Simple Battery Operated Spark Transmitter
The step up transformer takes the battery voltage and boosts it to many thousands of volts. Since transformers do not work on direct current some form of current interrupter is required. This takes the form of a simple clapper on the end of the coil, actuated by the primary current flow's magnetic field. In practice the interruptions occur, for explanation purposes, about 200 times a second. The collapsing magnetic field caused by the interruption to primary current flow cuts through the secondary winding turns inducing a voltage. The voltage induced is directly proportional to the "turns ratio" between the two windings.
In the case of the apparatus shown in the image, the core is a bundle of soft iron 22 SWG (Standard Wire Gauge) wires with a diameter of 2" and a length of 18". Primary winding is 300-400 turns of #12 Standard Wire Gauge (SWG) copper wire wrapped on the core. It is wound bobbin style around the soft iron core. The secondary coil is not wound smoothly back and forth (bobbin style) across the primary windings, but is done using many insulated disk shaped coils, joined together as illustrated on the left. This prevented high voltage breakdown within the transformer as the high voltages generated are now distributed along the width of the secondary coil, and not through the narrower diameter.
Thus the resonant antenna produces a burst of radio frequency energy at each arc of the spark gap. (A crude analogy would be a mallet [the energy pulse] hitting a gong [creating a diminishing tone].) The power induced into the antenna circuit is proportional to the voltage charge on the capacitors which, in practice, is governed by the spark gap spacing. A larger spacing allows the voltage on the capacitors to reach a higher value before the spark gap is overcome.
Of course the arcing gap momentarily shorts out the high voltage transformer. To keep any of the radio energy out of the high voltage transmforer some small choke inductances are connected on each lead. Energy induced into the antenna from the discharging capacitors will also be coupled back into the spark transmitter. The fix for this is to install a second spark gap, called a quench gap in series with the capacitor side of the RF transformer. The 'main bang' of the transmitter will energy will easily jump the quench gap, but any energy coming the other way will be much less and unable to jump the quench gap.
Hook a couple of capacitors and a coil of wire as shown in the top diagrams. The capacitors shown here are made using glass jars, a layer of tin foil inside and a similar layer on the outside. The glass serves as an insulating medium between the two layers and, as a result, can handle the high voltages. The value of the capacitor would typically be in the order of 0.1 mfd. The radio frequency (RF) coil/transformer will couple the stored energy to the antenna system.
Applying battery to the primary circuit causes current to flow and, due to the interrupter, be cycled off and on about 200 times a second. The voltage generated, due to transformer's step up action, on the secondary side is several thousand volts when the interrupter opens the circuit since the rate of change of the magnetic field is much faster collapsing. When the interrupter closes the circuit the magnetic field's rate of change is much slower due to the inductive effect of the primary coil. The resulting high voltage on the secondary alternates at the same frequency as the interrupter buzzes.
If no coil and capacitors were across the gap, the spark could be several inches long--hence references to a 10" spark coil transmitter in the old books. In practice the capacitors initially act as a short circuit across the coil's secondary, the voltage rising as the value of the capacitors and winding resistance of the spark coil's secondary allow. The applied high voltage isn't available for the length of time required, due to the quick action of the interrupter, and as a result the capacitors do not fully charge. Thus high voltage is quite a bit less than the peak supplied by the transformer and as a result the spark gap must be decreased to the order of 1/4 to 1/2 inch before a spark is obtained.
The antenna system has distributed capacity to earth and with the inductance of the coupling coil creates a resonant circuit. The amount of inductance is adjusted, by means of taps on the coil, to resonate at a particular frequency. In those days the wavelength would be somewhere in the region of 600 to 2000 meters (500 kHz to 150 kHz).
