US 3317846 A
Abstract available in
Claims available in
Description (OCR text may contain errors)
May 2, E967 Filed OCI. 24, 1965 I-IUGI-I L. DRYDEN,DEI=UTY 3,317,846 ADMINISTRATOR OE TI-IE NATIONAL AERONAUTICS AND SPACE ADMINISTRATION LINEAR ACCELERATOR FOR MICROMETEOROIDS HAVING A VARIABLE VOLTAGE SOURCE 4 Sheets-Sheet 2 HWI) I I I/\24 I I I I I I 2s I I// FROM VELOCITY DETECTOR i 23 22 sTAFITI- LW STOP I 3| I I I I MEASUREMENT I I I I COUNTER I CLOCK I I /33 I I I I FREQUENCY I I I COUNTER I I I I 37) I OUTPUT A |"I I--I I-II TO FL'P I ,CONTROL FREQUENCY FLOP I SYSTEM GENERATINC mm 40 SISTEM lo. OUTPUT 8 I L I INVENTORS David G. Becker Joseph FFfIIcITIenIChI Bernard Hamermesh Roberv. I On muir BYQQMS* Q CEMFM ATTORNEYS May 2 3957 HUGH I.. DRI/DEN, DEPUTY 3,317,846
ADMINISTRATOR OF THE NATIONAL AERONAUTICS AND SPACE ADMINISTRATION LINEAR ACCELERATOR FOR MICROMETEOROIDS HAVING A VARTABLE VOLTAGE SOURCE Filed Oct. 2A, 1965 4 Sheets-Sheet 5 TO CONTROL l SYSTEM 40 m u n @Il I E* o o V INVENTORS EE I- I-I DQvId G. Becker QI- j-w- J I Joseph F. Frchtencht gg I Eernafd Hamermesh Raber V. Langmuir ATTORNEYS May 2, E957 HUGH LA DRYDEN, DEPUTY 3,337,846
ADMINISTRATOR oF THE NATIONAL AERONAUTICS AND SPACE ADMINISTRATION I LINEAR ACCELERATOR FOR MIGROMETEOROIDS HAVING A VARIABLE VOLTAGE SOURCE Filed Oct. 24, 1965 4 Sheets-Sheet 4 H IG H VOLTAGE SYSTEM I3: I; E O I5 0 I I' Mm I I I I I- I I 2 Z :a D l O O U U I I l I (90 I MII E I gm i I f d S@ I I I S glow z I INVENTORS (T EE David@ EecII @seph F. Fmcmemch? OZ @E M; M ATTORNEYS Patented May 2, 1967 Filed Oct. 24, 1965, Ser. N0. 505,076
1 Claims. (Cl. 328-233) This invention relates to linear particle accelerators, and more particularly to apparatus for accelerating microparticles to hypervelocities.
Introduction It is well known that micron sized particles such as dust particles and micrometeoroids form an important constituent of interplanetary space. Signicant numbers of such particles exist having diameters in the order of ya fraction of a micron to a few microns, and travel through space at velocities of 20 km./seconds or more. Information as to the effects of the impact of such particles on various types of materials has a great deal of theoretical scientific value, and is of practical value in design consideration for the skin of spacecraft in order to protect instruments .and/or occupants thereof. It is therefore readily apparent that it is highly desirable to simulate these hypervelocity microparticles in a laboratory environment in order to study Various phenomena such as irnpact light flash, impact ionization and their interaction with gaseous and solid targets.
Hypervelocity research is presently limited by the fact that no practical apparatus exists for producing microparticles in the laboratory at average velocities greater than approximately 10 km./second, that is, applyin-g a 20 megavolt accelerating potential to dust particles of the approximate diameters mentioned and having a q/m (charge-mass ratio) of approximately l-l Coulomb/kg. At these values of q/m, much lower than those of nuclear particles, accelerators using magnetic guide iields are not feasible, and accelerators which use only electrostatic elds (i.e., Van de Graad' accelerators, linear accelerators and synchrotrons with electric guide elds) ofer the most promising potentialities.
The 1.6 mv. Van de Graai accelerator produces particles at 20 km./sec., with occasional small particles at 30 km./sec., although l0 km./sec. is the average. The repetition rates for particles above km./sec. are low, and the number of particles produced at higher velocities is considered too small to permit significant data to be accumul-ated for experimental work. In addition, Van de Graafr accelerators of a tandem type, requiring charge reversal of the charged particle at the mid-point of the acceleration process, have been developed which may be used to accelerate nuclear particles to 10-20 mv., but are not practical for use with charged dust particles because of their large q/m which gives rise to insuperable problems involving charge stripping and sign reversal. Electrostatic synchrotrons, on the other hand, give rise to severe structural problems in terms of dimensional tolerances and physical size, and tend to be ineiiicient because of loss of charged particles,\traveling in a relatively long circular path at high velocities, by way of collisions With residual gas molecules. These problems make electrostatic synchrotrons less desirable as practical laboratory apparatus than theoretical considerations would indicate.
