ON THE RECORDING OF CHANGING PROCESSES*

R. Vieweg, Berlin

Introduction

The recording of changing processes is as old as precision measurement physics. The characteristics of many physical phenomena make it necessary, instead of observing individual moments, to carry out a continuous recording of a prolonged process. In this, the time over which the process unfolds plays an extremely important role. Processes that occur very rapidly or very slowly cannot be studied by any method other than their continuous recording. Therefore the field of recording instruments encompasses every possible kind of process, from the most prolonged to the extremely rapid; the writing pen and the cathode oscillograph are the two limits defining the variety of instruments that record changing phenomena.

1. Classification of methods of recording

Depending on the method used, three main groups of recording instruments may be distinguished: the first group operates with the aid of mechanical recording devices; this group includes recorders proper—instruments that make a recording with ink. The second group includes instruments in which a light beam is used, giving the course of the curve under investigation on a photographic plate. Finally, the third group makes use of an electron beam acting on a fluorescent screen or else on a photographic layer.

2. Recording speed

In choosing one or another type of recording, the recording speed, the setting time of the recording part, and the force that must be applied to the writing element are of great importance. The advances achieved in recent years in the field of recording instruments are to a considerable extent connected with the development of devices that make it possible to reduce more and more the magnitude of the force acting on the recording element. One of the most important—

* R. Vieweg, Z. techn. Physik, 14, 441, 1933, translated by N. N. Malov.

The essence of these principles consists in the fact that the measured process itself is used not for direct action on the writing device, which is powered by an auxiliary source of energy, but only for controlling the latter, owing to which almost no energy of the process under investigation has to be expended. A broad field of amplifiers is based on this principle. At the present time a whole series of instruments containing electron tubes has been constructed, with the aid of which it is possible to record processes inaccessible to ordinary measuring instruments. We shall not dwell here on the details of amplification technique and shall confine ourselves to only one example of a recording instrument with an auxiliary device—a special self-recorder called a compensograph.

Fig. 1. Compensograph (schematic).
\(N\)—zero galvanometer, \(F\)—falling bow, \(A\)—Bugel drive, \(B\)—braking lever, \(L\)—runners, \(P\)—drive for the potentiometer, \(R\)—potentiometer (annular tube), \(R_1\), \(R_2\), \(R_3\)—bridge resistances, \(R_k\)—temperature-dependent resistance, \(T\)—thermoelement, \(M\)—milliammeter.

1. Siemens-Halske compensograph. According to Fig. 1, the voltage being measured, whose source is a thermoelement, automatically switches in a compensation voltage equal in magnitude but opposite in direction. To monitor the equality of the voltages, a zero galvanometer is used, the pointer of which is periodically lowered by means of a bow (Bugel) every 3 seconds. If it proves to be deflected from the zero position, it is pressed against a lever that releases a spring mechanism. Owing to the special construction of the bow, the duration of release of the mechanism is proportional to the deflection of the instrument. The spring mechanism, by means of a drive, rotates a tube \(R\) bent into a circle, which serves as a compensating resistance, until equality of the voltages is restored. The writing pen is connected with this tube. Thus

thus the compensated measured voltage acts on the instrument by means of a contact arc; subsequently, however, the instrument operates as an ordinary recorder, set in motion by a small synchronous motor. Thus the measured voltage serves only to control the mechanism driven by an auxiliary source of energy. The advantages of this method of control may be judged from the fact that, with full deflection of an ordinary recorder with a rotating coil, a direct current of about 0.5 μA may be observed, whereas the compensograph makes it possible to record currents of the order of 1 μA. If this instrument is connected with limiting contacts, it can also be used for regulating the process being recorded.

An even greater gain is achieved in recording by means of the bolometer proposed by Zelley1.

1. Use of a vibrating-reed instrument for recording frequency. The problem of continuous measurement of frequency with sufficient accuracy has become especially important owing to the expansion of parallel operation of power installations. The importance of this problem is clear from the fact that at present in some networks it proves possible to use synchronous clocks with a high degree of accuracy. In Fig. 2 the frequency of the current in the Berlin municipal network is shown, recorded by Hartmann and Braun’s precision self-recording frequency meter. The constancy of the frequency reaches 0.05 hertz, i.e. 0.1%. At night, as the upper curve shows, it is still higher. Of course, with such recording very high demands are imposed on the sensitivity of the self-recording instrument.