On the rise of the high voltage, the capacitors charge up via the antenna coil until the flash over voltage of the spark gap is realized. The spark appears, for all intents and purposes, as short circuit. The capacitors are now able to discharge their energy into the resonant antenna via the coupling coil, then the coil back into the capacitors via the still arcing gap. This back and forth transfer of the energy will continue until the arc finishes, or the resonant circuit's energy is dissipated in overcoming the resistance losses. In practice the resonant circuit's energy will diminish before the arc is extinguished due to the fall of secondary voltage or else stopped completely by the installation of a quenching gap.
Thus the resonant antenna produces a burst of radio
frequency energy at each arc of the spark gap. (A crude analogy would be a
mallet [the energy pulse] hitting a gong [creating a diminishing tone].)
The power induced into the antenna circuit is proportional to the voltage
charge on the capacitors which, in practice, is governed by the spark gap
spacing. A larger spacing allows the voltage on the capacitors to reach a
higher value before the spark gap is overcome.
A second spark gap, called the safety gap, was usually provided. This second gap was fixed and wider than the first, and served to protect the capacitor bank from break down due to over voltage.
Energy induced into the antenna from the discharging capacitors will also be coupled back into the spark transmitter. This energy would be wasted. To improve the efficiency by some 20-40% a second spark gap, called a quench gap, was placed in series with the capacitor side of the RF transformer. The 'main bang' of the transmitter will energy will easily jump the quench gap, but any energy coming the other way will be much less and thus be unable to jump the quench's gap.
A simple amplitude modulation receiver, such as a crystal set, tuned to the frequency of the transmitter, would hear a tone, the frequency of which is equal to that of the arc (400 Hz not 200 Hz), which in turn is governed by the frequency of the interrupter.
This unit would be quite similar to the transmitters carried on the first wireless equipped vessels. The only changes would be the insertion of a telegraph key in series with the battery which the operator would manipulate to form the Morse characters, and suitable coupling of the RF coil to an antenna.
This transmitter is inefficient, even by turn of the century standards.
As an aside, if the RF coil and capacitor are removed, put a spark plug in place of the spark gap, we will have a car ignition system. That explains why a passing car can often be heard on a radio--its ignition system is a wireless transmitter. Note that the missing coil and capacitor are still there--the ignition wiring has its own distributed inductance and capacitance.
The original five British Columbia coast stations were fitted with Shoemaker, instead of Marconi, apparatus at the start. Canadian Marconi felt they had an iron clad contract to build and man any Canadian Coast Station. Marconi already had the Canadian East Coast stations and also wanted the west coast stations. Dominion government went ahead built the stations, installed Shoemaker equipment and their own employees. Canadian Marconi threatened law suit. Eventually it all got smoothed over and by 1910 or so the Shoemaker apparatus had been removed and replaced by the better Marconi on the west coast stations.
Keep in mind the radiation from this new fangled science may be affecting your body--see this from 1911.
The professional images are from "An Elementary Manual of Radio-Telegraphy and Radio-Telephony, 3rd edition by J.A.Fleming/Longmans, Green & Co. London 1916)
From the 1939 Encyclopedia Americana:
"In making a condenser for radiography, the glass plate type is recommended as inexpensive and durable, and also much lighter than oil-immersed types. Photographer's negative glass, tinfoil and shellac are the materials. The foil should be cut to the required size (6x8 inches is convenient), and carefully smoothed to take out all wrinkles. The glass should be cleaned with alcohol and coated with shellac, then covered with the foil, and rolled or "squeegeed" so as to be perfectly smooth. In assembling the plates lugs should be placed between them. A unit may be made of 10 plates which are bound together with wire or suitable tape, and immersed for one hour in a bath of equal parts of hot melted rosin and beeswax, then allowed to drain and dry. This gives a condenser unit thoroughly moisture proof, with a capacity of 0.01 microfarad, which is suitable for the ordinary half-kilowatt wireless transformer for the standard 200 meter wave-length. It is good practice to make four such units, placing two multiple sets of two in each series; this reduces the strain on the condensers, without altering the capacity. They may be mounted in substantial open-side wooden boxes to protect the plates from injury."