Патент USA US2137419код для вставки
Nov, 22, 1938. F.H.SHEPARD.JR v2,137,419 _ AMPLIFIER CIRCUIT VFiled. May 28, 1934` 2 Sheets-Sheet 1 FRANCIS H.5HEPARD JR. BY )fgQ¿Mm ÁTTORN Ev Nov. 22,1938. ‘ ' F. H. sHIEÀPARD. JR 'y 2,137,419 AMPLIFIER CIRCUIT 5» C» è ,im . « ¿j ' kfz 6/ , 0f’ INVENTOR FRANCIS H. SHEPARD JR.` ì BY ` >ÀTTORN w Patented Nov. 22, 1938 2,137,419 ‘UNITED STATES PATENT oFFicE 2,137,419 ADIPLIFIER CIRCUIT Francis H. Shepard, Jr., Rutherford, N. J., as# signor to Radio Corporation of America, a cor poration of Delaware Application May 28, 1934, Serial No. 727,968 10 Claims. (Cl. 179-471) The present invention relates generally to am Figure 7 shows an improved circuit arrange plifiers and more particularly to so called D. C. ment over the system shown in Figure 6 eliminat-, ampliñers, of the type adapted to amplify D. C. ing the use of additional voltage sources; and low frequency A. C. Figure 8 illustrates a system like that shown For 'many applications it is desirable to have in Figure '7 except that means are provided for an amplifier that will efñciently amplify small eliminating the eii‘ects of line voltage variations; D. C. voltages. In systems heretofore known, and, Figure 9 is a diagrammatic representation of where more than one stage of amplification is needed, D. C. ampliiication is not obtained easily an improved and simplified ampliiier incorporat 10 for the reasons that either there must be sepa ing the features of the present invention. rate batteries for each stage, or a set of bucking Attention is now directed to Figure 5 wherein batteries for each stage, or a set of divider re lines 5 and 6 are adapted to‘be connected to a sistors foreach stage, or again, a very high suitable commercial A. C. power supply network voltage source across which the various stages .by means of terminals I and 2. A space dis 15 m-ay be placed in series with each other, or a charge device T1 provided with anode, cathode modulator system where the D. C. is converted and grid electrodes acts as a first amplifier for the into A. C., amplified and then re-converted back D. C. input. 'I'he input circuit of tube T1 com into D. C. All of the methods mentioned above prises a connection between the grid electrode are more or less complicated and furthermore are and the cathode and includes terminals 3 and quite critical to adjust.` It is also desirable to have 4, across which the D. C. to be amplified may 20 an amplifier oper-ating directly on raw A. C. for use with a photo tube and other industrial uses be applied in any desired manner. as for instance temperature control systems and burglar alarm systems. The present invention may be broadly stated so as to maintain the grid of the tube T1 at a predetermined operating potential with re spect to the cathode thereof. The cathode of 25 to comprise a method and means for cascading grid controlled recti?iers to produce a D. C. am plifier system which does not necessitate cas tube T1 is also connected to the line 6, while the anode of the tube is connected to the line 5 through a resistor R1 which is shunted by a condenser C1. The anode of tube T1 is also connected to the line 6 through a resistor l'tz` and a series condenser C2. A second amplifica tion stage may be provided and comprises an electronic tube T2 having an anode, a cathode and a control grid. 'I'he grid of the tube T2 is connected directly to a point intermediate 35 resistance R2 and condenser C2 and there is thus provided a. direct connection between the anode of tube T1 and the grid of tube T2 through resistor R2. The anode of tube T2 is connected to the line 5 through a suitable resistor Rs which is shunted by a condenser Ca while the cathode of the tube is connected directly to line 6. The anode of tube T2 is also connected to line 6 by means of a connection including resistor R4 cading the necessary tube energizing voltages. The invention also contemplates a method and means for correcting for variations in the sup ply voltage. ' Bro-adly speaking, in the present system the negative voltage developed on the plate of a grid controlled rectiñer tube connected across an A. C. line is used to bias the grid of a following tube also‘connected across the line. The energy to be amplified is impressed -across the input of the iìrst tubeA and the amplified energy is avail able across the output of the last tube. The various broad aspects of the invention are illustrated by the circuit diagrams in the ac companying drawings wherein, Figures 1-4 represent curve sheets showing various tube operating characteristics of use in describing the present invention; Figure 5 illustrates diagrammatically a circuit diagram incorporating, in a simple manner, the broad features of the present invention; Figure 6 is a diagrammatic representation of 50 a circuit arr-angement differing. from that shown in Figure 5 only in that means are provided for increasing the gain of the amplifier by causing the tubes to be operated at or near the steep 5.5. part of their grid voltage plate current curves; A bias bat tery B1 may be provided in the input circuit and a series condenser C4. 