US2465082A - Damped - Google Patents

Description

March 22, 1949.

ELECTRONIC AMPLIFIER FOR GALVANOMETER OSCILLOGRAPHS Filed Aug. 2, 1944 K. R. GEISER 2,465,082

2 Sheets-Shasta],

DEFLECTION AMPLITUDE DAMPED 53 '3 a. zzno LINE SE 0 Q Y O O 8 g Fig.5.

Z 9,ooo-

Q Fi .7 F336. 2

b WW 5 1) y His Attorney.

March 22, 1949. K. R. GEISER 2,465,032v

ELECTRONIC AMPLIFIER FOR GALVANOMETER OSCILLOGRAPHS Filed Aug. 2, 1944 2 Sheets-Sheet 2 Fig-.9.

PHASE ANGLE ADVANCE I AVERAGE CURRENT ANODE VOLTAGE E s, GRID REES n B 5 u 3 Q s Inventor: a Kenneth RGeiser,

u 300 EL 3:: 7 PLATE VOLTS E by 0 50 I00 I50 zoo 5 Attorney.

wave rectifier; Fig.

Patented Mar. 22, 1949 acac'rnomo ETER 0801 Kenneth R. Geiser, Schenectady, N.

AMPLIFIER FOR GALVANOM- LLOGRAPHS Y., assignor to General Electric Company, a corporation of New York Application August 2, 1944, Serial No. 547,794

2 Claims.

My invention relates primarily to a novel method of energizing galvanometers of the oscillograph typ and to amplifier arrangements therefor which enable the use of the oscillograph type of galvanometer to be extended to the measurement and study of rapidly varying feeble phenomena with much less amplifier equipment than has heretofore been considered necessary. In carrying my invention into effect, I energize the galvanometer by unidirectional current pulsations of a frequency well above the natural frequency of vibration of its moving element, and vary the magnitude of the current pulsations in response to the phenomena to be investigated. Such magnitude variation may be faithfully reproduced up to frequencies usually employed with oscillograph galvanometers.

The features of my invention which are believed to be novel and patentable will be pointed out in the claims appended hereto. For a better understanding of my invention, reference is made in the following description to the accompanying drawings in which Fig. 1 represents portions of an oscillograph galvanometer of the type used in my invention; Fig. 2 represents the deflectionfrequency characteristics of an oscillograph galvanometer energized by alternating current; Fig. 3 indicates an oscillograph galvanometer energized from an A.-C. source through a half- 4 represents the deflectionfrequency response of an oscillograph galvanometer when energized by unidirectional current pulsations; Fig. 5 represents a pulsating unidirectional current curve and the corresponding response curve of an oscillograph galvanometer when used in accordance with my method; Figs. 6, and 12 show different ways of amplifying feeble measurement currents for recording with an oscillograph type galvanometer in accordance with my invention using a gas-filled tube amplifier; Figs. 7, 8 and 9 will be referred to in explaining how the tube output current of Fig. 6 varies in response to changes in phase relation between the plate and grid voltages; Fig. 11 will be referred to in explaining the operation of Fig. 10; Fig. 13 represents a high frequency voltage measuring scheme using an oscillograph galvanometer and full-wave rectifier; Fig. 14 represents the use of a high vacuum tube in a system embodying my invention; and Fig. 15 are explanatory curves of the operation of Fig. 14.

The usual Duddell type oscillograph galvanometer consists of means for producing a strong unidirectional magnetic field between pole pieces 1 and 2, Fig. 1. In this field and at right angles thereto, is supported a bifilar type conductor suspension 3 which is the equivalent of a coil having one turn and consists of a fine conductor wire looped over a pulley 4 and secured at 5, and stretched taut by a lever B and spring I. The

pair of wires are held parallel in the field between bridges 8 and 9 and the plane of the wires is normally parallel to the direction of the field in which located. The current to be measured or investigated is sent through the loop, causing a motor action which causes the plane of the two wires to turn in a direction dependent on the current direction and by an amount proportional to the current strength. This motion is traced or recorded by a light beam photographic method by utilizing a tiny mirror H] which is fastened to the two wires between the bridges. When the current is removed, the tension of the wires returns the mirror to zero position. The moving element is contained in an oil-filled damping cell having a transparent window ll therein to accommodate the light beam. Reference is made to United States Patent 1,800,018 for a more detailed description of such galvanometers. The Duddell galvanometer is essentially a current sensitive instrument having low resistance and a high natural period of torsional oscillation. Practicable galvanometers of this type require a peak current of the order of milliamperes for full deflection and are capable of following current variations or reversals up to about 2500 cycles per second.

