US2135615A - Multipactor - Google Patents

Description

Nov. 8, 1938.

P. T. FARNSWORTH 2,135,615

MULTIPAGTOR Filed Feb. 11, 1936 2 Sheets-Sheet l llili J IN V E N TOR,

PH/LO 7. FA R/VSWOR TH- ATTORNEYS.

1938- P. T. FARNSWORTH 2,135,615

MULTIPACTOR FiledFeb. 11, 1956 2 Sheets-Sheet 2 flhl I l l l F I I I I W IN V EN TOR, PH/LO 7f FARNSWOR TH.

A TTORNEYS.

Patented Nov. s, 1938 UNITED STATES PATENT OFFICE MULTIPACTOR Philo T. Farnsworth, San Francisco, Calif., as-

signor to Farnsworth Television Incorporated, fSanr-ifrancisco, Calii'., a corporation of Cali-' Application February 11, 1936, Serial No. 63,424 3 Claims. (01. 250-215) provide a multipactor which will give any desired degree of amplification; to provide a multipactor which requires no external field for guiding the electron flow; to provide a multipactor wherein no filament circuit is necessary, so that any portion of the device may be operated above ground potential; to provide a device which may be used either as a detector or as an amplifier having linear characteristics; to provide a multipactor which may be excited by either direct or alternating potentials; and to provide a device wherein the degree of amplification obtained is completely under control.

My invention possesses numerous other objects and features'of advantage, some of which, to-

gether with the foregoing, will be set forth in the following description of specific apparatus embodying and utilizing my novel method. 'It

is therefore to be understood that my method is applicable-to other apparatus, and that I do not limit myself, in any way, to the apparatus of the present application, as I may adopt variousv other apparatus embodiments, utilizing the method, within the scope of the appended claims.

Referring to thedrawings,

Figure 1 is a view showing a preferred embodiment of the multipactor tube of this invention I in longitudinal cross section.

Figure 2 is a similar view, partly in elevation and partly in section, showing schematically the external connections of the device as used as a detector-amplifier in a radio receiving circuit.

Figure 3 is a diagrammatic representation of the fundamental circuit of the multipactor itself, with such external circuit elements as are necessary for an explanation of the operation.

Figure 4 is a sectional view of my invention utilizing a resistive coating as a voltage divider.

Figure 5 is a similar view showing the use of a tubular electrode of resistive material as a voltage divider.

Considered broadly the multipactor of this invention comprises a primary cathode from which initial electrons are emitted, these electrons being directed at a series of foraminated secondary cathodes which are capable of emitting secondary electrons at a ratio to impacting primaryelectrons which is greater than unity. Means are provided for maintaining a substantially uniform or increasing potential gradient along this series of secondary cathodes, each acting as an anode to the preceding cathode in the series, and means are also provided for collecting the electron fiow from the last cathode of the series. There is also preferably provided an electrostatic shield between the primary cathode and the first of the secondary cathodes which may be operated.

as a control electrode or grid, and for many purposes it is also advisable that the primary cathode should be photo-emissive rather than thermionic. This arrangement may be better understood by reference to the preferred form of the device shown in the drawings, wherein the reference character I refers to a tubular vitreous envelope which is evacuated to as high degree as is practical. This tube is symmetrical, and sealed through each end is a lead 2, 2' which is connected to a primary cathode 4, 4'. These cathodes may conveniently be-madeof silver, and provided with photo-electric surfaces of the well known silver-oxide-caesium type.

- Mounted in front of each of the cathodes and slightly spaced therefrom is an electrostatic shield 5, 5'. These shields comprise metal screens, each formed of very fine wire with relatively open mesh so it will be highly permeable to the electron stream. As it is undesirable that the shields should emit secondary electrons they may be made of nickel with carbonized surfaces.

