DE861866C - Device for amplifying modulated high-frequency vibrations, which contains at least one amplifier tube - Google Patents

Description

The invention relates to a device for reinforcement modulated high-frequency oscillations, which contains at least one amplifier tube and stands for the output stage of a transmitter is particularly suitable.

It is known to amplify modulated high-frequency oscillations to use an intensifier tube in which the grid bias is so great that the tube during half of each period the alternating voltage to be amplified allows current to pass through. Such an amplifier is known under the Designation B amplifier and has at full load

a useful effect of - if it is assumed,

that the amplitude of the occurring in the output circuit AC voltage corresponds to the anode DC voltage. In practice, however, one Maximum amplitude of the anode alternating voltage permitted, which is equal to 0.8 to o, c; times the anode direct voltage is, whereby the maximum useful effect of a B-amplifier in practice is not higher than 0.8

up to 0.9 times - = 67%. When using a 4

Such an amplifier for the purpose of amplifying modulated high frequency oscillations occurs Efficiency of 67% only in a positive modulation peak with 100% modulation of the carrier wave on, since only in this case the amplitude of the anode alternating voltage is equal to 0.8 to o, o times the DC anode voltage is. In the absence of modulation, the amplitude is the anode alternating voltage

only half of the value mentioned, so that the usefulness is no more than 33%. For radio stations, the mean modulation depth over a whole day is very small because the instantaneous value of the modulation depth rarely exceeds 30 ° / ₀ , ie the useful effect of a B amplifier is on average not significantly higher than 33%.

To achieve an amplifier with a higher efficiency, devices have been proposed in which a load impedance is supplied with energy from at least two amplifier channels and in which the output circuit of one of these channels is via a reciprocal network, i.e. a network whose input impedance is inversely proportional to the terminating impedance of the output circuit , is connected to the load impedance. A transmission line can be used as a reciprocal network, the length of which is a quarter of the wavelength of the vibrations to be amplified. Using these amplifiers, a useful effect of around 65 ° / _{0 can be} achieved even with a low modulation depth. The aim of the invention is to create a device for amplifying modulated high-frequency oscillations with a high useful effect, in which only a single amplifier channel is required.

According to the invention, between the anode circuit of a high-frequency amplifier tube and a load impedance a composite network switched on, its elements of reciprocal networks be formed at each connection point between two or more of these networks except at the point of the anode circuit mentioned and the load impedance, at least one Discharge tube operating as a switch is arranged, the impedance of which is between anode and cathode bridges the input or output circuit of the reciprocal networks in the connection point, wherein the discharge tubes are controlled so that with increasing amplitude of the amplified Vibrations one after the other and / or alternately at least a part of these tubes is blocked, whereby a stepped decrease in the input impedance of the composite network is achieved will.

According to an embodiment of the invention, the composite network can receive in the manner. be that each point of a group of points is connected via reciprocal networks to each point of a second group of points, where with two of the mentioned points the anode circuit of the high frequency amplifier tube or the load impedance, are connected.

In the case of a conventional B amplifier, the useful effect is proportional to the amplitude of the to be amplified. strengthening vibrations. At the highest value, the efficiency is around 67 ° / ,,, at half the value

33 Oj _{0 etc} _ Di _it is due to the fact that the

Änodenspannung the intensifier tube at the gxößten amplitude which is fully driven to be amplified oscillation at half value but only 5o _^0/0, etc. Among full scale of the anode voltage of the state is to be understood, in which the amplitude of the anode AC voltage is as large as possible, that is about 0.9 the anode DC voltage is. In conventional B amplifiers, the anode load impedance R _a is chosen to be constant and large enough that the output energy is the output energy when the anode voltage is fully driven, ie at the greatest value E _{ama! T of} the anode alternating voltage E _a

corresponds to the maximum energy that the tube must be able to deliver.

In the device according to the invention, the impedance R _{a is} not constant, but changes in steps with changing amplitudes of the vibrations to be amplified. With small amplitudes, R _{a is} larger, e.g. B. a few times larger than would be the case with a corresponding B amplifier. As a result, even with a small amplitude, the anode voltage E _{a is} fully controlled and the efficiency is high. In order to enable the output energy to increase as the amplitude of the oscillation to be amplified, the impedance R _{a is} abruptly reduced to a lower value, the modulation of the anode voltage and accordingly the useful effect dropping, but the output energy does not change. If the amplitude of the oscillation to be amplified increases further, the output energy can increase again until the anode voltage E _{a is} reached again at full level, whereupon the impedance R _{a is} again abruptly reduced, etc. This process can be repeated any number of times, until the last jump brings the impedance R _a to the same value that this impedance would have in the corresponding B amplifier.