There are many minute vibratory measurements which cannot be directly converted into a sufficient current change to afford a satisfactory measurement deflection of this type of galvanometer. One example of this are measurements taken with wire resistance type strain gauges of the type described in United States Letters Patent No. 2,292,549 to Simmons. Where such measurements are to be recorded with the galvanometer type oscillograph, it becomes necessary to employ amplifier apparatus between the measurement circuit and the oscillograph and heretofore this has consiststed of at least one stage of voltage amplification and a power stage employing a tube capable of supplying the high current necessary for good galvanometer deflections.

An important object of my invention is to reduce the amplifier equipment necessary for such measurements and in accomplishing this result, I energize the galvanometer with unidirectional current pulsations of a frequency higher than its natural period of vibration and modulate this high frequency supply in accordance with the desired measurements.

The vibration characteristics of the oscillograph type galvanometer when energized by constant voltage alternating current are represented in Fig. 2. If the galvanometer of Fig. 1 be energized by a constant voltage alternating current and we vary the frequency from zero up to 10,000 cycles per second, the amplitude of vibration of the galvanometer may be represented by the 3 curves of Fig. 2. With the usual damping the deflection amplitude of the galvanometer will remain constant and follow the oscillations of the alternating current up to about 3000 cycles per second. At higher frequencies the amplitude drops oil as shown by the full lines and becomes practically zero at about 6000 cycles per second. If the galvanometer were undamped as by removing the damping oil from the cell, the amplitude of oscillations would increase in the vicinity of the natural undamped period of vibration of the suspension as represented by the dotted line curves of Fig. 2. The resonant frequency occurs at about 4000 cycles per second, and again the oscillation amplitude becomes practically zero at about 6000 cycles per second.

If, now, I place a rectifier in the galvanometer circuit as represented in Fig. 3 and-repeat th test with the galvanometer damped, I have discovered that I can obtain th deflection amplitude response represented in Fig. 4. It is noted that the deflection is only on one side of zero due to the fact that the galvanometer is following a rectified current wave. The galvanometer follows the half-wave current pulsations to about 3000 pulsations per second, and for higher frequencies, drops off and becomes substantially constant between 6000 and 7000 cycles per second and does not drop to zero as in Fig. 2. This is due to the fact that the unidirectional current flowing through the galvanometer tends to deflect it to only one side of zero, and this forc is balanced by the restoring tension of the suspension so that th galvanometer willassume a balanced position corresponding to the mean value of the unidirectional current. It is to be understood that above 7000 cycles the galvanometer does not follow the current pulsations as it does below 4000 cycles but assumes a means deflection where the mean deflecting force and the restoring force are balanced. The transition between these galvanometer characteristics occurs between the two frequencies last mentioned. In producing the full curve of Fig. 4, the maximum value of the unidirectional current pulsations of variable frequency was held constant. If, now, the value of these current pulsations be increased or decreased, th deflection obtained at any frequency will rise and fall accordingly. Hence it is seen that below about 3000 cycles the oscillograph may be used in the usual Way for either I direct or alternating current measurements, and

the magnitude of its deflection will be independent of frequency variations and proportional to current magnitude. Also above 7000 cycles per second th galvanometer may be used to measure the mean value of direct current pulsations and its deflection will be practically independent of frequency variations. Below 3000 cycles the galvanometer may be used to study fundamental wave form, but above 7000 cycles the galvanometer will not follow the carrier wave, and the wave form of the carrier wave'does not influence the deflection except as it may vary the mean value of the current. It will, however, follow the lower frequency modulating wave.