45 A third stage may be similarly connected. This stage may also include an electronic tube T3 provided with anode, cathode and grid elec trodes. The cathode of T3 is shown connected to the line 6. The anode is connected to the 50 line 5 through a resisto-r R5 shunted by a con denser C5. The control grid is connected to a point intermediate resistor R4 and condenser C4. It is obvious that any number of succeeding stages may be added, it being understood that 55 2 2,137,419 causes a'pulsating D. C. current to flow through only does away with the necessity for a battery such as B2, (Figure 6) but also acts to partly counteract the above mentioned instability due to line voltage variations. For instance, in the system shown in Figure 7, if the line voltage falls, the drop across the plate resistor of a tube will tend to become less with the result that the grid biases of the various tubes Will tend to de crease. However, the superimposed D. C. on the line will also decrease and this will tend to make 10 the plate potentials or grid biases of the various tubes more negative. the tube so that the D. C. potential of its plate or anode becomes negative with respect to the Thus, the change in rectified superimposed D. C. due to the line voltage variations can be three stages have been shown for purposes of illustration only. In Figure 5 a utilizing circuit or load shown generally as a voltmeter V is connected across the resistor R5. From the above it can be seen that the cir~ cuit shown in Figure 5 cascades three tubes which have A. C. on their plates, for the purpose of amplifying D. C. and/or low frequency A. C. voltages applied to the input terminals 3 and 4. In operation, the rectifying action of each tube 15 line. The grid bias on a tube will determine its effective plate conductance while conducting, that is, during the half cycle when the plate is positive, and hence, will control the amount of rectified current passed through its respective plate resistor. The IR drop variations across the plate resistor are in excess `of the bias vari ations, hence in accordance Ywith Well known principles, the tube Will act as an amplifier. As the D. C. potential across both sides of an A. C. line is zero, the voltage drop across the plate re sistor of any tube» may be used after passing it through a suitable ñlter so as to obtain an average value, as bias for the next tube. In this way any number of stages may be used to ob made to oppose the effects of the A. C. line volt age variations, if it is so desired. In an actual circuit built as shown in Figure 7 it was found that the instability for a given sensitivity can be reduced to approximately 1/6 of that obtained in circuits such as those described in Figures 5 and 6. As pointed out above in connection with the description of the circuit arrangement shown in Figure '7 the instability in that system due to line . voltage variations is only partially counteracted. It is conceivable that in some systems it would be highly desirable to have a circuit arrangement for the purposes described which within certain in a circuit similar to that shown in Figure 5, a voltage gain of about Il per stage can be readily limitations is absolutely independent of line voltage variations. Such a circuit arrangement 30 has been shown in Figure 8 wherein a potentiom eter resistance R6 is connected across the lines 5 and 6 and a connection from the anode of tube realized. T1 to a point on said resistor Re including a In a circuit such 'as shown in Figure 5 the gain of any tube is limited by the factthat the D. C. variable tap arrangement 'l and condenser C1 is provided. Condenser C1 in Figure 8 instead of directly shunting the resistor R1 as in Figure 7 is connected to the supply lines through the po tentiometer resistor Re. In other respects the circuit arrangement of Figure 8 is similar to that shown in Figure 7. By means of the potentiom eter 1, Re the A. C. voltage on the plate of tube T1 can be varied independently as regards 30 tain the required'voltage amplification. In prac tice, it has been found that using type 53 tubes voltage drop in its respective plate resistor is limited to the grid bias voltage required for the next tube. This makes it necessary'to use a very low value of plate resistor or to operate the tubes with a very high negative bias in order to make the grid bias equal the plate load voltage drop. In either of the above cases a great sacrifice in eilìciency is made. The circuit arrangement shown in Figure 6 is concerned with a system for overcoming the above mentioned loss in efliciency. It is ap parent at once from a study of the circuit shown in Figure 6, that the only difference in