US2230473A - Means for increasing the compactness of high voltage electrostatic apparatus - Google Patents

Description

1941- R. J. VAN DE GRAAFF MEANS FOR INCREASING THE COMPACTNESS OF HIGH VOLTAGE ELECTROSTATIC APPARATUS Filed June 22, 1939 InvenZo 2 Roircri 1 VandeGz'aqjfi M W M rluddnw Patented Feb. 4, 1941 UNITED STATES MEANS 'FOR IN CBEASING THE OOIPACT- NESS OF HIGH VOLTAGE EIECIBOSTATIC APPARATUS Robert J. Van de Graafl, Belmont, Ila-5., auignor to Research Corporation, New York, N. Y., a corporation 01' New York Application June 22, 1939, Serial No. 280,556

8 Claims.

This invention relates to high potential electrostatic apparatus embodying the general principles disclosed in prior Patent No. 1,991,236. Electrostatic apparatus of that type has come into 8 increasing use for the generation of high potential direct current. Generators of this type generally comprise an electrode or high voltage terminal mounted on an insulating support or supports which constitute in effect an insulating column between the terminal and ground, charges being transferred from ground to the terminal by a movable charge conveying medium, usually in the form of an endless belt.

The need in such applications for voltages of the order of several millions has lead to the utilization of gases under high pressure in place of atmospheric air as the medium surrounding the component parts of such generators. The use of such increased gas pressure increases the voltage which can be supported across a given electrode gap, or, for the same voltage, reduces the gap spacing required. The eifect is therefore to increase the compactness of high voltage sources of this type and consequently to reduce their cost,

their space requirements, and to secure other attendant advantages.

Even with the use of gases under high pressure, however, voltages needed for nuclear research, which may be of the order of several megavolts, have required the construction of physically large generators. For example, generators have been constructed enclosed in pressure tanks, the latter measuring approximately 50 feet in height and 30 feet in diameter.

In general, the high voltage terminal of this type of electrostatic generator consists of a metal dome-like shell, usually hemispherical in form or having a rounded or spherical surface and supported on the insulating column. The pressure tank enclosing the generator usually has a rounded upper end, the wall of which is spaced from the terminal and may be concentric with the surface of the terminal.

For a given diameter of the tank there is an optimum diameter for the high voltage terminal at which a maximum voltage can be supported across the intervening gap. The use of either a larger or smaller diameter for the high voltage terminal would reduce the maximum voltage which could be supported. This is due to the fact that breakdowns ensue when a certain maximum gradient characteristic of the gas and pressure is exceeded at any point in the inter-electrode gap. The effect of reducing the diameter of the hemispherical terminal is to increase the spark-over distance, but, at the same time, it increases the relative electric gradient at the terminal surface due to the decreased radius of curvature and consequent increase in field distortion. 0n the other hand, the eifect of in- (CI. TIL-329) creasing the diameter over the optimum value is to increase the radius of curvature, but, at the same time, it reduces the intervening gap.

For maximum voltage strength the optimum relationship between tank diameter and terminal 5 diameter can be shown to be exactly as 2 is to 1, assuming the iield distribution to he that of concentric spheres and assuming a certain limiting potential gradient to be allowable for the gas insulation. With this optimum ratio, it is evident that the gradient at the surface of the terminal is exactly four times the gradient at the surface of the tank. When higher voltages are desired with this arrangement in a gas at a given pressure, it is not suflicient merely to increase the 15 tank diameter in order to secure a larger electrode separation. If the most eifective use of the insulation is made with such an arranganmt, it is necessary to increase both tank and terminal diameters so as to preserve the optimum ratio. 29 Due to these conditions, machines of many megavolts have been designed in which the terminal diameter is of the order of 10 to 15 feet or more,-

a dimension far in excess of the space required inside the terminal, but required in order to eb- 25 tain the most effective electrode geometry from the point of view of maximum voltage insulation strength. This method of insulating very high voltages in pressure tanks does not use the insulating properties of the compresed gas to the so fullest extent.