Linear accelerators are in many respects naturally suitable for [accelerating micron sized particles of the q/m involved in hypervelocity research. One type of such accelerators consists of a series of hollow drift tubes that are axially aligned, with alternate drift tubes connected to opposite terminals of an alternating voltage source. The frequency of this voltage, the length of the drift tubes and the width of the gaps therebetween are .adjusted so that a particle passing through each gap experiences an accelerating iield. The net acceleration potential is the peak voltage applied to the drift tubes multiplied by the number of gaps.
Linear accelerators of this type are normally operated in a phase stable mode so that bunching occurs, and many particles are accelerated at one time. Since phase stability and radial focusing cannot occur for the same accelerating potentials, radial focusing is provided -by electrostatic grids or quadrupole magnets. However, both focusing schemes are impractical for use with microparticles accelerators. The grids produce a large drop in transmission, and the magnetic sources required for focusing are extremely large because of the q/m ratios involved. In addition, in hypervelocity research it is desirable to accelerate one particle at a time and to insure that as great a proportion of particles as possible are successfully accelerated even though individual particles may vary as to q/ m and initial velocity.
Accordingly, it is an object of the present invention to provide new and improved apparatus for accelerating microparticlcs to meteoric velocities.
Another object of the invention is to provide a linear accelerator for accelerating single dust particles of varying initial velocity and charge-mass ratios in a manner which insures that .a maximum number of particles are successfully accelerated.
A further object of the invention is to provide a linear accelerator having apparatus for imparting extremely high accelerating energies to single dust particles from a relatively low voltage source.
Yet another object of the present invention is to provide a novel linear accelerator that is radially stable, thereby eliminating the need for auxiliary focusing arrangements.
The Yforegoing and further objects of the present invention are obtained by the provision of Ia linear accelerator which consists of a 'plurality of axially aligned hollow cylinders (or drift tubes), with alternate drift tubes connected to opposite terminals of an alternating voltage supply. The lengths of the tubes, the frequency of the alternating voltage and the width of the gaps between the drift tubes are adjusted so that a particle passing through the drift tubes is accelerated at each gap, with total acceleration for the particle equivalent to the product of the peak voltage applied to the drift tubes and the number of gaps. Individual particles having a positive charge are injected into the drift tubes by a dust accelerator. The particle accelerator is operated radially stable (and phase unstable), and the alternating voltage supplied to the drift tubes is a modified square Wave whose peaks have a constant, non-zero slope (for example, a drooping square wave whose peafks have a constant negative slope), such that a single particle passing through the drift tubes is always subjected to an accelerating voltage, with adequate radial focusing provided and time-phase instability minimized by the constant non-zero slope of the square wave. In addition, in order to accelerate incoming particles having widely varying values of q/m, each particle entering the drift tubes passes through a velocity sensor. An output signal derived from the velocity sensor controls the frequency of the drooping square wave to automatically accommodate each incoming particle. Accordingly, the frequency of successive drooping square wave pulses is adjusted to match the q/m of each particle being Iaccelerated, resulting in successful acceleration of a large portion of incoming particles.
The various features and attendant advantages of the invention will bec-ome readily apparent from the ollorwing description when considered in conjunction with the accompanying drawings, in which:
FIGURE 1 is a schematic diagram depicting the linear accelerator of the present invention;
FIGURE 2 is `a representation of a modified square wave utilized with the linear accelerator of the invention;
FIGURE 3 is a schematic diagram of one embodiment of the frequency generating system used in conjunction with FIGURE 1;
FIGURE 4 is a schematic diagram of one embodiment of the high voltage system used in conjunction with FIGUREl; and
FIGURE 5 is a simplified schematic diagram illustrating downstream velocity correction of the linear accelerator of FIGURE l.
Before describing specific embodiments of the present invention in detail, it will be helpful to review the operation of linear accelerators of the type consisting of a series of drift tubes and gaps spaced such that a particle travels between two gaps in the time required for the alternating voltage to change one-half cycle. S-uch llinear accelerators are described, for example, in the article, The Production of Heavy High-Speed Ions Without Use of High Voltages, by D. Sloan and E. O. Lawrence, published in Physical Review, vol. 31, Dec. 1, 1931, pages 2021-2032.