the sys , tems shown in Figures 5 and 6 is that in the lat ter a' battery B2 is placed in series with the A. C. line so that the drop in the plate resistor might be increased to a value equal to the battery voltage plus the grid bias voltage. This arrangement makes it possible to operate the tubes more ef iiciently and, in fact, it was found in an actual set up using type 53 tubes that a voltage gain of between 25 and 35 per stage may be easily at tained. l the D. C. voltage: In this way the A. C. and D. C. effects described above in connection with Figure 45 7 can be made to perfectly compensate for each other and make the amplification and level of output current independent of line voltage va-riations Within said limits. The theory under lying the above can best be illustrated by the q following: A triode may be operated with A. C. on its plate to amplify changes in the D. C. bias applied to its grid. ’The tube will act as a rectifier and will build up a negative voltage on its plate. The value of this voltage can be controlled by the ' grid potential of the tube. When pure D. C. is applied to the plate of a vacuum tube through a load resistor the assumed D. C. potential of the It is quite obvious from a consideration of the system disclosed in Figure 6 that the use of a separate source such as battery B2 is inconveni ent and should be avoided.v For this purpose, the system shown in Figure 7 will now be con sidered. In Figure '7 the battery B2 of Figure 6 65 has been displaced by the diode rectifier D1 con nected across the line 5, E and a condenser Ce, interposed between terminal l and the connec tion between the cathode of D1 and line 5. In plate is always positive and varies with the supply 7o; connection with the systems shown in Figures 5 bias variations. If the AC/DC ratio is too high 70 over-compensation will result and if it is too low there is obtained under-compensation. Be cause of the fact that a single stage of the ampli lier may be considerably over-compensated all 60 and 6 it was also found that when amplifying very small voltages through a number of stages, linevoltage variations which are also amplified make the ampliñers unstable. The use of a recti ñer or diode ‘D1 such as shown in Figure '7 not voltage. By placing both A. C. and D. C. on the plate of the tube simultaneously and keeping the ratio of A. C. to D. C. constant as is the case where the A. C. is rectified to supply the D. C. the two effects can be made to oppose each other 65 for the effects of line voltage variations. If the proper AC/DC ratio is chosen there can be ob tained practically perfect compensation over cer tain ranges of line voltage variations and grid the compensation for several stages in cascade 75 2,137,419 may be obtained in one of the stages only, hence, as shown in Figure 8 the compensating effect is ob tained only in the ñrst stage and may be of such a‘value as to compensate for all the succeeding stages of the amplifier. `While no mention has been made of the ef fects of heater current variations which cause effective changes in the temperature of and hence 'in the contact potential- between cathode and grid and changes in the velocity of emission of electrons from the cathode, which changes are equivalent to `actual changes in grid bias, it is. apparent that the AC/DC ratio may be adjusted so that this heater effect can also be compensated for. However, the compensation in the case previously discussed will take place immediately with line voltage variations while the effects due to cathode temperature will lag behind the volt age changes. If the time constant of the -re 20 sponse of the amplifier is greater than the heat ing time of the cathode, compensation can be ac complished `without any instantaneous de compensation. For a clearer understanding of the operation 25 of the system herein disclosed attenti-Onis now directed to Figures 1 through 4. connected to an A. C. power supply system through terminals I and 2. For purposes of il lustration only three tubes T5, T5 and T7 have 25 been shown comprising the amplifier, however, it is to be distinctly understood that any desired cation curves for a triode with various values of A. C. on the plate and with Zero bias on the grid. number of tubes may be’utilized and connected grid, the plate voltage current curve of a triode is shifted parallel to itself along the voltage axis, see Fig. 3. This causes the whole family of rec 35 tification curves to shift along the voltage axis. Figure 2 shows how the rectification curves have shifted from the position shown in Figure 1 when thegrid biasis` changed from zero to a negative value. A load line A--B placed in Fig as taught by the present disclosure. The ñrst tube T5 includes anode, cathode and grid elec 30 trodes. One side of the cathode is connected to line Ei while the anode is connected to conductor 5 through condenser C8. The anode is also