Spaced uniformly between the two electrostatic shields is a series of secondary cathodes B to l3, inclusive. Each of these cathodes is foraminated, the ratio of their over-all area. to the combined area of the foraminations being preferably of the order of 3 or 4 to 1. The cathodes may be made of sheet material and the openings punched therein, or they may be made of fine meshed screen suitably supported, as is shown in the drawings, the latter construction being somewhat preferable. Like the primary cathode the material of their construction is preferably silver and they are provided with silver-oxide-caesium surfaces which render them photoelectric and also capable of emitting secondary electrons at a ratio to impacting primaries of 6 to 8. His of some slight advantage if the holes. through the cathodes are not alined, but as will be shown later, this is of relatively small importance as regards the over-all efiect of the device.

The first secondary cathode 6 and the last secondary cathode l3 are provided with leads l5 and I6, respectively. It is also convenient for some purposes to provide individual leads II for the intermediate cathodes I to l2.

Preferably, and necessarily if the leads ll be not provided, the interior wall l9 of the tube l is metalized between the cathode 6 and the cathode l3, so that the tube itself forms a hollow resistor 50 with the secondary cathodes in electrical contact therewith. The shields 5 and 5' should, be insulated from this metalized surface. Any of the well known conducting coatings may be used, a coating of nickel which is so thin as to be transparent being quite suitable for the purpose, and from the standpoint of power conservation it is desirable that the resistance between the cathodes 6 and I3 be of the order of from 50,000 ohms to 1 megohm. These are not limiting values, however.

The fundamental circuit device as ordinarily used is shown in Figure 3. The signal voltage, which may be considered as supplied across the tuned circuit 20, is supplied to the cathode 4, and the shield 5 is grounded. A source of potential 2| supplies to the secondary cathode l3 a positive potential with respect to ground. The impedances 22 represent the resistive coating on the inside of the tube, or, when this is absent, they may be connected externally as shown. When so connected, the impedances 24 should be about half of the value of the impedances 22, and an additional impedance 25 connected between the cathode 6 and ground, should be of such value that the potential drop between ground and the cathode 6, and that between each of the successive cathodes of the secondary series is about the same, thus establishing a practically uniform potential gradient along the tube. For best results, however, the uniformity should not be exact, but the gradient should increase slight- 1y from the cathode 6 to the anode, as this assists in getting. all of the newly emitted secondaries 1 through the aperture. The device will operate even with a decreasing gradient, although less effectively.

The cathode 4 is connected through an output resistor 26 to a point on the source 21 which will give a like potential drop between the oathodes l3 and 4 to that between any pair of adjacent secondary cathodes, while the electrostatic shield 5' is, in most instances, a few volts positive to the cathode 4.

If, now, the cathode 4 be illuminated from a source of light 21, photoelectrons will be emitted therefrom, and will pass through the electrostatic shield 5 to impinge upon the secondary cathode 6. When reaching the cathode these electrons will be traveling at relatively high velocity and a proportion of them will impact the cathode while others will pass through it, the relative numbers being almost exactly in the ratio of solid area to open area in the cathode. If the cathode be of wire netting, with the spaces between the wires equal'to the diameter of the wire, the proportion intercepted will be almost exactly threefourths. v

The impacting primaries will release secondary electrons from the cathode 6 at a ratio which is dependent upon their velocity of impact. The secondaries will be of low velocity, and experience has shown that practically all of the released electrons will be drawn through the apertures and accelerated toward the secondary cathode I, where the phenomena of release of secondaries by impact will be repeated. The formation of the cathodes will, in general, cause a strong curvature of the lines of force at the apertures or foraminations, so that it is the exception rather than the rule when an electron travels straight through; in general they will be deflected, and diffused practically uniformly over the succeeding secondary cathode. As far as the number of electrons which impact the successive secondary cathodes is concerned, it therefore makes practically no difference whether the apertures be aligned or not.