So far as I am aware, the characteristics of the galvanometer type oscillograph when energized by unidirectional current pulsations having a frequency above the natural period of vibration of the galvanometer have never before been invesitgated or utilized for any purpose. Hence, I have discovered a new method of using this type of oscillograph which does not interfere with its uses as heretofore employed, and which may be exemplified by the curves of Fig. 5 representing, in curve l2, pulsating direct current of variable wave form and frequency of 7000 cycles per second or above, and curve [3 the corresponding mean value of the current. The oscillograph type galvanometer may be used to produce the curve I3 notwithstanding the fact that th frequency is above the natural period of oscillation of the galvanometer and still farther above the normal frequency measurement range of the galvanometer. Moreover, the discovery enables me to use the galvanometer type oscillograph for measurements of feeble currents with a reduced amount of amplifier equipment even though the measurement current be direct or alternating and within the ordinary frequency measurement range of such galvanometers as will. now be explained.

In Fig. 6, I have shown a transformer l4 supplied from a source of alternating current of pulsating direct current 15 having a frequency of the order of 10,000 cycles per second. A thermionic tube I6 of the tiny, gas-filled type has its cathode ll connected to the midpoint of the transformer secondary, its grid l8 connected to one end of the transformer secondary through a condenser l9, and its plate 20 connected to the other end of the transformer secondary through the oscillograph galvanometer 3. The grid and plate are also connected through a variable resistance 2l' which may be a wire resistance strain gauge of the type described in Patent No. 2,292,549.

It is evident that only direct current pulsations of a frequency of 10,000 cycles per second will flow in the plate-galvanometer circuit and hence are within the range 'of the galvanometer response which has been explained above, and becaus of the use of a tube capable of passing considerable current, ample current can be made to flow through the galvanometer to give good response. There are a variety of ways to cause the current flow in the plate-galvanometer circuit to be proportional to a desired measurement. In Fig. 6 this is done by varying the phase relation between the grid and anode voltages, thus utilizing a well-known characteristic of the tube. Each shaded area of the positive current pulses in Fig. 7 represents the current flow per cycle in the plate circuit of the tube when the grid voltage is nearly in phase with the anode voltage, and in Fig. 8 the shaded area represents such current flow when the grid voltage is appreciably out of phase with the anode voltage. Fig. 9 represents the average plate circuit current flow as a function of phase angle between grid and plate voltages where zero current flow corresponds to the maximum phase angle between grid and plate voltages at which current flow occurs. The average current can be found from the expression I lcrest g Referring again to Fig. 6, it is apparent that as resistance M is varied, the phase angle between plate and grid voltages will vary, and therefore the average plate current will change such that, by calibration, the average current through th galvanometer and its deflection may represent an amplified change in the resistance 2! corresponding to the change in strain or other factor which causes such resistance change. Instead of varying the resistance 21 I might vary the condenser E9 to obtain the desired results.

Fig. 10 is similar to Fig. 6 except that the condenser is replaced by an inductance again either the inductance 22 and here or the resistance 2| may be varied to control the amplified output current of the tube l6.

vector relations between the Fig. 11 represents the anode and grid voltages for Fig. 10. As the resistance 2| is changed,

its current 2 li is changed,

and the current 22f of the inductance is changed, varying the phase be-' tween the plate and grid voltages as indicated.

Fig. 12 represents another method in which a bridge circuit may be used to control angle between the plate gas-filled amplifier tube taken off of a bridge circuit through 9.

age is the phase and grid voltages of a 6. Here the grid volttransformer 23 and compared to the plate voltage by the connection 24. resistance arms, one, 2|,

may be a wire strain gauge.

is an inductance. The

The bridge contains three of which is variable and The other arm 25 bridge is connected across the high frequency source of supply for the tube such that variations in phase shift in the voltage will cause the desired across the bridge and accordingly. In Figs. may oscillate up to faithfully recorded by the vary 6, 10, and 12, the variable 2500 cycles per second and be the resistance 2| the output of the tube oscillograph.