As regards vibrating-reed instruments, it should also be pointed out that such instruments have now been technically developed for the simultaneous recording, on one chart, of several processes; an example of such an instrument is a six-color recorder, simultaneously recording six different processes in different colors. In this case only one measuring instrument (of the Desprez type or a dynamic one) is used, which is connected in succession into different sections of the circuit.

3. TIME OF SETTING OF THE PEN OF A SELF-RECORDING INSTRUMENT

As regards the time of setting of the pen of self-recording instruments, it is usually close to 1 sec.; however, it has been possible to reduce it to 0.2 sec. There also exist electrical instruments, for example electrostatic ones, possessing a considerable setting time.

For further increasing the sensitivity of self-recording instruments, attempts have been made to reduce the friction of the pen on the paper by means of additional vibrations of the pen. The pen itself plays an extremely large role in determining the sensitivity of self-recording instruments. The well-known difficulties that existed in instruments of old designs, for example thermographs or barographs, have in new designs been eliminated to a considerable extent thanks to proper

...to the correct choice of paper, pen, and method of supplying the ink. An extremely simple, but very reliable, method of operation, making it possible to obtain a very fine trace, is that proposed by Keil: a capillary needle of a syringe for injection is used, while a razor blade is used as the steel pen.

4. Movement of the Paper

Depending on the operating conditions of the instrument, the speed of movement of the paper may vary considerably, from a few millimeters per hour to one meter per second. Some instruments operate with automatic switching of speed depending on the character of the process being recorded. An example is the “distortion recorder,” used for analyzing disturbances in the regime of power networks and developed by various large

Fig. 2. Frequency in the Berlin electrical network.

firms. Under normal conditions it operates at a paper-motion speed of about 20 mm per hour. When an overvoltage or a sharp voltage drop occurs, the speed of paper motion is automatically switched (by means of a relay) to a higher one—up to 20 mm per second. The switching time depends somewhat on the intensity of the disturbance and amounts to from 35 to 200 milliseconds.

Of very great importance is the question of the uniformity of the movement of the paper, especially in chronographs. In these instruments the displacement is produced either by means of a precision spring mechanism, or by means of a synchronous motor controlled, for example, by quartz. In clock mechanisms, fine regulation is applied by means of Hipp’s braking and an oscillating plate; the advantages of this method are the possibility of establishing a very uniform regulating force and practically uninterrupted motion. The applications of chronographs are very numerous: from chronography proper—the registration of time—to ballistics; in addition, they find application in physio-

logy and psychology, for example, in establishing the speed of an irritating stimulus, and also in sporting competitions. The newest chronographs, equipped with a self-writing pen, make it possible to obtain a record with an accuracy down to one thousandth of a second. Worth mentioning are the quartz clocks of Scheibe and Adelsberger[^14], whose rate accuracy reaches \(\pm 0.001\) sec. per day. They also provide standard frequencies, constant with an accuracy up to \(\pm 4 \cdot 10^{-9}\).

5. Recording methods using scratching

For a long time in physics a smoked glass plate has been used, on which, for example, the vibrations of a tuning fork were recorded. Although at present this method is used comparatively little, it should not be thought that it has altogether lost its significance. Thus, for example, recently Tomlinson[^2] described a chronograph of very high accuracy, in which time marks are made by needles on a smoked drum; the accuracy of this chronograph reaches 0.0002 sec.

Fig. 3. Acceleration meter with a scratching device; natural frequency 15–20 hertz.

Fig. 4. Diagram obtained with a scratching diamond.

The method of diamond recording, in which a diamond needle scratches a celluloid, glass, or steel plate, has recently been applied in aviation by Seewald[^3]. The diamond draws an extremely fine line; the amplitudes of the oscillations of the curve amount to only fractions of a millimeter, but thanks to the exceptional cleanness of the record the curve can be enlarged 100–500 times. Figure 3 shows the scheme of an acceleration meter operating on this principle and giving a record on a glass cylinder. The diagram wound onto the cylinder in a spiral, owing to the fineness of the recording, occupies very little space, so that the small surface of the registering cylinder proves equivalent to many hundreds of meters of ordinary recording paper. Such an instrument is very suitable for studying flight dynamics, where, in addition to measuring acceleration, one can also carry out an investigation of defor-

...mations, naturally, give only a very small displacement of the writing stylus. Fig. 4 shows the acceleration curve during flight in windy weather. The ordinates of the curve have been enlarged approximately 100 times.