The current multiplication that takes place at each stage is dependent upon the ratio P of secondary to primary electrons and the ratio q of the total area of the secondary cathode to the combined area of cathode and the apertures therein. With an initial number of primary electrons m, the number of electrons leaving the secondary cathode 6 will be m(Pq+lq), the first term within the parenthesis representing the proportion of secondary electrons liberated at the secondary cathode, while 1q represents the proportion of the electrons which are primaries that have passed through the apertures. A similar increase takes place at each of the secondary cathodes, and for the total device the current amplification is expressed by the ,quantity m(Pq+1-q)", where n is the number of secondary cathodes.

Upon leaving the cathode 13, the electrons pass through the electrostatic shield- 5' (where a few of them will be collected) and will impact the cathode 4, where further release of secondaries will occur. These final electrons are attracted toward the shield 5' at low velocity and will be collected by the shield.

The degree of amplification which may be obtained by a device of this character may be readily computed. With the electron-emitting surfaces described, the value of P may easily be from 6 to 8, and can, by special precautions, be made still higher. Taking P as 6, and q as .75,

' the amplification at each of the foraminated secondary cathodes is, 4.75. As there are eight secondary cathodes shown in the illustrated embodirnent, the current amplification up to and including the cathode I3 is 4.75 or something over 244,000. With a ratio of six secondaries to one primary from the cathode 4', the current in the cathode circuit will be five times this 244,000+, while that to the shield 5' will be six times this value, giving a current amplification of about 1,220,000 or 1,460,000, depending upon whether the output circuit be taken as the circuit of the cathode 4 or that of the shield 5. Either electrode may be used as the output, the two being 180 out of phase, and the desirability of using one or the other electrode as the output circuit depending upon the phase upon which it is desired to use the output signal.

The ratio of the current to the collector 5' and that in the circuit of the cathode 4 will always be since the current through the impedance 2G is decreased by that supplied by the electrons impacting on the cathode I.

The over-all amplification of the device can be controlled by varying voltage supplied by the source 2| between ground and the cathodes 4' and I3. With practically all secondary emitters,

the ratio of secondary to primary electrons increases with voltage from zero to a maximum,

the increase being substantially linear for small voltages and falling on as the voltage increases.

Above the maximum point the ratio decreases .very gradually with increasing voltage, the maxior something less than one in 65,000. This, however, is basedupon the assumption that the impacting electrons are traveling parallel to the axis of the-device, and it has already been shown that they will, in general, make a slight angle therewith, and since the cathodes actually have a finite thickness the chance of any one electron getting all the way through without making an impact is even less than that set forth. Less than one electron in sixteen will pass through two successive cathodes and less than one in sixty-four through three. The final amplification may therefore be controlled very'accurately by regulating the potential drop through which the majority of the electrons will fall before an impact, the device ceasing to amplify altogether if the ratio of secondaries to primary electrons become unity.

Figure 2 shows schematically a method of utilizing the tube in a radio-receiving circuit, the impedances 22 being assumed in this case to be the metalized tube surface. Here the signalvoltage is supplied from an antenna or radio frequency transmission line 30 coupled to the inductor of the tuned circuit 20. The shield 5, insteadof being connected to ground directly, is connected thereto through a biasing resistor 8|, shunted by a "grid condenser 32. If the light from the source 21 be filtered so that that falling on the cathode is all in the red or infra-.

red region of the spectrum, the initial velocity of emission of electrons from the cathode 4 will be low and quite uniform, and the number collected by the shield will be very accurately controlled ,by the signal voltage, increasing the efficiency of detection. If white" or mixed radiahence the performance of the device. With a low value of resistance 3| and relatively large condenser 32. the bias may be made such that the device amplifies linearly, so that the voltage, which appears across the output impedance 26 and is applied through the blocking condenser N to any succeeding element in the circuit, will be at radio frequency. By increasing the resistance oi the impedance ll and decreasing the size of the grid condenser, so that the time constant of the circuit is of the proper value for gridleak detection, the device will act as a detector and the output will vary at audio frequency. It is obvious that the device can also be used as a. detector by so biasing the shield that flow to the cathode 6 occurs only during the negative half of the potential cycle as impressed on the oathode l.