Considering the usual prior art situation, a sine wave voltage is applied to the drift tube such that synchronous particles arrive at the center of each gap at a preselected phase yangle (the synchronous phase) which lies between 0 and 90 degrees. If a particle arrives too soon, e.g., at a slightly smaller phase angle, it receives a smaller voltage increase and takes a slightly longer time to arrive at the center of the next gap than does the synchronous particle. Hence this non-synchronous particle `arrives slightly later, in phase it receives too large a voltage increase -and takes less time to arrive at the center of the next gap, hence falling back in phase towards the synchronous particle. Therefore, a synchronous phase angle between 0 and 90 degrees creates a phase stable condition. It can likewise be shown that a synchronous phase angle between 90 and 180 degrees will produce a phase unstable condition in which non-synchronous particles diverge in phase.
A further problem is Vthe associated radial motion of the particles. The lines of force in a gap include a radial component of an electric :field which points inward during the passage of the particle through the rst half of the gap and which points outward during the second half of transit through the gap. If the voltage were constant in time these radial eifects would just cancel. But when the eld increases in the gap during transit of a particle i.e., when the synchronous phase angle lies between 0 and 90, the radial outward force exceeds the radial inward force and the particle is pushed away from the axis of the system. The opposite results obtained when the synchronous phase angle lies between 90 and 180. As la result the well known effect of incompatibility of phase and radial stability occurs; if the phase angle lies between 0 and 90 the system is phase stable but radially unstable, where-as if the phase lies between 90 and 180 vthe system is Iphase unstable but radially stable.
In an accelerator for nuclear particles, since a grouping of a large number of particles is desired, the accelerator is always operated in the phase-stable manner. Therefore, various techniques have ibeen developed in order to correct for radially outward motion. These techniques include focusing grids and various electric and magnetic lens systems.
In the case of micrometeor-oid simulation the specific problem is that of accelerating a single particle at a time. Therefore, there is no longer a restriction to operate the accelerator in a phase-stable condition; in ifact, it is a feature of the present invention to operate the accelerator in a phase-unstable condition. This will insure radial stability, a very desirable situation since the particles injected into the drift tubes have a small angular spread of about 10-3 radians. In addition, shifts in phase are more easily corrected than are changes in radial position.
It should be borne in mind that the phase must not be allowed to increase without bound since a particle will soon fail to be accelerated in a gap and will be lost. To prevent this, and as another feature lof the present invention, a modified square wave whose peaks have a constant non-zero slope (such as a drooping square wave Whose peaks have a constant negative slope) is provided for the high voltage pulses which are applied to the accelerating gaps. This produces a rate of change of phase which is proportional to the sl-ope of the square wave. A true square wave would be neither phase nor radially stable. However, by allowing each peak of the square wave to change slightly with time, it is possible to obtain an effect similar to that which would occur in a sine wave with a synchronous phase just over In contrast to a sine wave, however, the modied square wave has peaks with a constant non-zero slope such that for a given amount of radial focusing the phase instability is much less severe than would occur with a sine wave. This further allows the slope, and hence the rate of change of phase, to be controlled substantially independently of the frequency of the pulses applied t-o the accelerating gaps.
A further problem associated with a linear accelerator for micrometeoroid simulation (as constrasted with a nuclear accelerator) is the variation of the chargeto-mass-ratio (q/m) of the particles. For nuclear particles the q/ m of the particles is xed once the projectiles are chosen (protrons, nutrons, alpha-particles, etc.). This means that a iixed frequency may be chosen for pulses applied to the accelerating gap. In the case of microparticle acceleration, however, the q/m of the particles injected into the drift tubes is not constant and may vary over a range of one hundred to ione. Accordingly, it is another feature 'of the invention to adjust the frequency of the modified square wave pulses applied to the accelerating gaps to match the q/m of each particle that is being accelerated.
General description Referring now to a specific embodiment of the invention in FIGURE 1 positively charged micro-sized dust particles are injected into drift tubes 101-10n by Van de Graaff accelerator 12. Drift tubes 101-10,l are the drift tubes of a Wideroe-Sloan-Lawrence linear accelerator of the type described in the above cited publication, and in structural detail form no part lof the invention. It is suicient to note that a number of such drift tubes (92 in a practically constructed accelerator) are axially aligned in a suitable vacuum chamber, with the gap-to-gap spacing between drift tubes increasing with the square root of an integer representing the number of successive drift tubes to accommodate the increasing velocity of down-stream particles.
The dust particles may be micron-sized carbon or iron particles and are accelerated by a modified Van de Graaff positive ion accelerator wherein the ion source is replaced by a particle charging and injection system. Dust injection systems of this type are described in the article entitled, Two-Million-Volt Electrostatic Accelerator for Hypervelocity Research by I. F. Friichtenicht, published in the Review of Scientific Instruments, volume 3, No. 2, pages 209-212, February 1962.