con nected to conductor- t through two paths one of which includes the resistor Rin and the other a 35 resistor R11 and a condenser Cè in series. The D. C. input is applied across terminals 3‘ and 4, terminal 3 being connected to the grid of T5 while ures 1 and 2 will intersect the constant A. C. lines at different points on the two sets of curves. terminal 4 is connected to a point on the resistor R10 by means of a variable tap I2. A condenser C7 is connected between the connection from For instance, the load line in Figure 2 intersects terminal ll to the resistor R111 and conductor 6. . the siX volt A. C. line at a plate potential of Zero The second tube is also shown as including anode, cathode and grid electrodes, the anode being connected to the conductor 5 through a volts.`- . With a less negative bias, as shown in Figure 1, the plate voltage will be some value less than Zero. In this way the grid bias can be used to control the potential on the plate and since this plate potential can be normally around zero it can be coupled (through a simple filter to re 50 move the A. C.) directly to the grid of a follow ing amplifier stage (see, for instance, Figure 8 where the anode of tube T1 is coupled directly to the grid of tube T2 through the resistor R2) the condenser C2 acting as a ñlter element. Referring again to Figures 1 and 2, if the D. C. supply potential is increased the plate po~ tential will increase. If the A. C. is increased the plate potential will decrease. If both the A.` C. and D. C. supplies are increased together 60 inthe proper ratio there will be no change in the D. Cjplate potential. In the circuit arrange ment shown in Figure 8 a rectifier is used to pro videthe D. C. supply, while the A. C. is supplied from the potentiometer arrangement 1, R5 con 65 nected across the line through a low impedance condenser C1 to the plate of the tube. The proper ratio between the A. C. and D. C. for perfect compensation may be obtained by adjusting the value of the A. C. by means of variable tap 'I. As stated previously, it is possible to over-com pensate or under-compensate the circuit so that the plate potential may vary in either direction with varying line voltage. Because of this fact it is possible to compensate in one stage the 75 effects of line voltage variations in the other 70 ode temperature will have an effect, due to the velocity of emission of the electrons and contact potential on the cathode, equivalent to a change in grid bias. 10 Figure 4 shows a characteristic curve for bias which is equivalent to cathode exciter voltage changes. Since there is a definite relation be tween ñlament voltage and equivalent grid bias it is possible to compensate for filament voltage variations within limits in the plate circuit by the method outlined above. Attention is now directed to Figure 9 which shows in diagrammatic form a simpliñed A. C. operated D. C. amplifier circuit that is self biased and compensated for line voltage variations. As in Figures 5 through 8 conductors 5 and 6 are In Figure 1 there is shown a family of rectifi 30 This‘set of curves is similar to the curves usually shown for a diode detector. By means of the 55 3 stages and it is also possible to compensate for the eiîects of heater voltage variations, as will` be more fully covered below. As long as the plate current is small compared to the emission current of the cathode, that is, under unsaturated cathode current conditions, changes in the cath condenser C10 while one side of the cathode is connected directly to the conductor 6. The sig nal grid of the tube T5 is connected to a point on the resistor R11 by means of a variable tap I3. As in the case of tube T5 the anode of tube T5 is connected to the conductor E through two paths. One of these paths includes a resistor R12 while the other comprises resistor R13 and a condenser C11 in series. The third tube of the system T7 i-s also provided with anode, cathode and grid electrodes, the cathode being connected directly to the conductor E while the grid elec trode is connected to a point intermediate re sistor I3 and condenser C11. The anode of tube T1 is connected to an output terminal l I, ter 60 minal I0' shown at one end of the conductor 5 being the other output terminal of the system. It is to be understood that any desired utilizing system may be connected across the output ter Ininals` I0 and II. 65 For a proper understanding of the operation of the system shown in Figure 9, attention is di rected again to the curve sheets shown in Fig» ures 1 and 2. It will be noted that the load line OC starts from Zero as there is no D. C. supply 70 in the circuit shown in Figure 9. 