6. Recorders using a light beam

In the first place one should mention string oscillographs, which have recently undergone considerable technical improvements. At present there exist strings giving a deflection of 1 mm at a current of 2 mA and with the natural frequency of the string about 20 kilohertz. With a reduced natural frequency (1.2 kilohertz) it is possible to construct strings giving a deflection of 1 mm at a current of only 20 μA. The magnets of oscillographs are at present made permanent, the optics have been considerably improved, and it has become possible to observe the curve even during photography.

Fig. 5. Optical multiple recorder.

Another example is the optical multiple recorder of the firm “Askania,” shown schematically in Fig. 5. The figure shows the light sources, the slit system, the optics, and the mirrors. The curve is obtained on a rotating drum carrying photographic paper; time marks are made by means of a glow-discharge lamp. The mirrors of the measuring system are fastened to the membranes of an aneroid. This instrument is used for recording the breathing of pilots flying at great altitudes. It should be noted that with mechanical methods of recording it was not possible here to obtain good results. The advantages of the optical system are its high sensitivity and insignificant weight.

7. Application of piezoelectric forces

In designing all instruments that record a process by means of a light beam, as well as in designing ordinary oscillographs, a major difficulty is the need to preserve the minimum dimensions of the mirror connected with the measuring system, which is dictated by mechanical considerations.

But when the size of the mirror is reduced, the intensity of the light beam reflected by it falls. In this case it is sometimes possible to overcome the indicated difficulty, which in essence amounts to the fact that the action of the process under investigation on the system carrying the mirror may prove too weak. Fig. 6 shows

Fig. 6. Piezoelectric oscillograph.

Filippov’s piezoelectric oscillograph.^4 Its most important part consists of two flat quartz rods, provided with plates and arranged in the form of the recumbent letter V. Mirrors are attached to the narrow sides of these rods. If an electric field is applied to the electrodes, piezoelectric forces arise in the quartz, causing motion of the mirror. The magnitude of these forces considerably exceeds the forces developing in ordinary oscillographs, owing to which the mirror of the piezoelectric oscillograph can be made considerably larger, and the oscillogram, recorded by a stronger beam of light, will be much more distinct. The actual changes in the length of the quartz are of the order of 0.001 mm, so that large magnifications have to be used. This instrument is used in the investigation of the form of high-voltage curves, since several thousand volts are freely withstood by quartz. The natural frequency of such an oscillograph, which determines, as is known, the range of frequencies recorded without distortion, can be brought up to 16 kilohertz.

Fig. 7. Course of pressure in an internal-combustion engine, recorded by a piezoelectric indicator.

Piezoelectric forces in recent years have repeatedly been applied for various electrical measurements, and also for

registration. Thus, for example, in the Kluge and Link indicator⁵ the pressure in the ignition chamber of an engine acts on quartz; in the latter a piezoelectric voltage arises, which, after preliminary amplification in a tube voltmeter, is recorded by a Braun tube. Similar diagrams are shown in Fig. 7. The speed of motion of the tape reached 70 cm per second. The upper oscillogram was taken with twice the sensitivity of the instrument compared with the lower diagram, on which, in the central part, the course of the pressure during faulty ignition is visible.

8. Use of the electron beam

After the invention of the Braun tube⁶, despite the classical studies of high-frequency processes carried out by Zenneck, the tube at first found almost no application. However, after its applicability in television technology became clear, a number of firms arose that manufactured Braun tubes, which thus became inexpensive, readily available instruments. One of the most important parts of a cathode oscillograph is the system that gives periodic deflections of the electron beam proportional to time: the system of sweep oscillations (Kippschwingungsanordnung). The advances of recent years have considerably improved the proportionality of the deflection and simplified the synchronizing devices for recording periodic processes. In sweep circuits an electron tube is used as the resistance through which charging takes place, and for sweeping a thyratron is used, i.e. an ionic tube provided with a control grid. Designs of this kind have now been developed so well that they can be used even under operating conditions. The advantage of the thyratron over the glow-discharge tube formerly used consists chiefly in the fact that the control voltage is only fractions of a volt, whereas for a glow-discharge tube hundreds of volts are required. The sweep frequency obtained in such an instrument can reach 300 kilohertz.