Modified square wave pulses (to be subsequently described) are derived from the output of a high voltage system which applies voltage pulses of opposite polarity to leads 15 and 17. Leads 15 and 17, in turn, are respectively connected to alternate ones of drift tubes 101- n so that an alternating voltage wave providing voltage pulses of equal magnitude but opposite polarity is applied between adjacent drift tubes. This produces the desired accelerating field for each particle entering the gaps between drift tubes 101-10n.
There is further provided velocity detector 18, positioned between Van de Graaff accelerator 12 and first drift tube 101. Velocity detector 18 includes an insulated drift tube 19 mounted coaxially within grounded shield 21 having grids on either end. Details of velocity detector 18 are discussed in an article Electro-Static Acceleration of Microparticles to Hypervelocities by H. Shelton, C. C. Hendricks, Jr. and R. F. Wuerker, Journal of Applied Physics, volume 31, No. 7, July 1960, and also in the above cited article 'by J. F. Friichtenicht. Briefly, the passage of a charge particle through the detector will induce a voltage pulse having a Width equal to the time of liight through the detector. Thus for a known length of the detector the velocity of the particles may be determined. The signal thereby derived is coupled to amplifier 22 and thence to frequency generator 30. There is, accordingly, a pulse (illustratively shown by waveform 23) applied to frequency generating system 30 having a width inversely proportional to the speed of a particle entering drift tubes lOl-10D.
Frequency generating system 30 produces a periodic signal having a frequency component which is directly proportional to the initial velocity at which a particle enters drift tubes 10i-10u. This signal is c-oupled, via control system 40, the high voltage system 60. Details of frequency generating system 30, control system 40 and high voltage system 60 are to be subsequently described. It is sufiicient to note at this point that control system 40 responds to frequency generating system 30 to cause high voltage system 60 to produce the required accelerating voltage pulses. Thus driver stage 62 triggers switch tubes 64 and in conjunction with high voltage power supply 67 produces square wave pulses at a repetition rate established fby frequency generating system 30. These pulses are shaped (for example the peaks thereof provided with a constant negative slope) by wave shaping network 66 and applied to leads 15 and 17. The resulting high voltage Wave appearing on leads and 17 is balanced or symmetrical with respect to ground so that the accelerating field of successive gaps becomes negative as a positively change particle moves through successive ones of drift tubes 101-1011.
The modified square wave appearing between leads 15 and 17 and more applied to alternate ones of drift tubes 101-10n to provide an accelerating field of the desired polarity in the gaps therebetween is illustrated by FIG- URE 2. Waveform 24 (dotted) represents an ideal square wave with peaks having la constant negative slope, and waveform 26 (solid) represents an actual waveform produced as a result of the time constants of the system. A synchronous particle at the center of a selected drift tube is shown at 27, at the center of a gap 28 and at the center of the next drift tube at 29. The phase of a particle relative to the applied voltage is measured with respect to the center of the wave (=0), and E is a limiting phase which if exceeded will cause deceleration. The
limiting phase varies inversely with the rise time of the waveform, and accordingly rise time should be as small as possible. The fractional droop of the square wave a should be at least .005 at the highest frequency.
Values of significant parameters for a practically structured apparatus are given in Table I.
TABLE I Parameter Value Total number of drift tubes 92 Injection voltage mv 1.6
6 Peak applied voltage per gap kv Length of first section (drift tube gap) cm 4 Gap width to section length ratio Ms Bore radius cm 0.5 Median frequency of driving voltage kc Fractional droop at highest frequency 0.005 Equilibrium phase 0 The drift tube assembly is mounted inside a large stainless steel vacuum tank (24 feet long by 18 inches in diameter) fitted with a vacuum system capable of evacuating the tank to below l0*6 mm. Hg. The drift tube structure may be assembled in sections for ease of handling and mounted in the tank on supports which allow alignment with each other and with the Van de Graaff acceler- `ator. The overall effective accelerating potential of a linear accelerator having the above described parameters is six times the injection voltage. Thus particle velocities up to 70 km./seconds are obtainable, with the `average particle velocity of 30 km./ second. This average velocity is possible at a higher repetition rate and with significantly larger mass particles than the occasional 30 km./second particles produced by existing Van de Graaff accelerators.
Frequency generating system As noted, frequency generating system 30 produces a periodic signal having a frequency component which is directly proportional to the initial velocity at which a particle enters drift tubes 101-10n. This may be achieved by utilizing a variable frequency oscillator to generate the desired frequency and controllable Iby the signal produced by velocity detector 18. However, because of the accuracy required and the fact that .the phase of the oscillator must be precisely controlled, digital systems are preferred. One embodiment of a digital system is illustratively shown in FIGURE 3.
