US2504626A - Frequency changer - Google Patents

Description

A ril 18, 1950 D. A. BELL 2,504,626

FREQUENCY CHANGER Filed Sept. 20, 1944 2 Sheets-Sheet 1 art/11W DAVID A BELL April 18, 1950 D. A. BELL 2,504,526l

FREQUENCY CHANGER Filed Sept. 20, 1944 2 Sheets-Sheet 2' DJSVID FM BELL Patented Apr. 18, 1950 FREQUENCY CHANGER David ArthurBell, :London, "England, assignor ,to

1C. :CossorjLimited, London, England, a Britz-ish company (Application September 20, 1944, Serial'No. 555,029 In Great Britain October ll, 1943 Thls invention relates to "the operation of frequency changer valves of the kind having a cathodatwo output electrodes, a first input electrode system controlling the total current from the cathode :to the two output electrodes, and a second input electrode system controlling the distribution-of this current between the two output electrodes. An :example of such a valve is thehexode.

A "large part of the noise produced in valves ofthis kind is due'to current fluctuation arising from the random distribution of individual electrons between the -two output electrodes which occurs when the cathode current is divided between these electrodes. One purpose of this in-- vention is to eliminate this noise factor as far as possible.

According to one feature of the invention, one of the oscillations to be mixed is applied to the second input electrode system in such manner that substantially the whole of the electron current is switched from one output electrode to the other during a very-small fraction of a cycle. For example, the duration of each transition should not exceed one-sixteenth of the whole cycle period, *1. e. the two transitions occurring in each cycle should together occupy less than one-eighth of the "period.

In order to carry this feature into effect, a hexode having its screen grid aligned with its first input grid "may be employed, and the input voltage to the *second input grid may have rectangular waveform. The amplitude of this input voltage and the bias will be such that (I) when the second input grid is at its lower potential current is substantially wholly drawn by the screengrid, and (2) when it is at its higher potential the current is substantially wholly drawn by the anode. The degree of control efiected by the secondinput grid should be such that this'higher potential is not so high that-appreciable current is drawn by the second input grid; otherwise this will itself introduce partition noise.

While it is desirable that thewhole of the current should be rapidly transferred from screen grid-to anode and vice versa, a useful reduction in noise will result if such transfer of only 80% of the current is achieved. The expression substantially whollyf is therefore to be understood as covering a figure of this order.

In an "alternative arrangement, the electrodes are so constructedthat the whole of the cathode current isswitched rapidly from one output electrode to the other by a small fraction (say less '3 Claims. (01. 250 --20) n/ a ll than one-fifth) of the peak value of a sine wave alternating voltage on the second input electrode system. This arrangement can more easily be carried into efiect'if the secondtinput electrode system operates by beam deflection rather than by ordinary space-charge control.

If it'is desired that the sum or difierence'sbeatfrequency output from the valve shall be amplitude modulated in accordance with the amplitilde-modulation of one'of the input oscillations (as for example in the frequency-changer valve of a superheterodyne receiver for amplitudemodulated radio signals), then this input oscillation is-applied to the first input electrode system. The oscillation to be mixed therewith, the amplitude variations of which are irrelevantyis applied to the second input electrode system.

If the oscillation for the second input electrode system is to be locally generated, this may most conveniently be efiected in a separate electron stream, e. g. in the triode section of a triodehexode valve, the hexode section of which'is serving as mixer.

Another feature of the invention is the connection of the output circuit in push-pull between the two output electrodes of the frequencychanger valve. This gives the advantage that when the amplitude of the total electron current is changed, but not the proportion in which, when integrated over a number of cycles, itis distributed between theoutput electrodes, no difference beat frequency output signal is produced. From this it follows that an output signal cannot be produced solely by a voltage appearing at the first control electrode; but only by co-operation between separate voltages on the first and second control electrode systems Whose frequencies differ by the frequency to which the output circuit is tuned. This gives protection againstseveral types of interference. For example, an unwanted signal, having the frequency'to which the output circuit is tuned, and which has not been sufficiently rejected by the input circuit to the first control electrode, will cause severe interference in a conventional type of frequencychanger, but will have no effect in a frequencychanger having .a push-pull output circuit. For another-example, there maybe two powerfulsignals, separated in frequency by the frequency to which the output circuit is tuned, both present at the first control electrode, whether or not one of them is the signal to which the input circuit to that electrode is tuned. In a conventional frequency-changer, either -(a) rectifying action due to the flow of grid current, or (b) non-linearity of the space-current/grid-voltage characteristic, may then result in a component at difference frequency in the space-current, and this will cause an interference output signal. This again is avoided by using the push-pull output arrangement.

