Split section high voltage tube

Jan. 27, 1959 j. lempert 2,871,402

SPLIT SECTION HIGH VOLTAGE TUBE

Filed Sept. 20, 1954 2 Sheets-Sheet 1

<img file="US2871402A_D0001.tif" />

<img file="US2871402A_D0002.tif" />

<img file="US2871402A_D0003.tif" />

<img file="US2871402A_D0004.tif" />

WITNESSES'.

<img file="US2871402A_D0005.tif" />

<img file="US2871402A_D0006.tif" />

<img file="US2871402A_D0007.tif" />

Jan. 27, 1959

J. LEMPERT

2,871,402

SPLIT SECTION HIGH VOLTAGE TUBE

Filed Sept. 20, 1954

Sheets-Sheet 2

Fig.5.

<img file="US2871402A_D0008.tif" />

United States Patent Office _PitOTted &lt;ίΓ SIXΔ1 t? eel 33 e &amp; a j Awvt*

2,871,402

SPLIT SECTION HIGH VOLTAGE TUBE

Joseph Lempert, West Elmira, N. ¥., assignor to Westinghouse Electric Corporation, East Pittsburgh, Pa., a corporation of Pennsylvania

Application September 20, 1954, Serial No. 456,940

Claims. (Cl. 315—14)

My invention relates to high voltage electron tubes and in particular relates to structures and arrangements for preventing arc-over and other unwanted discharges between the electrodes of very high voltage electron tubes, which X-ray tubes may be taken as one example.

As is well known, the space inside high vacuum tubes is capable of withstanding much higher voltage gradients than is ordinary air so that the distance between electrodes of opposite polarity may be much smaller than air spacings. However, when it is desired to design tubes in which the voltages between electrodes of opposite polarity is of the order of five to six hundred kilovolts, difficulties arise from flashovers even when the voltage gradient per inch is kept down by wide spacing of the electrodes to values which tubes of lower overall voltage are able to withstand successfully. Avoidance of any surface areas of small radius of curvature on the electrodes and sectionalizing of the insulating portions of the tube walls have been resorted to as expedients for minimizing the probabilities of breakdown of the insulating properties of the interelectrode space but without attaining a completely satisfactory solution of the problem. I here present other expedients which make it possible to further minimize the probability of arc-over between electrodes operating at extremely high potential differences.

Broadly speaking the object of my present invention is to provide improved structures for high voltage electrical devices enclosed within vacuum-tight containers.

One more specific object is to provide arrangements for minimizing the probability of electrical flashovers in high vacuum electrical devices containing parts operating with high potential differences.

Another more specific object is to make possible the operation of electrical apparatus with higher potential differences between parts included within the same enclosure than was possible with arrangements of the prior art.

Still another more specific object is to provide an improved type of X-ray tube.

Other objects of my invention will become evident through reading the following description taken in connection with the drawings, in which:

Figure 1 is a diagrammatic showing in longitudinal section of an improved structure for a high voltage Xray tube embodying one form of my invention;

Fig. 2 is a similar showing of an improved form of high voltage X-ray tube embodying another form of my invention;

Fig. 3 is a similar showing of the mid-portion of another form of X-ray tube in which the principles of the same invention are applied in a somewhat different way;

Fig. 4 is a similar showing of a high voltage X-ray tube which embodies the principles of my invention in still a different form;

Fig. 5 is a showing like that in Fig 3. of another form of X-ray tube embodying still another modification of my invention;

Fig. 6 is a schematic showing of certain shielding mem- bers of Fig. 5 taken in a direction normal to the tube axis;

Fig. 7 is a showing of a modified form of said shielding members;

Fig. 8 is a view similar to Fig. 7 of a still different form of said shielding members; and

Fig. 9 is a view similar to Fig. 5 of another modification of my invention.