With particular reference to FIGURE 3 there is provided a measurement counter 31 and a frequency counter 33. Counters 31 and 33 may be conventional binary counters. Each counter 31 and 33 receives a clock signal from a fixed frequency oscillator 35. Counter 31 starts counting the clock pulses produced by oscillator 35 at the beginning of the pulse received from velocity detector 18 via amplifier 22, i.e., when a particle enters drift tube 19 of velocity detector 18. Counter 31 is stopped by lthe trailing edge of the pulse from velocity detector 18, i.e., when a particle leaves drift tube 19. At the end of this coun-t the value of the count is shifted in parallel into frequency counter 33. Frequency counter 33 then repetitively counts the clock pulses produced by oscillator 35, beginning with the value established by measurement counter 31 and finishing at zero. The time required for counter 33 to reach Zero is inversely proportional to the measured velocity of a particle passing through velocity detector 18. When frequency counter 33 reaches zero it is reset to the value of the count stored in measurement coun-ter 31 and the process repeats. When a successive particle enters drift tubes 101-1011, or when downstream velocity correction is provided, as subsequently described, a new count is entered and stored by measurement counter 31. This, in turn, causes frequency counter 33 to count to zero in a different time interval if the detected particle has a different velocity.
Each time frequency counter 33 reaches zero a pulse is produced to trigger flip-Hop 37, generating a square wave having a fundamental frequency representative of the velocity of a particle as measured by velocity detector 18. This square wave and its complement appear as outputs A and B of flip-flop 37, and are supplied to control system 40.
In addition to the above described system, a further digital system having improved accuracy with the use of a lower frequency clock signal which may be utilized for frequency generating system 30 is described in detail in copending application Ser. No. 505,321 filed Oct. 24, 1965 by Charles A. Gilkison and commonly assigned.
High voltage system High voltage system 60 functions to apply high voltage square wave pulses of the desired frequency and with peaks having the desired negative slope to drift tubes 101-10n. As previously discussed, the basic elements of such a system includes driver stages 62, high voltage switch tubes 64, wave shaping network 66 and high voltage power supply 67 A preferred embodiment of high vol-tage system is shown in detail in FIGURE 4, and incorporates circuitry to perform the functions of the kabove mentioned basic elements.
With reference now to FIGURE 4, the high voltage switch tubes collectively illustrated by reference numeral 64 include vacuum tube triodes '71 and 72, 7 3 and 74 connected in a bridge configuration in the manner shown. One pair of opposite corners of the bridge (the junction of the anodes of vacuum tubes 71 and '73 and of the cathodes of vacuum tubes 72 and 74) is connected across high voltage power supply 67. By way of example, vacuum tube triodes 71-74 may be Machlett DID-15 pulse modulation triodes. High voltage power supply 67 includes two equal high voltage sources 77 connected in series, with their common point return to ground. Each high voltage source 77 is also shunted by a storage capacitor 79. This arrangement provides a balanced or symmetrical high voltage power supply; that is, one providing equal positive and negative voltages with respect to ground reference potential.
The other pair of the opposite corners of the bridge (the junction of the anode and cathode of vacuum tubes 71, 74 and 72, 73, respectively), is connected to leads and 17 and thence to drift tubes lill-10m. The equivalent distributed capacitance of drift tubes 101-10n is representatively shown by capacitor 80. Resistors 82 and 84 return leads 15 and 17, respectively, to ground reference potential. The RC time constant provided by equivalent capacitor 8@ and resistors 82 and 84 imparts a constant negative slope or droop to the peaks of the square wave. The equivalent capacitor 811 and resistors 82 and 84 function as wave shaping network 66. Since the amount of the droop is fixed in time, the percentage droop which determines the focusing properties of the accelerator varies inversely with frequency. However, it has been found that by utilizing values for resistors 82 and 84 which give the minimum tolera-ble droop at the highest frequency, the droop at the lower frequencies is not excessive.
Grid drive for switch tubes 71-74 is obtained by seeondary windings 87a, 87b and 89a, 8919 of pulse transformers 86 and 88, respectively. These secondary windings are phased so that opposite pairs of switch tubes 71-74 are rendered conductive in the manner subsequently described. The primary of transformer 86 is driven by a pair of pulse drive tetrodes 93a and 93b. The primary of pulse transformer S8 is similarly driven by a pair of pulse driver tetrodes 94a and 94b. Grid drive for tetrodes 93a, 93b and 94a, 94h is obtained from control system 40 as will be subsequently described. To apply a modified square wave to drift tubes 1111-10,rl high voltage switch tubes 71 and 72 are momentarily pulsed on. This charges drift tubes 10140 (represented by capacitor 80) with a voltage of a given polarity from high voltage power supply 67. Turning switch tubes 71 and 72 off leaves drift tubes 101--10n in this state. The voltage applied to drift tubes lof-10n is reversed by momentarily pulsing switch tubes 73 and 74 on. Similarly, turning switch tubes 73 and 74 off leaves the drift tubes in this new state. Thus, by applying appropriate driving pulses to switch tubes 71`74 a modified square wave voltage, shaped such that its peaks have a constant negative slope, is applied to drift tubes 101-10n- By using the above described symmetrical configuration insulation problems are considerably reduced, since the Voltage from any point to ground is onehalf the total voltage across the gaps. In addition, the presence of resistors 82 and 84 insures that after the acceleration cycle has been completed the drifty tubes Will return to ground potential.