Sweep circuits are also used for recording very rapid processes (lasting 1 microsecond or less), and also for the registration of non-periodic, single processes. The most technically perfect is the so-called high-power cathode oscillograph. When working with it, the most difficult task is the exact recording of processes that arise arbitrarily, independently of the experimenter’s will. Lightning may serve as an example of such processes. In this case the most difficult thing is to construct a sweep relay that operates without distortion. Usually in this case, for deflection, it is no longer the charge of a capacitor that is used, but a very brief discharge of the voltage under investigation.

In Germany, high-power cathode oscillographs were constructed by Rogowski⁷, and also by Matthies. Fig. 8 shows the circuit

of a cathode oscillograph with a cold cathode. The cathode ray, excited by a voltage of the order of 50 kV, passes through a preliminary concentrating coil and a narrow diaphragm into a rarefied space free from the auxiliary field producing the discharge; here it enters a blocking capacitor, which eliminates preliminary illumination of the screen. It then passes through plates that produce a time deflection and a deflection caused by the process under study, and finally reaches a fluorescent screen or a photosensitive layer. Preliminary concentration and blocking of the beam are very important achievements of oscillograph technology, very substantially increasing their resolving power. A typical cathode oscillogram of a normal wave is shown in Fig. 9. For research purposes the rise and fall of the voltage surge used are normalized, and in such investigations the cathode oscillograph becomes an indispensable control instrument. The wave shown in Fig. 9 was recorded at a maximum voltage equal to 10 kV. The rise time to half the maximum value reaches 1 microsecond, and the time during which the voltage remains greater than half the maximum value is about 50 microseconds. The time marks are the oscillations of a three-hundred-meter tube generator, so that the distance between two maximum amplitudes of the lower curve corresponds to 1 microsecond.

Labels in Fig. 8:

• Discharge chamber
• Cathode
• Glass tube
• Concentrating coil
• Tube with diaphragms
• Main diaphragm
• To pump
• Movable diaphragm
• Beam-blocking chamber
• Main collecting coil
• Beam-deflection chamber
• Time plates
• Process plates
• To pump
• Observation window
• Screen
• Recording chamber

Fig. 8. Cathode oscillograph according to Rogowski (schematic).

In Fig. 10 a cathode oscillograph is shown, allowing one to pro-

make it possible to record four processes simultaneously. The cathode beam, by means of blends, is divided into separate beams, as was proposed by Knoll⁸. In each of the two tubes two electron beams are produced (one of the tubes is shown in section in the figure).

Processes whose frequency does not exceed 10 kilocycles can be recorded by means of a mechanical time sweep. In addition, it is possible, with the aid of a Lenard window, to let the beam fall on the photosensitive layer of the lower drum, i.e., the range of application of this oscillograph is further extended. Finally, a combination of both methods at once is possible, in which the low-frequency process will be recorded in one part of the instrument, while

Fig. 9. Normal wave.

Fig. 10. Fourfold cathode oscillograph.
a — discharge tubes, b — anode blend, c — locking plates, d — locking blend, e — zero adjustment, f — concentrating cathode, g — process plates, h — time plates, i — Lenard window with locking grid, k — fluorescent screen, l — drum for recording through the Lenard window, m — drum for photographing the screen, n — optics for photographing the screen, o — path of the cathode beam, p — path of the light beam.

at the same time, as with a tipping relay, the rapid process superposed on it will be registered in another part of the instrument. The recording speeds attainable with the cathode oscillograph are practically unlimited; at present speeds have been attained—

growth on the order of 60,000 and even 100,000 km/sec. Considerably greater difficulties are presented by the questions of time deflection and beam control. However, even now the field of application of the cathode oscillograph is broader than that of all other measuring instruments. Yet up to the present time the cathode oscillograph is still not as simple an instrument as other measuring devices, since the need for a certain skill in working with it somewhat hinders its wide dissemination.

One could point to a whole series of further applications of the cathode oscillograph, such as, for example, Binder’s proposed use of the oscillograph for the direct measurement of high voltages without the aid of a voltage divider.

9. Special methods

Short-duration varying processes, such as, for example, overvoltages, can also be studied with the aid of the stepped spark gap proposed by Binder and Geine[^9]. This method, of course, is inferior to the cathode oscillograph, but is of interest as an attempt to revive the old spark chronographs.