One object of applicant's invention is to avoid such limitation and its attendantdisadvaniages and to increase the voltage which may be insulated under given conditions by the establish- 5 ment of a substantially or approximately uniform potential gradient in the region between the walls of a substantially uniform field in such regions surrounding the carrier and the insulating support for the terminal, and by the establishment an of a substantially imiform field in such regions irrespective of the ratio of tank and dimensions.

One means for accomplishing this, which is herein shown in the illustrative embodiment of .35 the invention, is through the provision of a number of thin shields presenting each a continuous metallic conductive surface surrounding the terminal, such shields being spaced one another in the gap between the terminal and the so walls of the tank. In case the dome of the tank and the terminal are in hemispherical form, the upper portions of these shields are preferably concentric with the terminal.

In general, solid insulation has a lower flash- 55 over strength than that of a corresponding gap in the gas, and the allowable gradient along such insulation is much less than directly across the gas. It is therefore necessary to provide an insulating support of relatively great length beso tween the terminal and the ground. The region along and within such support and along and about the charge-carrying belt is accordingly particularly susceptible to flash-over. It therefore becomes quite important to control potential distribution and to minimize field distortion in this region as well as in the region adjacent the terminal.

Accordingly, there is provision for maintaining a substantially uniform potential gradient between the terminal and ground along the insulating support on which the terminal is mounted. For this purpose, the shields surrounding the terminal comprise also cylindrical downward extensions which are connected to and supported on the column at appropriate and succeeding points so that the proper intended potential is placed throughout on each intermediate shield. Means are aLso preferably provided so that the field due to carrier charge is distributed along the length of the carrier through controlled potentials effectively applied to a succession of short sectional lengths of the carrier.

Such shields then act in each case to define a substantially equi-potential area in the space surrounding the terminal and the insulating support and around and across the carrier which minimim the distortion of the electric field, leaving the gradient between the tank and the terminal more uniform.-

By this means the wide disparity in the ratio of the gradient at the outer surface of the inner electrode and the inner surface of the tank can be reduced, and, moreover, the requirement of placing an optimum ratio between the terminal and tank diameters can be disregarded. The use of such intermediate shields in a relatively compact tank or other enclosure thus permits high voltage terminals of small diameter, large enough merely to house the various component parts of the apparatus,--a consideration of particular importance when the apparatus is used for the acceleration of electrons, sources of which are relatively simple and compact.

Ideally considered, the high voltage terminal, if a sumcient number of intermediate shields were used, could be made vanishingly small so that the full radius of the spherical tank would be available for the insulation of the voltage. With an infinite number of shields of negligible thickness, all at controlled potential, the gradient from terminal to tank could be substantially uniform. Under such an ideal supposition four times the voltage could be insulated over that when there is employed an open or single gap arrangement with the optimum diameter ratio of two. Due to practical considerations such an ideal condition is obviously impracticable. The high voltage terminal cannot thus be made vanishingly small, nor can the number of intermediate shields be excessively large. The following comparative illustration, however, shows the nature of the gain in insulated voltage for a given tank diameter resulting from applicant's method.

Let there be considered a tank 40 inches in diameter, with a hemispherical dome and a hemispherical high voltage terminal, the tank containing gas at such a pressure that the maximum gradient which can be supported is 400,000 volts per inch. In prior constructions, with an uninten-opted gap between the tank dome and high voltage terminal, the optimum diameter of the latter would be inches, the ratio of the gradient of the surface of the terminal would be as 4 is to 1, and the maximum voltage which could be insulated would be 2,000 kilovolts.

Let it be assumed that a single intermediate shield at controlled potential is provided of such diameter that the gradient at the outer surface of the shield should be the same as the gradient at the outer surface of the terminal. The maximum voltage which could be supported on the terminal would then be 2,830 kilovolts, and the ratio of the maximum and minimum gradients in this double gap arrangement would be as 2 is to 1. The intermediate shield under these conditions would be at a potential of 1,665 kilovolts above ground.