'I'he plate volt age will always be negative as shown by the in tersections with the triode rectification curves. The actual value of D. C. plate voltage with a constant A. C. voltage on the plate will vary with 4 2,137,419 tential due to the rectifying action of the tube. This potential, for best amplification, will be too to that given above in connection with the de scription of curve sheet shown in Figure 4. In the system shown in Figure 9 the slider I3 should be adjusted along the resistor R11 so that the grid of the tube Ts swings into the op erating range of the tube but never swings posi tive with'respect to the cathode. high a Value to use directly for self-bias. How ever, since ’the load resistor is returned to ground or zero potential, a slider may be placed on this of none of the resistances and condensers used are critical. In one of a number of practical the grid bias. This is shown by observing the intersections of the constant A. C. lines with the D. C. load line in Figures 1 and 2. In normal operation, considering the tube T5, the anode will assume a highly negative average D. C. po resistor to obtain any average D. C. Voltage from zero to the average D. C. potential of the plate. It is possible in some particular instance that the D. .C. potential may be in the order of -80 volts. If -4 volts isV the normal bias for the tube in this circuit, the slider is adjusted along the load resistor R10 (see Figure 9) until a potential of -4 volts is reached. Condenser C1 is üsed as a filter to by-pass the A. C. on the top portion 'lil of the divider resistor R10. This will keep the A. C. oiî the grid of the tube T5. If the normal a 2000 ohms resistor was used for the A. C. line shunt while the plate resistances were each one megohm, the grid resistors two megohms and the condensers 0.1 mmf. While the invention has been described by cer tain particular embodiments, it is to be under stood that the principles underlying the inven tion may be carried out in many other forms f without departing from the spirit of the inven tion, and no limitations upon the invention are -80 volts, the slider will be set approximately V20 of the way up the load resistor R10 and the degeneration due to the self-bias will be one part in 20, times the actual realized voltage gain. It should be understood, of course, that the values given above are approximate values that can be intended other than those imposed by the scope of the appended claims. I claim: 1. In an alternating current operated direct current amplifier, an electronic tube including The potentiometer R11 as far as D. C. is con cerned is merely a resistor of negligible Value in series with the grid of tube Ts. Condensers C10 and C11 have no eiïect on the D. C. voltage, however, potentiometer R11 as far as A. C. is 36 concerned will act as a voltage divider resistor across the line. Condensers Ca and Cs have neg ligible impedance to A. C. The instantaneous potential on the grid of tube Te is the sum of the D. C. voltage on the plate of tube T5 and the 40 instantaneous value of A. C. at the point to which the slider I3 is set on the resistor R11. During the part of the cycle that the tube Te is conduct ing. the instantaneous A. C. voltage applied to the grid of tube Te is in such a direction as to oppose the normal D. C.,bias supplied by the plate 45 of the vacuum tube T5. Thus, the grid during the conducting part of the cycle will be less nega tive than the normal bias and will be of such a value that the grid of the tube Ts will be at a 50 embodiments of the invention, and in particular one that followed the diagram shown in Fig. 8, bias is ' -4 volts, when the plate potential is expectedV in normal operation of the system. 30 In practice it has been found that the values suitable operating potential. » Suppose that the A. C. line voltage, that is, the A. C. voltage on the anode is increased it can be seen from the rectification curves in Fig ures 1 and 2 that the negative plate potential This will cause the normal grid bias of tube Ts to become more negative. How ever, increasing the A. C. line Voltage would also cause an increase in the superimposed A. C. on the grid on the tube T6 and this will tend 60 to make the grid more positive during the op erating part of the cycle. Since the two above mentioned effects can be made to oppose each other the eñîects of line voltage variations may 55 will increase.> be nulliñed. Compensation adjustment is made an anode, a cathode and a grid electrode, a source of alternating current energy including a pair of line conductors, a connection including a re 30 sistor element between one of said conductors and the anode of said tube, a connection between the cathode and the other line conductor, an in put circuit for said tube including said grid elec- ~ trode and the cathode, a capacitive impedance shunted across said resistor element and a filter circuit comprising an impedance device and a condenser connected between the anode of said tube and the cathode. 