This article discusses the history and gradual development of the Oscilloscope.
The earliest method of creating an image of a waveform was through a laborious and painstaking process of measuring the voltage or current of a spinning rotor at specific points around the axis of the rotor, and noting the measurements taken with a galvanometer. By slowly advancing around the rotor, a general standing wave can be drawn on graphing paper by recording the degrees of rotation and the meter strength at each position.
This process was first partially automated by Jules François Joubert with his step-by-step method of wave form measurement. This consisted of a special single-contact commutator attached to the shaft of a spinning rotor. The contact point could be moved around the rotor following a precise degree indicator scale and the output appearing on a galvanometer, to be hand-graphed by the technician. This process could only produce a very rough waveform approximation since it was formed over a period of several thousand wave cycles, but it was the first step in the science of waveform imaging.
Automatic paper-drawn oscillograph
The device known as the Hospitalier Ondograph was based on this method of wave form measurement. It automatically charged a capacitor from each 100th wave, and discharged the stored energy through a recording galvanometer, with each successive charge of the capacitor being taken from a point a little farther along the wave. (Such wave-form measurements were still averaged over many hundreds of wave cycles but were more accurate than hand-drawn oscillograms.)
To perform a waveform measurement, a photographic slide would be dropped past a window where the light beam emerges, or a continuous roll of motion picture film would be scrolled across the aperture to record the waveform over time. Although the measurements were much more precise than the built-up paper recorders, there was still room for improvement due to having to develop the exposed images before they could be examined.
In the 1920s, a tiny tilting mirror attached to a diaphragm at the apex of a horn provided good response up to a few kHz, perhaps even 10 kHz. A time base, unsynchronized, was provided by a spinning mirror polygon, and a collimated beam of light from an arc lamp projected the waveform onto the lab wall or a screen.
Even earlier, audio applied to a diaphragm on the gas feed to a flame made the flame height vary, and a spinning mirror polygon gave an early glimpse of waveforms.
Moving-paper oscillographs using UV-sensitive paper and advanced mirror galvanometers provided multi-channel recordings in the mid-20th century. Frequency response was into at least the low audio range.
Cathode ray tubes (CRTs) were developed in the late 19th century. At that time, the tubes were intended primarily to demonstrate and explore the physics of electrons (then known as cathode rays). Karl Ferdinand Braun invented the CRT oscilloscope as a physics curiosity in 1897, by applying an oscillating signal to electrically charged deflector plates in a phosphor-coated CRT. Braun tubes were laboratory apparatus, using a cold-cathode emitter and very high voltages (on the order of 20,000 to 30,000 volts). With only vertical deflection applied to the internal plates, the face of the tube was observed through a rotating mirror to provide a horizontal time base. In 1899 Jonathan Zenneck equipped the cathode ray tube with beam-forming plates and used a magnetic field for sweeping the trace.
Early cathode ray tubes had been applied experimentally to laboratory measurements as early as 1919  but suffered from poor stability of the vacuum and the cathode emitters. The application of a thermionic emitter allowed operating voltage to be dropped to a few hundred volts. Western Electric introduced a commercial tube of this type, which relied on a small amount of gas within the tube to assist in focussing the electron beam.
V. K. Zworykin described a permanently sealed, high-vacuum cathode ray tube with a thermionic emitter in 1931. This stable and reproducible component allowed General Radio to manufacture an oscilloscope that was usable outside a laboratory setting.
The first dual-beam oscilloscope was developed in the late 1930s by the British company A.C.Cossor (later acquired by Raytheon). The CRT was not a true double beam type but used a split beam made by placing a third plate between the vertical deflection plates. It was widely used during WWII for the development and servicing of radar equipment. Although extremely useful for examining the performance of pulse circuits it was not calibrated so could not be used as a measuring device. It was, however, useful in producing response curves of IF circuits and consequently a great aid in their accurate alignment.
Allen B. Du Mont Labs. made moving-film cameras, in which continuous film motion provided the time base. Horizontal deflection was probably disabled, although a very slow sweep would have spread phosphor wear. CRTs with P11 phosphor were either standard or available.
The triggered-sweep oscilloscope
Oscilloscopes became a much more useful tool in 1946 when Howard Vollum and Jack Murdock invented the triggered-sweep oscilloscope, Tektronix Model 511. Howard Vollum had first seen such 'scopes in Germany. Before triggered sweep came into use, the horizontal deflection of the oscilloscope beam was controlled by a free-running sawtooth waveform generator. If the period of the horizontal sweep did not match the period of the waveform to be observed, each subsequent trace would start at a different place in the waveform leading to a jumbled display or a moving image on the screen. The sweep could be synchronized with the period of the signal, but then the sweep speed was uncalibrated. Many oscilloscopes had a synchronization feature which fed a signal from the vertical deflection into the sweep generator circuit, but the equivalent of trigger level had at best a narrow range, and trigger polarity was not selectable.
Triggering allows stationary display of a repeating waveform, as multiple repetitions of the waveform are drawn over exactly the same trace on the phosphor screen. A triggered sweep maintains the calibration of sweep speed, making it possible to measure properties of the waveform such as frequency, phase, rise time, and others, that would not otherwise be possible.
More importantly, triggers can occur at varying intervals, and unless too closely spaced, each trigger creates an identical sweep. There is no requirement for a constant-frequency input to obtain stable traces.