Let it be now assumed that two spaced intermediate shields are placed between the electrodes in the same way. The maximum voltage which could be insulated on the terminal would now be 3,180 kilovolts, and the ratio of the gradients in the electrode system would be reduced to 1.59 to 1. The inner shield under these conditions would be maintained at 2,355 kilovolts. above ground, and the second shield at 1,315 kilovolts above ground.

With three intermediate shields, the maximum voltage becomes 3,370 kilovolts, and the ratio of the gradients reduces to 1.41 to 1. With a very large number of shields of negligible thickness, the voltage would be 4,000 kilovolts, and the ratio of gradients would be unity.

It is evident that relatively few shields, provided as suggested, result in a large increase in the voltage which can be insulated. The above example does not illustrate, however, the additional advantage that accrues from a construction permitting a terminal diameter of less than the above stated optimum value. Should the space requirement permit a diameter of the high voltage terminal smaller than one-half the tank diameter, as may be the case particularly in very high voltage generators, it is evident that the shield construction can start from any desired terminal diameter and that a larger proportion of the radial distance can thereby be devoted to insulation of voltage.

Besides the advantage of minimizing field distortion through the use of shields as described, a further advantage arises from the fact that in general the breakdown strength across a large gap, even when the ileid is uniform, is increased by dividing the gap into a number of smaller gaps by the use of shields at appropriate potentials, since the aggregate of the break-down strength of a number of small gaps is apt to be greater than that of the previous single gap of the same aggregate length.

The invention will be better understood by reference to the following description when taken in connection with the accompanying illustrations of one specific embodiment thereof, while its scope will be more particularly pointed out in the appended claims.

In the drawing:.

Fig. 1 is an elevation, in partial section, of a gas pressure-insulating electrostatic generator embodying one form of the invention;

Fig. 2 is a section in plan on the line 22 in Fig. 1; and

Fig. 3 is a diagram representing the gradient distribution along the axis of the generator tank where two intermediate shields are employed, as in the generator of Fig. 1.

Referring to the drawing and to the illustrative embodiment of the invention there shown,

'the electrostatic generator comprises the high potential electrode or terminal ll consisting of a hollow and herein approximately hemispherical shell of conductive material, such as brass, this resting on a ring I3 of conductive metal, which in turn is mounted on the top of three spaced pillars 15 of insulating material, providing the effect of an insulating column of a generally tripod shape.

The bottoms of the pillars rest on a base plate I! of conductive metal, supported by the walls of the steel tank l9, which latter provides a chamber filled with gas under pressure. This, for example, may be air maintained at a pressure such as from to 200 pounds per square inch above atmospheric. The tank above the base plate I! is herein shown as cylindrical in form, terminating at the top in a hemispherical dome, the walls of which are concentric with those of the terminal ll, there being provided a substantial clearance between the walls of the tank and the terminal and insulating column.

The charge carrier herein is in the form of an endless belt 21 of insulating material, such as rubber fabric. Adjacent the base plate, this belt passes over a metallic grounded pulley 23, driven by a suitable motor (not shown), and extends vertically upward and downward in parallel runs, passing over a second metallic pulley 25 within but insulated from the body of the terminal ll. At its lower end charges of either a positive or negative sign, as may be desired, are established on the belt 2| by any suitable means, such as the usual brush electrode or comb-like series of points 21, connected to a charge supply 29 which may comprise a source of alternating current, a transformer and a rectifier. Such charges are transferred to the terminal, and charges of the opposite sign are transferred from the terminal to the descending run of the belt by any suitable means, herein conventionally represented by the brush electrode 3| connected to the pulley 25 and operatively related to the belt, and a second brush electrode 33 operatively related to the belt at the top of the pulley and connected to the electrode.

In the illustrative form of generator, an X-ray tube 35 is conventionally indicated, mounted vertically on the base plate H with its upper end extending into the hollow terminal II, which latter serves as the source of high potential for the tube. A grounded metallic continuation 31 of the tube extends through the bottom of the tank I9 and terminates in an X-ray target.