In the accompanying drawings, Figure 1 is a circuit diagram of the frequency-changer stage and associated circuits of a superheterodyne radio receiver; Figures 2 and 3 are diagrammatic elevation and sectional plan views respectively of the electrode structure of an electron beam deficction valve suitable for use in the circuit of Figure l; and Figure 4 is'a circuit diagram of an alternative frequency-changer stage for a superheterodyne radio receiven wherein is employed a triode-hexode valve using space-charge control throughout.

In the arrangement shown in Figure 1, the total current emitted by cathode I of valve 26 is spacecharge controlled by the first control electrode 2; it is" then partially focussed by an electrode 3 maintained negative to the cathode, and is next passed through an accelerating electrode 4. Electrodes 2 and 4 are a pair of aligned grids, so that no appreciable current is drawn by electrode 4. After leaving electrode 4, the electron stream is subjected to deflection control by a pair of deflector plates 5. According to the voltage applied between plates 55, substantially the whole electron current falls either on anodes 6 and T or on anodes 8 and 9. It is merely for diagrammatic convenience that anodes I and S are drawn below anodes 6 and 8, as these are actually disposed all in one plane perpendicular to the mean axis of the electron stream.

Amplitude-modulated radio signals are picked up by aerial Ill and selected by radio frequency tuned circuit I I. They are they applied through condenser I2 to the first control electrode 2, which is biassed through grid leak I3 from a tapping on a self-biassing network I 4, I5, IS. The focussinlg electrode 3 is more heavily negatively biassed from the lower end of this network, which is shown as earthed. Thus the electron stream passing upwards through accelerating electrode 4 will be amplitude-modulated in accordance with the signal input applied to the first control electrode 2.

Deflector plates are push-pull connected through tank circuit I9, and anodes I and 9 are push-pull connected through tickler coil 25. A local oscillator system is thus formed, the frequency of which is determined by the tuning of tank circuit I9. The mid-points of coil and of the coil of tank circuit I9, and also accelerat ing electrode 4, are maintained at a steady high positivepotential by means of dropping resistor I1 from the high tension source and condenser I8 to earth.

The deflections of the beam produced by deflector plates 5 should be large relative to the width of the gap between anodes 6 and 8. The effect of the deflection voltage applied to deflector plates 5 will then be that substantially the whole electron current is rapidly transferred between anode E .and anode 8; it will only be divided between them for a very short transition period every half-cycle.

The primary of I. F. transformer 20 is push pull connected between anodes 6 and 8. Both primary and secondary are tuned to the intermediate frequency. The intermediate frequency output from the circuit is developed between terminals.

Risk of feedback to the first control electrode is reduced by the fact that both output and oscillator circuits are balanced. The electronic coupling which is produced in a space-charge controlled valve, and which causes feedback from second control electrode system to the first, is avoided by the use of balanced deflection control for the second control electrode system.

A preferred electrode arrangement for the valve 26 of Figure 1 is diagrammatically illustrated in Figures 2 and 3. Figure 2 is an elevational view with parts broken away to reveal the interior construction. Figure Sisa cross-sectional view tak en on the plane indicated at AA in Figure 2.

It will be seen from Figure 3 that the valve is symmetrical, current from both faces of the flat cathode I being usefully employed. Next to the cathode I is the first control grid 2 wound on backing wires 22. Outside grid 2 the beam-forming electrode 3 is placed, consisting of two channels of sheet material. All of these partsare surrounded by grid 4 which is supported on back ing wires 24 and is aligned with grid 2 in knownmanner. Separate pairs of deflector plates 5 are provided for the electron streams issuing in opposite directions through grid 4. Finally, the whole structure is surrounded by anodes 6 and 8.

The part of the electrode structure seen at the bottom of Figure 2 serves for the generation of the local oscillations. It comprises separate anodes I and 9 and extensions of cathode I, beamforming channels 3 and deflector plates 5. The grids 2 and 4 do not extend into this section of the valve, and the cross-section on plane BB of Figure 2 will therefore be identical with Figure 3, except that these grids are omitted. One advantageous result from their omission is that the capacity between cathode I and first control electrode 2 is minimized.