The new method of using the oscillograph galvanometer as herein described may be used to measure the mean value of any alternating current voltage having a frequency above about 3500 cycles by energizing such frequency source through represented in Fig.

galvanometer from the a full-wave rectifier as 13. This may be useful to record rapid fluctuations in the mean value of the alternating current voltages to be observed here that with and currents. It is full-wave rectification the rectified pulsations will be at double the frequency of the alternating current source.

In Figs. 6, 10, and 12 a plifier tube was employed,

gas-filled type of amsince the plate current of such a tube is suflicient to operate the oscillograph type of galvanometer directly. In Fig. 14, I have represented a modification of my invention employing an amplifier uum type having a plate to the order of about tube 26 of the high vaccurrent output limited milliamperes. To obtain sufficient output current to operate the galvanometer, the plate current of the tube is stepped up by a current circuit is controlled in 6 as by varying the resistance 2|.

transformer 21. The

the same way as in Fig.

However, the

characteristics of this form of tube are as represented in Fig. 15.

r In the upper left portion of Fig. are found the transfer characteristics at various plate voltages E age Ep E9 860.,

In the lower left portion of Fig. 15 is shown at E: a sine wave of grid; and also the sine wave hold constant at 200 volts but voltage applied to the of applied plate voltat different phase relations with respect to the grid voltage, Ep

meaning a O-degree phase angle,

Ep a 30-degree phase angle, etc. The zero line of grid voltage wave E;

is positioned at any point of the grid valtage scale of the transfer characteristics by choosing the voltage determined by the plate current proper amount of bias the bias of battery 28in Fig. 14. At the upper right in Ip for different phase angles of Fig. 15 is shown grid and plate voltages for a plate voltage of 200 volts. It is seen that the plate current Ip is a maximum when the plate and grid voltages are in phase and decreases as this phase angle increases, the plate currents for 0, 30, 60, 90, and

degrees phase angle being represented in decreasing magnitude.

The output of tube 26, Fig. prise direct current pulsations of a frequency of, say, 8000 per second. of varying magnitude, the magnitude being determined by the phase angle between plate and grid voltages. While the output current is not sufficient to obtain good response of the galvanometer 3, it can he stepped up by the required current transformer at 21 so that it is suflicient. The current transformer secondary output is rectified and influences the oscillograph galvanometer as explained in connection with Fig. 5.

In accordance with the provisions of the patent statutes I have described the principle of operation of my invention, together with the apparatus which I now consider to represent the best embodiment thereof, but I desire to have it understood that the apparatus shown is only i1- lustrative and that the invention may be carried out by other means.

What I claim as new and desire to secure by Letters Patent of the United States is:

1. The method of measuring wave form phenomena with an oscillograph type galvanometer such that at least complete half cycles of the wave may be visualized, such wave form phenomena varying at a frequency within the range of 25 to 4000 cycles per second and which is well below the natural period of vibration of the galvanometer, which consists in energizing said galvanometer with direct current pulsations of a constant frequency which is at least 7000 cycles per second and well above the natural period of vibration of the galvanometer, and superimposing on such energizing current the wave to be measured by amplifying such pulsations solely in direct proportion to such wave.

2. The method of tracing the wave form of varying current phenomena having a frequency between 25 and 4000 cycles per second with an oscillograph type galvanometer so that at least complete half cycles of the wave may be visualized and where such varying current is too weak to produce a good response of the galvanometer, which consists in energizing the galvanometer with direct current pulsations of good galvanometer response magnitude but at a frequency of at least 7000 cycles per second and which is well above the natural period of vibration of the galvanometer, and directly modulating the value of such pulsating energizing current by amplification solely in response to the varying current phenomena to be traced.

KENNETH .R. GEISER.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,651,150 Ramsey Nov. 29, 1927 1,728,835 Petch Sept. 17. 1929 1,957,511 Weinberger May 8, 1934 1,987,539 Razek et al. Jan. 8, 1935 2,125,608 Gerlach Aug. 2, 1938 OTHER REFERENCES G. E. Ci. Research Laboratory publication No. 491, June 1930, pages 14-21.

14, will then com- I

Images