Means are employed to establish and maintain a uniform potential gradient down the insulating column between the terminal H and the ground plane which is established at or about the base plate ll. Such means herein consist of a. series of closely spaced but separated rings 39, each of conductive metal and preferably tubular in form, encircling the three pillars l5 and supported in their assigned positions thereby, so as to provide a column-like series of successive rings extending from the upper end of the insulating support to the bottom thereof, the complete showing of all the rings being omitted in part in the drawing to avoid confusion.

To control the field due to belt charge, means are provided which are described more in detail in respect to function and construction in the co-pending application of John G. Trump, Serial No. 258,268, filed February 24, 1939, but which comprise the following:

A vertical series of conductors 4|, each tubular in form, is provided closely spaced from and extending transversely across the outside face of the ascending run of the belt and a similar series on the outside face of the descending run of the belt. These series of tubular conductors preferably extend over at least the same vertical range or distance as the rings 39 and conform in number and spacing to the rings, each conductor 4! being prolonged in length to have mechanical and electrical connection to one of the rings.

Such field control means also comprises two additional vertical series of similar conductors 43 positioned in the same transverse plane as the conductors 4| but located one series closely spaced from the inner face of the ascending run of the belt and the other from the inner face of the descending run thereof, these conductors also extending across the full width of the belt and being prolonged to have electrical and mechanical connection to the rin 39 lying in the same plane.

The provision of the series of spaced conductors assists in maintaining a controlled potential gradient lengthwise and adjacent the belt from the electrode to ground by the establishment of equi-potential surfaces, much as the rings serve to maintain such gradient lengthwise and adjacent the insulating column. The establishment of such a controlled potential gradient is further assisted by the employment of flat, equi-potential surfaces in the form of thin, fiat, conductive plates. A series of such plates 45 is indicated (Figs. 1 and 2) as connected each between a ring 39 and the corresponding tube 4| on the ascending side of the belt and a similar series 41 is shown extending from each ring to the corresponding tube 4| on the descending side of the belt. These are shown as covering substantially the entire area lying in any one plane between the conducting tube and the ring left uncovered by the insulating pillars, X-ray tube or other in terfering parts. The insulating pillars may consist of a built-up series of approximately cylindrical but preferably corrugated (as partly indicated in Fig. 1) sections separated one from another by the thin conducting plates. A similar series 01' plate 49 is provided to extend between opposite conductors 43 of each pair, and each such plate 49 extends to and is connected at opposite ends to the inner periphery of the corresponding ring 39.

A uniform potential gradient throughout the series is assured by a slight leakage flow of current from each transverse conductive assembly to the adjoining underlying one, as by the resistances 5| conventionally indicated in Fig. 1 between the plates 45.

Surrounding the terminal I I there are provided a number (herein two) of intermediate shields 53 and 55 of the character and construction hereinbefore referred to. These comprise each a sub stantially hemispherical portion concentric with the terminal and having each a cylindrical portion extending downwardly between the insulating column and the walls of the tank, each cylindrical extension having its lower end connected to and supported by one of the conductive rings 39.

The spacing between the shields is such as to provide three successive gaps and to create a substantially uniform gradient between the tank and terminal, in accordance with the principles heretofore set forth and as indicated by the gradient diagram shown in Fig. 3. The cylindrical extensions are each connected to the series of rings 39 at appropriate points, that is to say, at those points along the column at which a corresponding potential is established. The inner shield accordingly is shown connected to such a ring at a point a short distance down the column and the second shield at a point further down the column and herein somewhat more than one-half the distance from the electrode to the ground plane. The result is that the entire volume around the electrode, column and belt and between thoseparts andthe walls of the tank is separated or sectionalized by equi-potential areas or surfaces, 0 thus minimizing the distortion of the electric field in the entire region between the terminal and the tank.

