments. In particular, without their great interest
it would have been impossible to overcome the difficulties in construction
of the shortest wave oscillators for 1.2 meters.
1. Converting to the English system and
in terms of frequency, t=146.25f, approximately, where t is thickness
and f is frequency in kilocycles.
The thickness dimension is parallel to the Z axis for tourmaline whereas
it is generally
parallel to either the X
or Y axis for quartz. - EDITOR.
2. Cf. Parsons, "Silvering
Electrodes on Quartz Crystals," QST, March, 1932. - EDITOR.
We understand that the common black variety, known as
schorl, is generally unsuitable. - EDITOR.
Posted April 18, 2013