Referring in detail to Fig. 1, an X-ray tube intended to generate X-rays of several hundred electron-volts energy has a vacuum tight container 1 having an electronemissive cathode 2 of conventional type adjacent one end and an anode or anticathode 3, also of conventional type, adjacent its opposite end. The cathode 2 is backed by a conventional shield 4 and supported on a glass stem 5 through which heating-current in-leads 6 are sealed. The cylindrical tube wall 7 of glass extends from the stem 5 surrounding the cathode 2 and shield 4. A similar glass tube-wall section § surrounds the anode 3 and is attached to a tube 9 extending from the anode by an annular seal, thus permitting a cooling-fluid duct 11 to reach the anode 3. A shielding sleeve 12 extends from the anode 3 and provides an aperture through which the electron stream from the cathode 2 is projected onto anode 3.

It has been a practice in the prior art to sectionalize the tube wall at its mid-portion by connecting the glass walls enclosing the cathode and anode as described above with a metallic ring or tube to which the glass end-portions were respectively sealed. This ring was kept at ground potential and surrounded a sleeve w'hich, in turn, surrounded the electron stream flowing from the cathode to the anode, and was electrically connected to the ring. The supply voltage source had its mid-point grounded, and hence connected to the aforesaid ring, and maintained the cathode at a high negative potential relative to the midpoint ring and the anode at a high positive potential relative thereto. The ring therefore sectionalized the tube wall into two parts, and in some degree subdivided the electric gradient inside the tube into two regions respectively adjacent the anode and the cathode at opposite ends of the tube. This subdivision of the electric gradient into two regions remote from each other gave a certain amount of assurance that if an arc-over ionized one of the regions, the trouble would clear up before the ions had time to reach the other region, and the arc-over would be limited to one end of the tube and be energized by only half the voltage impressed between the anode and cathode of the tube.

However, occasions arise in which the ionization of one region communicates itself to the region at the other end of the tube with the result that the two arc-overs and the mid-point sleeve surrounding the electron stream form a dead short circuit between the anode and cathode energized by the full voltage of the tube-supply voltage.

In Fig. 1,1 have modified the above-described prior-art structure by connecting the glass tube-portions 7 and 8 by means of a composite wall-structure comprising two metallic ring-portions 14 and 15 which are themselves interconnected by a tubular glass wall-portion 16. Each of the ring-portions 14 and 15 surrounds and supports a metallic sleeve 17 (or 18) through which the electron beam from cathode 2 to anode 3 passes. The ring-portions 14 and 15 are connected to ground respectively through high resistors 19 and 21. The voltage-source supplying the tube consists of two, preferably similar, units one of which 22 has its negative pole connected to cathode 2 and its positive pole grounded, while the other 23 has its negative pole grounded and its positive pole connected to the anode.

If, with this arrangement, no arc-over occurs, the combined voltage of sources 22 and 23 is impressed between

2,871,402 anode 3 and cathode 2, and no substantial current flows through resistors 19 and 21 so that they waste neither energy nor tube voltage regardless of their high resistance. The voltage gradient is of course concentrated in the two regions between cathode 2 and sleeve 17, -and between sleeve 18 and anode 3. Any arc-over that occurs will take place at one of these regions of voltage gradient concentration; let us say- between cathode 2 and sleeve 17. Current from voltage of source 22 will then flow through resistor 19 and the voltage of the source will be used up in drop through that high resistor, limiting the energy of the arc-over between cathode 2 and sleeve '17 to a low value that will result in its rapid disappearance. The low concentration of energy in the arc-over and the shielding presence of sleeves 17 and 18 greatly reduces the probability of ionization’being communicated to the other region of concentrated gradient between sleeve 13 and anode 2 to cause arc-over there. Resistor 21 affords a similar protection in case of arc-overs between shield. 18 and anode 3. In the remotely probable instance of simultaneous arc-overs at both cathode 2 and anode 3, the insulation of shields 17 from shield 18 would make it necessary for arc-over to occur in the gap between shields 17 and 18 before a direct discharge from cathode 2 to anode 3 could occur; and there is little chance of this occurring before the discharge at cathode 2 or anode 3 is extinguished.