Control system Interposed between frequency generating system 30 and high voltage system 60 is control system 40, the details of which may be seen in FIGURE 1. Outputs A and B of flip-flop 37 of frequency generating system 30 are fed to differentiating amplifiers 41 and 4Z to provide pulse trains (shown as waveforms 41aV and 42h respectively) in response to the leading edges of the complementary square Waves derived from flip-flop 37. These pulse trains have a phase relationship with respect to one another. One pulse train provides on trigger pulses for iiip-iiops 43a and 43b and the other pulse train provides on trigger pulses for fiip-flops 44a and 44h.
Flip-Hops 43a, 43h and 44a, 44h receive off trigger pulses from pulse height discriminators 45a, 45h and 46a, 46b, adapted to select and pass voltage pulses exceeding a specified minimum amplitude. These pulse height discriminators, in turn, receive a voltage pulse having a magnitude proportion to the accelerating voltage wave appearing on leads 15 and 1'7 and applied to drift tubes 101e10n from capacitive probes 47a, 47b and 48a, 48h. As shown, pulse height discriminators 45a and 45b (supplying off pulses to flip-ops 43a and 43b) are respectively coupled to leads 15 and 17 by probes 47a and 47b, and pulse height discriminators 46a and 46h (supplying off pulses to flip-fiops 44a and 44h) are respectively coupled to leads 15 and 17 by probes 48a and 48b. The triggering level of pulse height discriminators 45a and 45b and 46a and 46h are set to turn respective ones of flipops 43a, 43b and 44a, 44h off when the accelerating voltage applied to drift tubes 10i-10n exceeds predetermined level. This level may be varied with respect to ground reference potential, for example, by potentiometer 49. In addition, pulse height discriminators 45a and 45b are adapted to operate on input voltages of opposite polarity wtih respect to one another, as are pulse height discriminators 46a and 46h. Thus, when the accelerating voltage applied to drift tubes lOl-10u exceed a specied value of one polarity pulse height discriminators 45a and 45h turn flip-flops 43a and 43h off, and when the accelerating voltage is of the opposite polarity pulse height discriminators 46a and 46h turn flip-flops 44a and 44b off.
The outputs of iiip-fiops 43a and 431) are coupled, via amplifiers 53a and 53h,V to the control grids of driver tetrodes 93a and 93b of high voltage 6i) to thereby control switch tubes 71 and 72. Similarly, outputs of flipflops 44a and 44b are coupled, via amplifiers 54a and 54b, to the control grids of driver tetrodes 94a and 94b of high voltage 60 to thereby control switch tubes 73 and 74.
In operation, When frequency generating system 30 starts, as a result of the passage `of a particle through velocity detector 18, the pulse train derived from output A of flip-flop 37 triggers flip-flops 43a and 43b on. This causes conduction of driver tetrodes 93a and 93b and hence switch tubes 71 and 72, and a voltage pulse of a given polarity is applied between leads 15 and 17. When the voltage for eac-h set of drift tubes exceeds a specified level with respect to ground reference potential (for example, 50 kv.), hip-flops 43a and 43b are turned off by pulse height discriminators 45a and 45h, thereby ending the pulse applied to the drift tubes by switch tubes 71 and 72. Subsequently the second pulse train derived from output B of fiip-flop 37 triggers flip-flops 44a and 44b on to similarly cause conduction of switch tubes 73 and 74, thereby applying a voltage pulse of the opposite polarity between leads 15 and 17. In a like manner, this pulse is terminated when the voltage applied to the drift tubes exceeds the specified level. This action continues as long as signals are generated by frequency generating system 30, with the repetition rate of the voltage pulses applied to drift tubes 101-10n controlled by the velocity of a particle to insure that the particle is synchronized with the accelerating potential across the gaps between drift tubes. When a subsequent particle of a different velocity enters drift tubes 101-10n, or when downstream velocity correction is employed, the repetition rate of the voltage pulses is altered correspondingly so that synchronous phase is maintained.