During World War II, a few oscilloscopes used for radar development (and a few laboratory oscilloscopes) had so-called driven sweeps. These sweep circuits remained dormant, with the CRT beam cut off, until a drive pulse from an external device unblanked the CRT and started one constant-speed horizontal trace, which could have a calibrated speed, permitting measurement of time intervals. Once the sweep was complete, the sweep circuit blanked the CRT (turned off the beam) and the circuit reset itself, ready for the next drive pulse. The Dumont 248, a commercially available oscilloscope produced in 1945, had this feature.
Long-persistence CRTs, sometimes used in 'scopes for displaying quite-slowly-changing voltages, used a phosphor such as P7, which comprised a double layer. The inner layer fluoresced bright blue from the electron beam, and its light excited a phosphorescent "outer" layer, directly visible inside the envelope (bulb). The latter stored the light, and released it with a yellowish glow with decaying brightness over tens of seconds. This type of phosphor was also used in radar analog PPI CRT displays, which are a graphic decoration (rotating radial light bar) in some TV weather-report scenes.
Triggered-sweep oscilloscopes compare the vertical deflection signal (or rate of change of the signal) with an adjustable threshold, referred to as trigger level. As well, the trigger circuits also recognize the slope direction of the vertical signal when it crosses the threshold—whether the vertical signal is positive-going or negative-going at the crossing. This is called trigger polarity. When the vertical signal crosses the set trigger level and in the desired direction, the trigger circuit unblanks the CRT and starts an accurate linear sweep. Each start can happen at any time after the preceding one (but not too soon) – provided that the preceding sweep is complete, and the sweep circuit has completely reset itself to its initial state. (This dead time can be significant.) During the sweep, the sweep circuit itself ignores sweep-start signals from the trigger-processing circuits.
Having selectable trigger polarity and trigger level, along with the driven sweep, made oscilloscopes into exceptionally valuable and useful test and measurement instruments. Early triggered-sweep oscilloscopes had calibrated time bases, as well as vertical (deflection) amplifiers with calibrated sensitivity. The trace speed across the screen was given in units of time per division of the graticule.
As oscilloscopes have become more powerful over time, enhanced triggering options allow capture and display of more complex waveforms. For example, trigger holdoff is a feature in most modern oscilloscopes that can be used to define a certain period following a trigger during which the oscilloscope will not trigger again. This makes it easier to establish a stable view of a waveform with multiple edges which would otherwise cause another trigger.
Vollum and Murdock went on to found Tektronix, the first manufacturer of calibrated oscilloscopes (which included a graticule on the screen and produced plots with calibrated scales on the axes of the screen). Later developments by Tektronix included the development of multiple-trace oscilloscopes for comparing signals either by time-multiplexing (via chopping or trace alternation) or by the presence of multiple electron guns in the tube. In 1963, Tektronix introduced the Direct View Bistable Storage Tube (DVBST), which allowed observing single pulse waveforms rather than (as previously) only repeating wave forms. Using micro-channel plates, a variety of secondary-emission electron multiplier inside the CRT and behind the faceplate, the most-advanced analog oscilloscopes (for example, the Tek 7104 mainframe) could display a visible trace (or allow the photography) of a single-shot event even when running at extremely fast sweep speeds. This 'scope went to 1 GHz.
In vacuum-tube 'scopes made by Tektronix, the vertical amplifier's delay line was a long frame, L-shaped for space reasons, that carried several dozen discrete inductors and a corresponding number of low-capacitance adjustable ("trimmer") cylindrical capacitors. These 'scopes had plug-in vertical input channels. For adjusting the delay line capacitors, a high-pressure gas-filled mercury-wetted reed switch created extremely-fast-rise pulses which went directly to the later stages of the vertical amplifier. With a fast sweep, any misadjustment created a dip or bump, and touching a capacitor made its local part of the waveform change. Adjusting the capacitor made its bump disappear. Eventually, a flat top resulted.
Vacuum-tube output stages in early wideband 'scopes used radio transmitting tubes, but they consumed a lot of power. Picofarads of capacitance to ground limited bandwidth. A better design, called a distributed amplifier, used multiple tubes, but their inputs (control grids) were connected along a tapped L-C delay line, so the tubes' input capacitances became part of the delay line. As well, their outputs (plates/anodes) were likewise connected to another tapped delay line, its output feeding the deflection plates. (This amplifier was push-pull, so there were four delay lines, two for input, and two for output.)
The first Digital Storage Oscilloscope (DSO) was invented by Walter LeCroy (who founded the LeCroy Corporation, based in New York, USA) after producing high-speed digitizers for the research center CERN in Switzerland. LeCroy remains one of the three largest manufacturers of oscilloscopes in the world.
Starting in the 1980s, digital oscilloscopes became prevalent. Digital storage oscilloscopes use a fast analog-to-digital converter and memory chips to record and show a digital representation of a waveform, yielding much more flexibility for triggering, analysis, and display than is possible with a classic analog oscilloscope. Unlike its analog predecessor, the digital storage oscilloscope can show pre-trigger events, opening another dimension to the recording of rare or intermittent events and troubleshooting of electronic glitches. As of 2006 most new oscilloscopes (aside from education and a few niche markets) are digital.
Digital scopes rely on effective use of the installed memory and trigger functions: not enough memory and the user will miss the events they want to examine; if the scope has a large memory but does not trigger as desired, the user will have difficulty finding the event.
1 ^ Hawkins (1917, p. 1844) Fig. 2589
Abramson, Albert (1995), Zworykin, pioneer of television, University of Illinois Press, ISBN 0-252-02104-5