It can be seen that the mean benefit at Amplitudes of the oscillation to be amplified between the value zero and the maximum value at the described method is considerably higher than with an ordinary B amplifier.

The invention is closer to the drawing explained.

The illustrated in Fig. 1 embodiment of the device according to the invention contains a power amplifier tube U of an amplifier for modulated high-frequency oscillations and a load impedance R, the z. B. can be an antenna to which the tube U supplies energy. Between the anode circuit of the tube U and the load impedance R a composite network is turned on, the elements of reciprocal networks R _0, R _1, R _3, R _1, .RG ', R _3', R ₁ ', U _2', R _z " . With the exception of J? _o , three of these networks are connected in series, while the resulting chains of three networks each are connected in parallel to the network R _0. In the connection points between the networks R ₁ and R ₅ ,, R ₂ and R ₃ , R ₁ and R ₂ ' , etc. are arranged as switch-working discharge tubes A, B and A' , etc., the anode circuits of which

are connected to the networks that the impedances between the anode and cathode of the tubes A, B, A ' etc. individually bridge the input and output circuits of the networks R ₁ , R ₂ , R ₁ etc. in the connection points, ie the networks are terminated in the mentioned connection points by the anode-cathode impedances of the tubes, B, A ' etc. The input impedance of the composite network forms the load impedance R _{a of} the tube U, the resistor R forming the terminating impedance of the network. For the sake of clarity, the connecting lines between the tubes and the networks are shown as simple lines, while the sources of the supply and control voltages are omitted.

The reference symbols R ₀ , R ₁ , R ₂ , R ₁ etc. denote reciprocal networks, ie networks which have the property that the input impedance is inversely proportional to the terminating impedance of the output circuit, while the output impedance is also inversely proportional to the terminating impedance of the input circuit is. The product of the input or output impedance and the terminating impedance is constant, and it is assumed below that for the reciprocal network R ₀ or R ₁ , R ₂ etc. the product of the input impedance and the terminating impedance is equal to R ₀ ² or R _{1 is} ² , R _{2 is} ² , and so on. The value R ₀ , R ₁ , etc. will hereinafter be referred to as a characteristic impedance. Furthermore, for reciprocal networks, e.g. B. for the network R ₀ ,

V = jR _o i,

where V denotes the voltage across the input terminals of the network R ₀ and i_ denotes the current flowing through the terminating impedance of the output circuit. This current is shifted by 90 ° with respect to the voltage V , is independent of the value of the terminating impedance and thus depends only on the input voltage V for a certain characteristic impedance. The same applies to the other networks R ₁ , R ₂ , R ₁ , etc.

The sudden reduction in the anode load impedance R _a is achieved in the device according to Fig. Ι as follows: It is assumed that the tubes, B, A ' , etc. are controlled in such a way that they are conductive and have a low impedance between anode and cathode and consequently practically short-circuiting the reciprocal networks at the junctures of these networks where the tubes in question are located. The terminating impedances of the networks R ₁ , R ₂ etc. formed by the tubes ./4, B, A ' etc. are thus very low, and the input impedances of the reciprocal networks at the point of the tube U and the resistor R are thus very high , ie the networks R ₁ , R ₂ , R ₃ , R ₁ etc. do not let any current through. The following applies to the input impedance Z _{1 of} the network R ₀

No. ₁

so that under these conditions the load R _{a of} the tube U is equal to Zi .

If the tubes A and B are blocked in a manner to be described in more detail, as a result of which the circuit containing the networks R ₁ , R ₂ and R ₃ begins to let current through, the resulting characteristic impedance R _{res of} the series connection of the networks R ₁ , R ₂ and R ₅ from the formula

K ₁ K ₃

Rres = -

R,

to be determined. R _res is therefore negative if one of the quantities R ₁ , R ₂ and R ₃ or all three are positive. The parallel connection of the network R ₀ to the circuit R ₁ , R ₂ and R ₃ is equivalent to a network with a characteristic impedance R _t , in which

R _n

Rr,

that is to say that the wave resistances R ₀ and R _{res are,} as it were, parallel. The characteristic impedance R _t resulting from this parallel connection is smaller than R ₀ when R _{Tes is} positive. Under these conditions, the anode load impedance R _a corresponds to the input impedance Zi of the networks R ₀ , R ₁ , R ₂ , R ₃

existing structure, where Z / = - ^ - smaller than

J? ²
Z \ - - ^ -. As a result of the blocking of tubes A and B , the value of R _{a is} reduced.