It is to be noted that should pulse height discriminators 45a, 45b and 46a, 46b fail to turn off their aS- sociated flip-flops, one pair of switch tubes 71-74 would remain on as the other pair is turned on. To prevent this possibility the output of each fiip-ilop 43a, 43b and 45a, 451; is coupled to one of monostable multivibrator 55a-55b, and the output of these monostable multivibrators is returned to the input its associated flip-flop. Thus, each monostable multivibrator is triggered on by an output of one of flip-flops 43a and 43b and 44a and 44h, and is adapted to turn the same ip-ilop olf after a present timing period. In the event that the related pulse height discriminator fails to turn the flip-flop ofI, such flip-flop is automatically turned oi by one of monostable multivibrators 55u-55h. This insures that Ione pair of the high voltage switch tubes 71-74 is off before the next pair is turned on.
Downstream velocity correction Although for the linear accelerator specifically described proper frequency control is obtainable by using a single velocity detector 18 for particles entering drift tubes 101-1011, downstream velocity correction is desirable where a larger number of drift tubes are used. One manner in which this downstream velocity correction may be achieved is illustrated by FIGURE 5, wherein like reference numerals denote like elements of the basic accelerator of FIGURE l.
Additional velocity detectors 18a and 18b, each similar in construction and mode of operation to velocity detector 18, are disposed at periodic intervals along drift tubes 191-10n. The output of each additional velocity detectors is coupled, via amplifiers 22a and 22b, to measurement counter 31 of frequency generating system 30. Passage of a particle through velocityY detector 18 initiates an output from frequency generating system 30 as previously discussed. As the particle subsequently passes through velocity detector 18a a new count of clock pulses from oscillator 35 is stored by measurement counter 31 and shifted into frequency counter 33. This new count may be the same or different from the previous count. If different, this new count provides a variation of the frequency of the square waves generated by flip-flop 37 in order to provide the necessary frequency correction. In a like manner further frequency correction is provided as the particle passes through velocity detector 18h, and any subsequent velocity detectors which may be utilized.
The invention provides, therefore, a linear particle accelerator capable of accelerating microparticles to greater velocities than heretofore obtainable. The accelerator is particularly adaptable to accelerate single particles, is provided with radial focusing and minimum phase instability by application of a modied square voltage with peaks having a constant non-zero slope to the drift tubes thereof, and automatically compensates for variations of initial particle velocities and charge-to-mass ratios to insure that a maximum number of particles are successively accelerated.
Although a specic embodiment of the particle accelerator of the invention has been described with particularity, it is not limited to the specific system and particular circuit arrangements herein disclosed, and modications and variation thereof should be obvious to those skilled in the art. It is therefore to be understood that Within the scope of the appended claims the invention may be practiced otherwise than specifically set forth.
What is claimed as new and desired to be secured by Letters Patent of the United States is:
1. Particle acceleration apparatus comprising:
a series of longitudinally spaced, axially aligned hollow means for injecting charged microparticles into said drift tubes;
means for applying a periodic accelerating voltage wave to said drift tubes having a phase and frequency such that each particle passing through said drift tubes receives an accelerating electric field in the gaps between said drift tubes;
said periodic accelerating voltage wave being a moditied square wave whose peaks have a constant nonzero slope. 2. Particle accelerating apparatus comprising: a series of longitudinally spaced, axially aligned drift tubes disposed within a vacuum chamber;
electrostatic generator means for injecting charged microparticles into said drift tubes at an initial velocity;
high voltage supply means producing a substantially square wave accelerating voltage; and
wave shaping means for modifying and applying said square wave accelerating voltage to alternate ones of said drift tubes with opposite polarity,
said wave shaping means operable to impart a constant non-zero slope to the peaks of said square wave accelerating voltage, thereby causing said particle accelerator to operate radially stable, with phase instability being minimized by the constant non-zero of the peaks of said square wave accelerating voltage.
3. Particle acceleration apparatus comprising:
a series of longitudinally spaced, axially aligned hollow drift tubes;
means for injecting charged microparticles into said drift tubes at an initial velocity;
means for applying a periodic accelerating voltage to said drift tubes, said periodic accelerating voltage being substantially a modied square wave whose peaks have a constant non-zero slope; and means responsive to the velocity of a particle passing through said drift tubes to control the repetition rate of said square wave such that each particle experiences an accelerating electric field in the gaps between said drift tubes. 4. Particle accelerating apparatus comprising: a series of longitudinally spaced, axially aligned hollow drift tubes disposed within a vacuum chamber;
electrostatic generator means for injecting charged micropaiticles into said drift tubes at an initial velocity;
high voltage supply means producing a square wave accelerating voltage;
wave shaping means for modifying and applying said square wave accelerating voltage to alternate ones of said drift tubes with an opposite polarity, said wave shaping means operable to impart a constant non-zero slope to the peaks of said square wave accelerating voltage; and
circuit means including detection means responsive to the velocity of a particle passing through said drift tubes to control the repetition rate of said square Wave accelerating voltage such that each particle experiences an accelerating electric field in the gaps between said drift tubes.