2. In a direct current amplifier, a pair of con 40 ductors adapted to be connected to a source of alternating current, a controllable rectifier corn prising an electronic tube having anode, cathode and control electrodes, a connection including a high plate resistor between the anode and one 45 of said conductors, a connection between the cathode and the other conductor, an input cir cuit for said tube including the cathode and control electrode thereof and means adapted to connect the input circuit to a source of energy 50 to be amplified, a second controllable rectifier comprising an electronic tube having anode, cath ode and grid electrodes, a connection between the anode of the last named tube and one of said conductors including a plate resistor, a con nection between the cathode of the last named tube and the other conductor, a direct connec tion between the anode of the iirst tube and the control electrode of the second tube including an impedance device, whereby a controlling volt 60 age is impressed upon the control electrode of the second device as determined by the drop across the high plate resistor associated with the first tube due to the flow of current therethrough, 65 by varying the A. C. voltage on the grid of tube said controlling voltage being normally too nega Ta by adjusting the slider I3 on the potentiom eter R11. As previously stated, it is possible to obtain overcompensation, hence, it is possible to perform all the necessary compensation for sev 70 eral stages in the one stage. It is also possible as previously pointed out to accomplish compen sation for the effects of filament voltage varia tions. The explanation for filament compensa tion in the circuit shown in Figure 9 is similar tive as regards the normal operating character istics of the second tube to produce proper op eration thereof and mea'ns for impressing a po tential upon the anode of the ñrst tube with re spect to the cathode thereof of such direction 70 and amplitude to partially compensate for the high drop across the plate resistor of the first tube whereby there is impressed upon the control electrode of the second tube a control voltage 75 2,137,419 commensurate with the proper operation of said tube. - `1 , ‘ ` ~ 3. In a direct current amplifier, a pair of con ductors adapted‘to be conneotedto a sourceof alternating current, a controllable rectifier com prising an electronic tube having anode, cathode and control electrodes, a connection including a between the alternating current and direct cur rent voltage applied to the plate of the tube is maintained constant by deriving the direct cur rent voltagefrom the alternating current voltage whereby variations in voltage of one causes `like , voltage variations of the other. . „ 8. In `anv alternatingcurrent operated direct high plate resistorbetween the anode and one of currentv` amplifier, . a source» of l alternating cur said conductors, a connection between the cath rent, a plurality of cascaded electronic rectifiers connected across the source in parallel, each of said rectifiers including a control element ar ranged so as to control the rectifying action of the rectifier, a source of energy to be amplified, lO ode and the other conductor, an input circuit for said tube including the cathode and control elec trode‘thereof and means adapted to connect the input circuit to a source of energy to be ampliñed, a second controllable rectifier comprising an elec tronic tube having anode, cathode and grid elec trodes, a connection between the anode of the last named tube and one of said conductors in cluding a plate resistor, a connection between the cathode of the last named tube and the other con ductor, a direct connection between the anode of the first tube and the control electrode of the second tube whereby a controlling voltage is impressed upon the control electrode of the second device determined by the drop across the high 25 plate resistor associated with the first tube due to the flow of rectiñed current therethrough, said controlling voltage being normally too negative as regards the normal operating characteristics of the second tube to produce proper operation 30 thereof and means for impressing a potential across said two conductors of such direction and amplitude as to partially compensate for the high drop across the plate resistor of the first tube whereby there is impressed upon the control elec 35 trode of the second tube a control voltage com means for connecting said last named source to the input of the first of said cascaded reotifi-ers 15 to control its action in accordance with energy to be amplified, means for controlling the recti iying action of each of the other rectifiers in accordance with the rectiñcation of energy by the respective preceding rectifier, means for impress 20 ing a unidirectional potential across the space path of at least the first rectifier, the direction and amplitud-e of said potential acting to par tially compensate for the potential applied to the next successive tube. 25 9. In a direct