In the early years of radio technology, spark transmissions ruled the (air) waves. They occupied a relatively wide bandwidth, in what came to be known as the long wave band. The receivers used had a corresponding bandwidth, as ‘wide open’ as the proverbial ‘barn door’. Most were simple detectors without an amplifier stage. Today when you operate an electric light switch you produce a wideband spark that’s audible on some radios as a crackle from long through to short waves. The same occurs with intermittent breaks in cables, high voltage strikes, defective transformers, poorly suppressed electric motors and all kinds of contacts that open and close.
With a suitable receiver, it’s possible to track down the source of these problems. Tests using normal radios are largely unsuccessful, simply because they display restricted bandwidth and are too effective at suppressing short interference pulses. After some research, the best results were obtained with a wideband Audion receiver.
The requirements for this kind of receiver are totally different from normal radio reception: the receiver must have bandwidth as wide as possible, with maximum sensitivity in the longwave region. A further special request: since the wave packets of a single spark are often extremely short, the receiver should integrate them into a longer pulse whose spectrum should lie well inside the audible range. As for the circuit, the audion stage in the collector circuitry detunes the input circuit. To prevent self-oscillation we need to add 10 kΩ resistors. Using an oscilloscope you can see extremely short pulses on the emitter of the BC557 being broadened. The amplitude is frequently sufficient to drive the final amplifier into limiting. A 1 µs long input pulse results in a circa 1 ms long audio pulse in the loudspeaker.