5. The particle acceleration apparatus of claim 4 in which said detection means is disposed between said electrostatic generator and said drift tubes and produces a voltage pulse of width inversely proportional to the initial velocity of each said particle, and wherein said circuit means further includes:
frequency generating means responsive to said voltage pulse for providing periodic signals having a frel 1 quency component proportional to the initial velocity of each sa-id particle; and controls means coupled between said frequency generating means and said high voltage supply means and responsive to said periodic signals to correspondingly vary the repetition rate of said square wave accelerating voltage. 6. The particle acceleration apparatus of claim 5 wherein said periodic signals are complementary square waves, and wherein said control means includes;
rst and second pairs of bistable circuit means, each pair of -bistable circuit means coupled to said frequency generating means and triggered to a rst state by one of said complementary square Waves;
means responsive to the magnitude of the square wave accelerating voltage applied to said drift tubes coupled to trigger said rst and second pairs of bistable circuit means to a second state; and circuit means coupling signals indicative of the state of said first and second pairs of bistable circuit means to said high voltage supply means to control the phase and frequency of said square wave accelerating voltage. 7. The particle acceleration apparatus `of Iclaim 6 wherein said means responsive to the magnitude of said square wave accelerating voltage applied to said drift tubes includes:
capacitive probe means `coupled with `said drift tubes to induce a voltage proportional to the magnitude of said square wave accelerating voltage; and
voltage level discriminator means operable to supply trigger pulses to said lirst and second pairs of lbistable circuit means as said square wave accelerating voltage exceeds a predetermined level to thereby trigger said first andV second pairs of bistable circuit means to said second state.
8. The Iparticle acceleration apparatus of claim 7 further including means to automatically trigger said rst and second pairs of bista'ble circuit means to said second state after a preset timing interval.
9. The particle acceleration apperatus of claim 4 wherein said high voltage supply means includes electronic switching devices in circuit between a D.C. voltage source and said drift tubes, and driving means responsive to said control circuit means to switch said electronic switching devices toV thereby apply high voltage pulses of opposite polarity to alternate ones of said drift tubes.
10. The particle acceleration apparatus of claim 6 wherein said high voltage supply means includes first and second pairs of ele-ctronic switching devices connected in series between alternate ones of said drift tubes and terminals of lopposite polarity of said D.C. voltage source, and driving means responsive to said control circuit means such that said rst pair of electronic switching devices is switched on as saidrst pair of bistable circuit means is triggered to said rst state and switched off as said first pair of bistable Icircuit means is triggered to said second state, and said second pair of electronic switching devices is switched on as said second pair of bistable -circuit means is triggered to said first state and switched o as said second pair of bistable circuit means is triggered to said second state.
,11. The particle acceleration apparatus of claim 4 wherein said wave shaping lmeans is comprised of a resistance-capacitance network including the effective distributed capacitance -of said drift tubes, operable to impart a constant negative slope to the peaks of said square wave accelerating voltage.
12. The particle acceleration apparatus of claim 5 wherein a plurality of further detection means are positioned at periodic intervals along the longitudinal -aXis of said series of drift tubes, each said detection means providing a voltage pulse of a width inversely proportional to the velocity of each Iparticle a-t the corresponding point along the longitudinal axis of said series of drift ltubes, each said `further detection means coupled to said frequency generating means to thereby vary the repetition rate of said square wave accelerating voltage in accord'- ance with particle velocity at said corresponding point along the longitudinal axis of said series of drift tubes.
References Cited by the Examiner UNITED STATES PATENTS 2,683,216 7/1954 Wideroe 328-233 2,970,273 1/1961 Bennett 313-63 JAMES W. LAWRENCE, Primary Examiner. S. D. SCHLOSSER, Assistant Examiner.
UNITED STATE SI PATENT oEFlCE CERTIFICATE l0E CORRECTION Patent No. 3,317,846 May 2, 1967 Hugh L. Dryden, Deputy Administrator of the National Aeronautics and Space Admini It is herebj)T certified that error appears in the abov ent requiring correction and that the said Letters Patent should read as stration e numbered patcorrected below.
Column l0, line 4, after "hollow" insert drift tubes; line 6, after "drift tubes;" insert and Signed and sealed this 6th day of August l968 (SEAL) Attest:
EDWARD I. BRENNER Edward M. Fletcher, ,I r.
Commissioner of Patents Attesting Officer