current amplifier, a source of alternating current including a pair of terminals, a first conductor connected to one of said ter minals and a second conductor connected to the other of said terminals, a diode rectifier element 30 including an anode and a cathode, the cathode the positive potential side of the rectifier is con of said rectifier element being connected to said first conductor and the anode thereof being con nected to the other of said conductors, an elec' tronic tube provided with an anode, a cathode 35 and a control electrode, means including a plate resistor for connecting the anode to the ñrst named conductor, means for connecting the cath ode to the other of said conductors, an input circuit for said electronic tube including the 40 cathode and control electrode thereof, a connec tion between the anode of said electronic tube and the cathode thereof including a resistor and nected to that conductor which is connected to a capacity connected in the order named, a ca mensurate with the proper operation of said second tube. 4. The system described in the next preceding claim further characterized by that the means 40 for impressing the potential across the two con ductors comprises a rectifier device connected across the two conductors and arranged so that pacity element shunted across said plate resistor, 45 45 the anode of the first tube. 5. The system described in claim 3 further characterized by that the means for impressing the potential across the two conductors comprises a rectifier device connected across the two con 50 ductors and arranged so that the positive po tential side of the rectifier is connected to that conductor which is connected to the anode of the ñrst tube and wherein the rectiñer device is adapted to be connected to a source of alter 55 nating current through a condenser said system being further characterized by that the compen sating voltage also acts to partly counteract in» stability of the amplifying system due to line voltage variations. 6. A method of amplifying a direct current voltage applied to a control grid of an electronic tube which comprises impressing both an alter nating current voltage and a direct current Volt age upon the anode of the tube so that the tube 65 will act as a rectifier and thereby build up a 60 negative voltage on its plate, controlling the value of this negative voltage by the grid po tential applied to the control grid and main taining the ratio of the alternating current volt 70 age to the direct current voltage applied to the plate constant and of such a value that effects of alternating current line voltage variations are i compensated. Y 7. The method described in the next preceding 75 claim further characterized by that the ratio a second electronic tube including an anode cath ode and control electrode, a connection between the control electrode and the cathode thereof in cluding said first named capacity element and means includinga plate resistor for connecting 50 the anode of said second named electronic tube to the ñrst named conductor, a utilizing circuit and means for connecting the utilizing circuit across said last-named plate resistor. 10. In an amplifier circuit including at least 55 two cascaded electron discharge devices each of said devices including anode, cathode and grid electrodes, a source of alternating current pro vided with a pair of terminals, a line conductor connected to one of said terminals and a line 60 conductor connected to the other of said ter minals, a resistor device connected between saidV two conductors, a diode rectifier having its cath ode connected to the first of said conductors and its anode connected to the second, a connection 65 between the anode of the first of said two electron discharge devices and the first of said conductors including a load resistance, a connection between the last named anode and a point of said first named resistor including a condenser, a connec 70 tion between the other line conductor and the cathode of said first electron discharge device, an input circuit for said first electron discharge device including the grid and cathode thereof, a connection between the anode ofthe second of 75 6 2,137,419 said electron discharge devices. and the iirst line conductor including a. load resistor shunted by a condenser, a connection between the anode of said second electron discharge device and the cathode thereof' including an impedance element,V means for connecting the cathode of the second electron discharge device and the cathode of the ñrst electron dischargei device, a connection be tween the anode of the first electron discharge device and the cathode thereof including a re sistor and a condenser in series and a connection between the grid electrode of thev second named electron discharge device and a point of said last named circuit between the resistor and con denser. ' FRANCIS H. SHEPARD, JR.