A further step-by-step reduction in the anode load impedance R _{a of} the tube U can be achieved in that the tubes 'and B' and then also the tubes A " and B" are blocked. In Fig. Ι three circles of series-connected networks are shown; of course, this number can be arbitrarily large.

In Fig. 2, the curve shown corresponds to the course of the anode alternating voltage E _{a of} the tube U in Fig. Ι as a function of the amplitude P of the vibrations to be amplified by the tube U. Since the gain is almost linear, P also denotes the amplitude of the current / flowing through the output impedance R. P ₄ indicates the maximum value of the vibrations to be amplified and the output current /, while P ₁ , P ₂ and P ₃ denote the levels at which the anode load impedance R _{a of} the tube U is abruptly lowered. The dimensions of the reciprocal networks are assumed to be such that the resulting wave resistance R _res of each of the circles R ₁ , R ₂ , R ₃ , R ₁ ', R ₂ ', R ₃ ' and R ₁ ", R ₂ ", R ₃ "is equal to R _0. The maximum current / flowing through the load resistor R , ie the current when the anode voltage of the tube U is at full output, is thus in each case when one of the circuits R ₁ , R ₂ , R ₃ or" I ₁ ', R ₂ , R ₃ and R ₁ ", R ₂ " and R ₃ " by the same amount

ι =

R _n

enlarged. E _amax denotes the amplitude of the anode alternating voltage E _a at full output. the

Distances OP ₁ , P ₁ P ₂ , P ₂ P ₃ and P _s P _i on the P-axis all represent this amount

and are therefore shown in the same way.

The dashed curve in Fig. 2 corresponds to this

Course of the anode impedance R _{a of} the tube U. As already explained, this impedance is im

ίο amplitude range between O and P ₁ , with only

J?

the network R ₀ is turned on - ~ .. In the area

between the limits P ₁ and P ₂ or P ₂ and P ₃ and P ₃ and P ₄ are two 'or three and four circles with characteristic impedance R ₀ connected in parallel, and the

7? 7? 7?

• Anode impedance is - ■ - or - ~ - and.

FIG. 2 shows how the anode alternating voltage E _a drops with every jump from full modulation to a smaller value.

_{To avoid a discontinuity,} the output energy I _a , R _a = I ² R (I _a denotes the alternating current component of the anode current of the tube U) must be almost the same immediately before and every sudden change in the impedance R a. Since R _a falls sharply, this means that I _a must increase sharply. If the tube U is a three-pole tube, an increase in I _{a will} readily take place; however, it is possible that this increase is actually not necessarily of the correct size. Any change in shape that may occur can, however, be eliminated by negative feedback. If the blocking of the tubes .4, B, A ' etc. is more or less gradual, the change in shape to be eliminated by negative feedback is much smaller than would be the case with extremely abrupt changes.

It is not absolutely necessary that the resulting wave resistance of the circles R ₁ , R ₂ , R ₃ or R ₁ ', R ₂ ', R ₃ ' or R ₁ ", R ₂ ", R ₃ "be Ji ₀ , ie the amplitude _{levels P 1} to P ₄ do not need to be selected at the same distance from one another. Because the level P ₁ is selected to be low, a greater useful effect results, but the amounts by which the impedance changes one after the other at the levels P ₁ , P ₂ and P _{3 are} abruptly reduced, also become larger, which causes a greater distortion. The levels P ₁ and P ₃ are expediently made such by selecting the networks R ₁ , R ₂ , R ₃ and A ₁ ', R ₂ , R ₃ determines that the amounts by which the anode alternating current I _a must increase when one of the circuits A ₁ , R ₂ , R ₃ , etc. are added, are equal to one another, since in this case the largest jump that occurs is as small as possible.