The valve, the electrode structure of which is digrammatically shown in Figures 2 and 3, may be conventionally mounted on an ordinary valve foot and enclosed in an ordinary glass valve envelope. Conventional means may also be used 1 for supporting and spacing the electrodes.

Another arrangement for the frequency-changer stage of a superheterodyne radio receiver is shown in Figure 4. A triode-hexode valve 36 is employed having a triode section formed of cathode I, grid G and anode P, and the hexode sec-. tion including cathode I, grid Ga, screen S, grid GI) and anode A. The input circuit to first control grid Go. of the hexode mixer from aerial Ill is conventional and comprises radio frequency tuned circuit I I, grid coupling condenser I2, grid leak 3| and biassing network 32.

The triode oscillator section of valv 36 comprising cathode I, grid 2 and anode P is connected into a tuned-anode local oscillator circuit comprising tank circuit 33, tickler coil 34, grid condenser 35 and grid leak 36. A resistor 3i, having a conductance which is small relative to the grid-cathode conductance, is inserted in the oscillator grid lead for two purposes:

(1) It minimizes the current collected by the second control grid Gb of the hexode which is directly connected to the oscillator grid 6-. A substantial current collected by this second control grid would itself introduce partition noise into the hexode circuit.

(2) It prevents the grid condenser 35 and grid leak 35 from performing their usual function of biassing back the oscillator grid G so that zero or small positive potential is only reached at the peak of the oscillation cycle. It is therefore ef:

L; fective to produce an approximately square wave of positive potential on the oscillator grid in place of the usual part of a sine wave.

The output circuit comprises I. F. transformer 33 having both primary and secondary tuned to the intermediate frequency. The primar is. con nected in push-pull between anode A and screen S, through blocking condenser 39. Screen current is supplied from the high tension source through one half of the primary and through potentiometer 48, M. The intermediate frequency output signal is developed between terminals 42.

I claim:

1. A frequency-changer comprising a thermionic v'alve having a cathode. an anode, a screen grid, means connecting said cathode and anode to establish an electron stream from the oathode directly to the anode, said screen grid being positioned in said stream between said cathode and said anode, a first input grid aligned with said screen grid and disposed in said stream between said cathode and said screen grid to control the total current from said cathode to said screen grid and anode, and a second input grid disposed in said stream between at least part of said screen grid and said anode to control the distribution of current from said cathode between said anode and said screen grid, means to apply an oscillation of a first frequency to said first input grid, means to apply to said second input grid an oscillation of a second frequency and of rectangular waveform to switch over the electron stream from said anode to said screen grid and vice versa during very small fractions of a cycle in every half-cycle of said second frequency oscillation, and a circuit connected between said anode and said screen grid responsive to current of diiference frequency between said first and said second frequencies.

2. A frequency-changer comprising a thermionic valve having a cathode, an anode, a screen grid, means connecting said cathode and anode to establish an electron stream from the cathode directly to the anode, said screen grid being positioned in said stream between said cathode and said anode, a first input grid disposed in said stream between said cathode and said screen grid to control the total current from said cathode to said screen grid and anode, and a second input grid disposed in said stream between at least part of said screen grid and said anode to control the distribution of current from said cathode between said anode and said screen grid, means to apply an oscillation of a first frequency to said first input grid, means to apply an oscillation of a sec ond frequency to said second input grid, and output circuit connected in push-pull between said anode and said screen grid responsive to current of difference frequency between said first second frequencies.

3. A frequency changer according to claim 2 wherein said output circuit is tuned to said difference frequency.

DAVID ARTHUR BELL.

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

UNITED STATES PATENTS Number Name Date 2,924,97 9 Metcalf Dec. 1'7, 1935 2,073,454 Peterson et al Mar. 9, 1937 2,107,333 Jonker Feb. 8, 1938 2,190,515 Hahn Feb. 13, 1940 2,191,903 Aldous Feb. 2'7, 1940 2,225,330 Cage Dec. 17, 1940 2,269,588 Bath Jan. 13, 1942 2,270,791 Strutt et a1 Jan. 20, 1942 2,271,716 Salzberg Feb. 31, 1942 2,325,654 Cohn Aug. 3, 1943 2,342,987 Bosch Feb. 29, 1944

Images