. It will be seen that the performance of the device may be varied greatly in character by relatively slight changes in the output circuit, particularly by varying the relative potentials of the cathode 4' and shield 5'. If the shield be sufficiently positive so that all the electrons emitted by the cathode 4' are collected, the last stage of the device operates as a linear amplifier and increased current causes the electrode 4' to swing positive. If the shield be made sufficiently negative so that all electrons emitted by the electrode 4 are returned thereto, the final stage of multiplication is without effect, the amplification derived from the device is simply that due to the foraminated secondary cathodes and the electrode 4' becomes the true anode of the device and swings negative with increased current. If, however, the relative potentials of the shield and electrode 4' be such that heavy current from the cathode would swing the electrode 4' positive with respect to the shield, while with lighter currents the shield is the more positive, the current through the impedance 26 can build up only to a certain definite maximum, but below this maximum will vary in proportion to the electron flow. Under these conditions the output circuit acts as a detector, maximum detection occurring if the device he so adjusted that maximum current flows in the impedance 26 at zero signal.

It will also be apparent that as far as the relative potentials of the cathode I and shield 5' are concerned, it makes no difference to which of these electrodes the impedance 26 is connected, the only diflerence being the phase of the voltage drop therethrough.

Where detection is accomplished in the output circuits, the initial cathode 4 and shield 5 may be used to supply automatic volume control for the system, the integrating circuit 3l-32 being given a time constant whichis long in comparison with the period of the lowest modulating frequency present in the signal, and the impedance of the circuit being sufilciently low so that the shield will block the emission from the cathode l to only the desired degree.

With proper adjustment of the voltages on the electrodes 4' and 5', and proper choice of the impedance 26, the negative voltage characteristic will cause oscillation in this circuit, and either heterodyne or zero-beat detection may be accomplished with the device. Furthermore it is not necessary that the input be applied at radio frequency, for the apparatus will operate equally well on audible or sub-audible frequencies.

Because the device is symmetrical it is possible to replace the D. C. source 2| by an A. C. source of low or high frequency, .and the multipactor will operate to amplify or modulate at such frequency giving many of the advantages of the radio-frequency type of multipactor .as described selenium, or Thyrite, having appreciable thickness, and may be self-supporting or deposited on the glass wall as required to give the necessary structural strength together with the desired resistance. Furthermore, the secondary-emitting surfaces need not be photo-electric, but may be of aluminum,- nickel, or any other material whose secondary'electron primary electron ratio is greater than unity.

I claim:

1. A current amplifier comprising an evacuated envelope, *9, primary cathode within said envelope, a series' of secondary cathodes each comprising awoven screen having a surface ca- 1 pable ofemitting secondary electrons at a ratio to impacting primary electrons greater than unity, an electrostatic shield screen positioned between said primary cathode and said secondary cathodes, and means for collecting the electrons emitted from the lastsecondary cathode of said series, said shield screen having a mesh larger than the mesh of said cathodes.

2. An amplifying and detecting tube comprising a photo-emissive cathode, an electrostatic screen of small projected area mounted in front of said cathode, a series of foraminated. secondary cathodes mounted in spaced relationship in front of said screen, each of said secondary cathodes having a surface capable of emitting secondary electrons at a ratio to impacting primaries greater than unity and the foraminations in adjacent cathodes being offset to reduce the probability of an electron emitted from said primary cathode traversing the entire series of seconda'ry cathodes without impacting thereon, and

means for collecting the electron current emitted from the last cathode of the series.

3.'A tube in accordance with claim 2 wherein the last cathode and collecting means comprise respectively a photo-sensitive electrode and an electrostatic screen substantially similar to the cathode and its screen therein previously described.

PHILO T. FARNSWORTH.

Images