From the above it follows that the tubes A ₁ B, 4'.USW. work exclusively as a switch, ie they are either conductive as much as possible or completely blocked. The blocking of tubes A and B or A ',

B 'and A "and B" takes place when exceeding the amplitude level of P ₁ and P ₂ or P ₃ to be amplified vibrations in which levels, as seen from Fig. 2 shows a full scale of the anode voltage E _a of the tube U occurs In one embodiment of the device according to the invention, a low-frequency voltage, optionally obtained after rectifying the vibrations to be amplified, which corresponds to the envelope curve of the vibrations to be amplified, is supplied with negative polarity to a control electrode of the respective tubes. The tubes are z. B. adjusted by means of different grid biases so that the different sets of tubes, e.g. B. the tubes A and B, are blocked one behind the other at the desired amplitude levels of the vibrations to be amplified. The control electrodes of tubes A and B or A ', B' and A ", B" can be connected to one another for this purpose.

Another embodiment of the device according to the invention is shown in FIG. Here, the composite network is obtained in that a number of parallel branches (two of which are shown in Fig. 3) are connected between the anode circuit of the tube U and the load impedance R, each of which consists of two reciprocal networks connected in series, which are shown in Fig 3 are denoted by R ₁ and R ₂ or R ₁ and R ₂ , in the connection point of which a discharge tube A or A- ' working as a switch is arranged, the anode-go cathode impedance of which the input or output circuit of the relevant Bridged networks in the connection point. The mode of operation corresponds essentially to that of the device according to FIG. 1, but with the difference that only one branch can be in operation at the same time, so that the branches must have different wave resistances from one another. That two branches cannot be connected in parallel can be seen as follows; Good for a branch consisting of two reciprocal networks connected in series, for example R ₁ and R ₂

R ₁

11th

where V _n denotes the voltage between the output terminals of the second network and Fj the voltage between the input terminals of the first network. The following applies to the branch consisting of the networks R ₃ and JR ₄

R3

If the two branches, ie the input and output circuits of the two branches, are connected in parallel, the voltages F «and Vi for the

J? 7?

Both branches are the same, what if -— - ~ in contradiction

with the two equations above. Then only the value V _n = F; = 0 simultaneously the two equations, ie the two branches are short-circuited on both sides. Parallel connection of two or more branches is therefore not possible in the device according to FIG. 3.

Another embodiment of the device according to the invention is shown in FIG. Here is the

The anode circle of the amplifier tube and the load resistor R are connected to two opposite corner points of a hexagon, the sides and diagonals of which are formed by reciprocal networks R ₁ to R _9. In all corners of the hexagon, with the exception of the two already mentioned, a discharge tube working as a switch, labeled A, B, C or D, is arranged. Input circuit of the networks in the relevant ίο connection point bridged.

By means of this device, six different values of the anode load impedance R _{a of} the tube U can be achieved by the following combination of conductive and blocked discharge tubes.

I A, B, C, D blocked

II A ₁ B ₁ C ₁ D conductive

III A, B conductive, C, D blocked

IV A ₁ B blocked, C ₁ D conductive

V A ₁ C conductive, B ₁ D blocked

VI A ₁ C blocked, B ₁ D conductive.

4 it can be seen that in combination II only network R _{9 is} switched on, while in the other combinations one or more chains formed by series connection of three of the other networks are connected in parallel to network R _9. The blocking and reopening of the tubes A to D can take place by means of a device with a cathode ray tube, in which the electron beam is deflected under the effect of a low-frequency voltage, which may be obtained after rectifying the vibrations to be amplified and which corresponds to the envelope curve of the vibrations to be amplified, and at increasing amplitude of this voltage one after the other a number of electrodes each connected to a control electrode of the tubes A to D hits, such that the tubes A to D z. B. are blocked in the sequence shown in the table above one after the other and / or alternately.

If only four of the six possible combinations I to VI, specifically the combinations III to VI, are used, the control of the tubes A ₁ B, C and D can be achieved with simple means. This can be seen from the table below, in which the states of the tubes A ₁ B ₁ C ₁ D (g = blocked, 1 = conductive) are indicated with the amplitude of the vibrations to be amplified increasing from zero to the maximum value. The networks A ₁ to R ₉ should be dimensioned in such a way that the combinations used in the order III, V, IV, VI cause the anode load impedance of the tube U in the manner shown in FIG. The amplitude levels at which a jump occurs in the amplitude input impedance of the tube U are denoted by P ₁ , P ₂ , P _3.

A. B. C. D. small amplitude Ill G G' 1 ι the to V G 1 G ι - ¹ reinforcing IV 1 1 G g P ₂ , Vibrations VI 1 G 1 R. 3 ^N

large From this table it can be seen that the tube D or A is _{blocked or opened at level P 2} with increasing amplitude of the oscillations to be amplified, while the tube C or B is blocked or opened at level P ₁ and is opened or blocked at level P _3.