Fig. 3 shows the potential gradient at various points along the axis of the tank, between the high voltage terminal and the tank. The full line ordinates represent the gradient at the particular shell to which the ordinate is tangent. Thediagram also shows the gradient at points between the shells. For example, the ordinate Y shown in dotted lines indicates the magnitude of the potential gradient at a distance X above the center of the high voltage terminal.

For simplicity, the shields have been indicated in the drawing as having their centers coinciding with the center of the spherical terminal, Inv

practice it would be desirable, in general, to locate the center of the hemispherical portion of the first shell or shield somewhat above the center of the terminal, the center of the second shell somewhat above that of the first shell, and so on.

This has the efiect of increasing the length of ductive rings encircling said pillars arranged in:

successive transverse planes, a movable charge conveyer for conveying charges between the tero minal and ground, means to effect a substantially uniform distribution of the electric .field along the length of the conveyer, a tank enclosing said apparatus, and a plurality of spaced shields intermediate the terminal and the tank and between the column of conductive rings and the side walls of the tank, said shields being each connected to one of said rings. r

2. An electrostatic apparatus, comprising a high voltage terminal, insulating supporting means, a movable charge-carrying mediulnfor conveying charges between the terminal and,

ground, a tank enclosing said apparatus, resistances for establishing a substantially potential gradient along the insulating supporting means, and means for rendering field distortion between the terminal and tank at a minimum, including the region adjacent to and within the insulating supporting means and adjacent to and along the charge-carrying medium,

'0 said means comprising intermediate shields, said shields being each electrically connected to a point on the insulating support having a substantially corresponding potential.

3. An electrostatic apparatus, comprising a high voltage terminal, insulating supporting means, a movable charge-carrying medium for conveying charges between the terminal and ground, a tank enclosing said apparatus, and

means for rendering the field distortion at a minimum between the tank and the terminal, including the region surrounding the chargecarrying medium, said means comprising intermediate conductive shields surrounding the terminal and means including resistances for maintaining said shields at a controlled potential.

4. An electrostatic apparatus, comprising a high voltage terminal, insulating supporting means, a movable charge-carrying medium for conveying charges between the terminal and ground, a tank enclosing said apparatus, and means for producing a substantially uniform field in and around the terminal, supporting means and carrying medium irrespective of the ratio of tank and terminal diameters, comprising spaced shields within the tank and surrounding said terminal, insulating support and chargeca rying member. V

5. An electrostatic apparatus, comprising a high voltage terminal, an insulating column, a

chargea y g belt for conveying charges bctween the terminal and'ground, a tank containing gas at a high pressure enclosing said apparatus, and means within the tank'and surrounds ing said terminal, insulating column and belt for producing a substantially uniform gradient be' tween the tank and the enclosed parts of the apparatus. 7

6. An electrostatic apparatus, a

high voltage terminal, insulating supporting means, a movable charge-carrying medium for* conveyin charges between the terminal and ground, a tank containing gas at a high pressure enclosing said apparatus, and means within the y tank and surrounding'said high voltage terminal,

supporting means and carrier for establishing a substantially uniform field throughout the space.

immediately around the terminal, supporting means andcarrier.

7. An electrostatic apparatus comprising a high voltage terminal, insulating supporting means} a movable charge-carrying medium, means containing a medium of higher dielectric strength than that ofthe atmospheric air and enclosing said terminal, said supporting means and said conveying medium, means for establishing a substantially uniform" potential gradient lengthwise the supporting. means and the I carrying medium, and means for establishing qui-potential surfaces in the space surrounding I I said terminal, insulating support and conveyer.

8. An electrostatic apparatus comprising a high voltage terminal, insulating supporting 'means, a movable charge-carrying medium,

means containing a medium of higher dielectric strength than that of the atmospheric air and enclosing said terminal, said supporting means and said conveying medium, and means for establishing substantially equi-potential areas in the space surrounding said terminal, insulating support and conveyer.

ROBERT J. VAN as GRAAFF.

Images