Fig. 11. Clidonogram.

Fig. 12. Positive Lichtenberg figure.

The possibility of recording short-duration overvoltages with the aid of a clidonograph has long been proved, i.e. by developing the pattern of an electric discharge on a photographic layer. With the aid of calibration curves, the magnitude of the voltage can be judged from the diameter of the figures obtained. It is also possible to connect a clidonograph with a slowly moving film into a high-voltage network and observe the moments of formation of the Lichtenberg figures and their magnitude, which makes it possible to determine the overvoltages that occurred in the circuit. In Fig. 11 overvoltages in a 10-kilovolt ...

...of a network during a thunderstorm, recorded with the Siemens-Schuckert klydonograph[^10].

Garthe[^11] in Hanover succeeded, by reducing the capacitance of the system (practically—by introducing an intermediate layer of air), in obtaining considerably larger and more distinct figures, thanks to which the accuracy of measurements could be brought to 3% instead of the usual 20%. In Fig. 12 a positive figure corresponding to 45 kV is given. The duration of the voltage required for the formation of the figures is extremely small—of the order of hundred-millionths of a second. The figure in Fig. 12 was obtained within one ten-millionth of a second.

10. Photographic Means

It is obvious that advances in obtaining highly sensitive films, plates, and paper have a very favorable effect on the technique of recording methods. Here one should also mention television, in which, when an image is formed from separate points, the exposure time can be reduced to millionths of a second, which makes it possible to raise the question of constructing new, very sophisticated instruments.

Finally, photographs as such are also a means of recording variable processes. Indeed, there is a whole series of very important physical photographs, such as, for example, photographs of lightning and of the arc. Further, one should mention sound recording, which can be carried out by scratching a plate or by photographing a Kerr condenser, as well as by other methods.

11. Statistical Methods

These include the summation of variable quantities by counting mechanisms. For example, the summation of discharges in a Geiger counter by means of a thyratron (Kluge and Eger[^12]) makes it possible to construct a high-speed relay capable of registering up to 50 pulses per second. If the counter is photographed over short intervals of time, then the registration of individual pulses becomes possible.

Finally, the so-called time transformer of Steenbeck and Strigel[^13] deserves mention. It makes it possible, for example, to study the statistical distribution of the moments of breakdown of a spark gap. Of course, one can take a huge number of cathode oscillograms and subject them to statistical processing; however, the time transformer works differently: during short intervals of time (microseconds) a well-calibrated lamp and a system of capacitors are charged (the charging is produced by a quite definite rectangular wave); subsequently the system discharges according to the usual laws of capacitor discharge over time intervals of the order of a second, the discharge being marked by an ordinary...

...by a chronograph, the pen of which is connected into the discharge circuit. The time transformer makes it possible to determine the interval-puncturing time with an accuracy of up to 10%.

Literature

1. The paper is being published in Z. Techn. Physik.
2. G. A. Tomlinson, Z. Instrumentenkde 51, 271, 1931.
3. F. Seewald, Maschinenbau 10, 725, 1931.
4. W. v. Philippoff, Elektrotechn. Z. 53, 405, 1932.
5. J. Kluge und H. E. Linckh, Forschung a. d. Geb. Ingenieurwesens 4, 177, 1933.
6. Alberti, Über Braun’sche Kathodenstrahlröhren und ihre Anwendung. Berlin, Jul. Springer, 1932; there is a Russian translation, GTTI, 1933; Manfred v. Ardenne, Die Kathodenstrahlröhre und ihre Anwendung in der Schwachstromtechnik. Berlin, Jul. Springer, 1933.
7. W. Rogowski, Elektrotechn. Z. 51, H. 17, 1933.
8. Elektrotechn. Z. 53, 1101, 1932.
9. L. Binder und H. Heyne, Arch. Elektrochem. 24, 469, 1930.
10. D. Müller-Hillebrand, Siemens-Z. 7, 547, 1927.
11. F. Hartje, Elektrotechn. Z. 53, 939, 1932.
12. J. Kluge und R. Jäger, Z. Instrumentenkde 52, 229, 1932.
13. M. Steenbeck und R. Strigel, Arch. Elektrotechn. 26, 831, 1923.
14. A. Scheibe u. U. Adlersberger, Phys. Z. 33, 835, 1932; Hochfrequ. u. Elektroakustik 43, 37, 1934.