Because a low-frequency voltage, possibly obtained after rectifying the vibrations to be amplified, which corresponds to the modulation envelope of the vibrations to be amplified, is fed together with suitable bias voltages to a control electrode of tubes A and B , tube A or B can easily be at higher negative values than a specific _{instantaneous value of the low-frequency voltage corresponding to the level P 2} is blocked and is conductive at higher positive instantaneous values or vice versa. The desired control of the tubes B and C can be achieved by rectifying the aforementioned low-frequency voltage without smoothing and supplying the resulting direct voltage surges to the control electrodes of the tubes B and C in such a way that the tube B at positive or negative instantaneous amplitudes of the low-frequency voltage below one certain value corresponding to the level P ₃ or P ₁ is conductive and is blocked at instantaneous amplitudes above this limit and the tube C vice versa.

FIG. 5 shows a further embodiment of the device according to the invention, which is very similar to that according to FIG. In the table below, the six useful combinations of blocked and conductive discharge tubes A through D are indicated.

I A blocked, B ₁ C and D conductive

II B blocked, A ₁ C and D conductive

III C blocked, A, B and D conductive

IV B ₁ C and D blocked, A conducting

V A ₁ C and D blocked, B conductive

VI A ₁ B and D blocked, C conductive.

With reference to the devices of Figures 3, 4 and 5, it can be noted that the circuitry of the composite networks composed of the various reciprocal networks used in these devices and still other composite networks not described, also used can be achieved by connecting each point of a group of points via reciprocal networks to each point of a second group of points, the anode circuit of the high-frequency amplifier tube U and the load impedance R being connected to two of the points mentioned. The discharge tubes A, B ₁ C , etc. are connected to the remaining points. 6a and 6b serve to explain this, in which two groups of points a _v « ₂ ,« 3 or B ₁ , b ₂ and b _{3 are} connected by straight lines which each represent a reciprocal network. In Fig. 6 a, the anode circuit of the tube U is connected to the point a _x and the load impedance R is connected to the point O ₁ , if the remaining 1-25 points a ₂ , a _s , b ₂ and b _{3 are} connected with discharge tubes.

are assumed to be bound, this figure is identical to Fig. 5..

Instead of a Röhret or B; To arrange A ' etc. in the connection points of the networks, two tubes connected in push-pull can advantageously be arranged, with no anode voltage being required for the tubes. A suitable device for this is indicated in FIG. L ₁ and L ₂ denote coils coupled to one another, which are tuned to the frequency of the vibrations to be amplified _{by means of capacitors C 1} and C _{2, respectively.} The anodes of the two discharge tubes A ₁ and A ₂ connected in push-pull are directly connected to the ends of the coil L ₂ . The control electrodes of the tubes L ₁ and L ₂ are connected to one another and thus receive the same control voltage. The input terminals k are connected to the input and output terminals of the reciprocal networks in the connection point of the networks. If the tubes A ₁ and A _{2 are} blocked, it can easily be seen that the impedance between the terminals ß has a very small value, while the impedance is high when the tubes A _{1 and A 2 are} conductive, as already noted above ,, the resulting characteristic impedance R _{res of} the circles of three series-connected networks, such as B, the circles R ₁ , R ₂ , R ₃ used in the devices of Figures 1 and 4 are positive; This can be achieved in that one of the networks A ₁ , R ₂ and A ₃ or all three are chosen negative, ie assembled in such a way that the output current _ * _ has a phase shift of -90 ⁰ with respect to the input voltage V, i.e. opposite that of the network R _a . An advantageous embodiment of these circuits is obtained in that a filter network is used for the network R _{2, which has a phase}

• brings about a shift opposite to that of the network R ₀ , while R ₁ = R ₂ = R ₃ , FIG. 8 a shows the circuit thereof. Since R ₁ = R ₂ = R ₃ , the self-induction L and the capacitance C, which are connected in parallel with the tubes A and B , form an oscillation circuit that is tuned to the oscillation to be amplified and has a constant high impedance. This circle can thus be omitted, so that the simple circuit according to FIG. 8b is produced.

Fig. 9 shows a circuit diagram of the embodiment

the device according to the invention according to FIG. 1, in which the networks R ₁ to R ₃ "are all chosen to be the same and are assembled in the manner shown in FIG. 8b.

Instead of reciprocal networks composed of concentrated self-inductions and capacities can also use reciprocal networks with distributed self-induction and capacity, e.g. B. Transmission Lines with a length equal to an odd number of quarter wavelengths to be amplified Vibrations, are applied. Each reciprocal network can also consist of an odd number series-connected reciprocal networks exist.

Claims (10)

PATENT CLAIMS: i. Device for amplifying modulated High-frequency oscillations containing at least one amplifier tube, characterized in that that between the amplifier tube and a load impedance. a composite network is turned on, the elements of which are formed by reciprocal networks, being in each Connection point between two or more of these networks, except at the point mentioned Amplifier tube and the load impedance, at least one discharge tube acting as a switch is arranged, the impedance of which between the anode and cathode the input and output terminals the reciprocal networks bridged in the connection point, which discharge tubes be controlled in such a way that with increasing amplitude of the vibrations to be amplified one after the other and / or alternately at least a part of these tubes is blocked, whereby results in a stepped decrease in the input impedance of the composite network. 2. Device according to claim 1, characterized in that that a possibly obtained after rectification of the vibrations to be amplified Low frequency voltage, which corresponds to the envelope curve of the vibrations to be amplified, together with a bias voltage that is suitably selected for each interrupter, a control electrode the interrupter is fed. 3. Device according to claim 1, characterized in that that an optionally obtained after rectification of the vibrations to be amplified Low frequency voltage, which corresponds to the envelope curve of the vibrations to be amplified, the deflection means of a cathode ray tube is fed, each of which has a number with a control electrode the switching tubes contains connected electrons, which alternate from the cathode ray beam can be taken. 4. Apparatus according to claim 1, 2 or 3, characterized in that the composite Network of a parallel connection between the anode circuit of the high-frequency amplifier tube and the load impedance reciprocal network and a number of chains, each consisting of at least three series-connected reciprocal networks exist, with the control electrodes working as switches Discharge tubes, which belong to a chain of three reciprocal networks connected in series, with one another are connected and controlled in such a way that with increasing amplitude of the amplified Vibrations one after the other blocked the discharge tubes belonging to each chain will. 5. Apparatus according to claim 1, 2 or 3, characterized in that the composite Network is preserved in such a way that everyone a group of points via reciprocal networks with each of a second group of Points is connected, with two of the points mentioned the anode circuit of the high-frequency amplifier tube or the load impedance are connected. 6. Apparatus according to claim 5, wherein each of the two groups contains three points, characterized in that one optionally after Rectification of the vibrations to be amplified obtained low-frequency voltage, that of the envelope corresponds to the vibrations to be amplified, together with suitable pre-stresses is fed to the control electrodes of two of the discharge tubes operating as switches, in such a way that that one of these tubes (.4) at below one ίο a certain level lying amplitude of the vibrations to be amplified, the other tube (D) is blocked above this level, at the same time the mentioned low-frequency voltage is rectified without smoothing and fed to the control electrodes of the other two discharge tubes working as switches that when below a second or At the third level positive or negative instantaneous amplitudes of the low-frequency voltage mentioned, one of the last-mentioned two discharge tubes (C) is blocked, while the other of these two tubes (B) is blocked at instantaneous amplitudes above this level.
7. Device according to one of the preceding claims, characterized in that the amplitude level the vibrations to be amplified, in which the discharge tubes working as switches are locked, are selected for the various discharge tubes so that the Amount by which the alternating current component of the anode current of the relevant high-frequency amplifier tube each time a discharge tube is blocked or several discharge tubes are blocked at the same time increases, is always the same size. 8. Device according to one of the preceding claims, characterized in that optionally any changes in shape that occur are eliminated through the use of negative feedback will. 9. Device according to one of the preceding claims, characterized in that in the connection points of the reciprocal networks are arranged two discharge tubes connected in push-pull, working as switches, to which the Control electrodes with each other and the anodes with the ends of one on the frequency of the to be amplified Vibrations coordinated oscillation circuit are connected to the relevant reciprocal networks is inductively coupled. 10. Apparatus according to claim 4, 5, 6, 7, 8 or 9, characterized in that each chain has three series-connected reciprocal networks of a parallel capacitance, a series self-induction, a series capacitance, a second series self-induction and a second parallel capacitance are formed with discharge tubes working as switches between the series capacitance and one of the two series self-inductions are arranged and all self-inductions or capacities are equal to each other. 1 sheet of drawings © 5606 12.52