The Fig. 1 arrangement is thus an improvement over the above-described prior-art sectionalized high voltage tube.

In the X-ray tube shown in Fig. 2, I employ a different principle for minimizing the tendency to arc-over between points of different electrical potential. I find that probably the chief agencies active in producing heavy arc-overs in high voltage evacuated tubes are relatively heavy charged particles such as positive ions of the residual gas always present at even the highest attainable vacua. Such ions may also emanate from the metallic electrode surfaces. Because their ratio of mass to electric charge is far greater than that of electrons, such ions follow different paths than electrons do through a magnetic or a non-uniform electric field, and I make use of this fact to separate the ions from the electrons and divert them to regions remote from the portions of the interelectrode space where electric graidents are concentrated, thereby minimizing the probability that the ions will cause flashovers.

Referring to the details of Fig. 2, an X-ray tube has a cylindrical anode section 31 having walls of suitable glass and a cylindrical cathode section 32, these having axes which intersect at an angle to each other. The anode 3 and cathode 2 have the same structure as anode 3 and cathode 2 in Fig. 1 and are believed to need no detailed description here. The anode section 31 and cathode section 32 are sealed to opposite ends of a tubular metallic mid-section 33, provided near its opposite ends with metal diaphragms 34 and 35 having apertures 36 and 37. Mid-section 33 may be connected to the grounded termials of voltage sources 22—23 by a resistor 40. A unidirectional magnetic field normal to the plane of Fig. 2 is induced in the mid-section 33 by any suitable means such as magnetizing windings (not shown), and is given such a strength that an electron emitted by cathode 2 and accelerated to the voltage of source 22 through - aperture 37 will follow a path 38, emerge through aperture 36, and pass through aperture 13 into impact with anode 3. However, any negative ions accelerated through aperture 37 from the space inside tubeportion 32 will follow a nearly straight path 39 and discharge on diaphragm 34. Similarly, any positive ions accelerated from tube-portion 31 through aperture 36 will follow a nearly straight path 41 and discharge on diaphragm 35. Thus, highly accelerated ions are prevented from passing but of the confines of mid-sectiond 33 into the regions of concentrated electric gradient near cathode 2 and anode-aperture 13 where their impact with residual '

gas molecules would be likely to cause cumulative ionization and arc-overs.

Fig. 3 shows the middle portion only of a tube otherwise similar to that of Fig. 2, but in which an electric field parallel to the plane of the figure replaces the magnetic field of Fig. 2. This electric field is set up between a pair of electrodes 45 and 46 which are supported within mid-portion 33 by insulated in-leads 47 and 48 respectively. Electrons projected from tube-portion 32 through aperture 37 follow a curved path 38 through aperture 36 to the anode. Negative ions, however, by reason of their greater ratio of mass to electric charge follow a nearly straight path 39 into impact with electrode 46. Positive ions, for similar reasons, follow a nearly straight path 41. For reasons similar to those given in discussing Fig. 2 both types of ion avoid regions where they would be likely to cause flashover across the entire tube.

Fig. 4 shows a modified form of high voltage vacuum tube embodying arc-over protective arrangements embodying the principle of my invention. Again, I use an X-ray tube to illustrate the invention. In this embodiment I subdivide the interior of the tube into three sections which are interconnected only by channels through two chambers having interiors which are free from electric fields and which have relatively small entrance and exit apertures. Such channels render very low the probability that ionization present near one electrode will communicate itself to the region adjacent another electrode; but in addition the entrance-and exit aperture in each chamber lie on opposite sides of a diaphragm having only a single small aperture positioned so that a particle must follow a curved path to pass through the chamber. Magnetic fields analogous to that in Fig. 2 are arranged to cause electrons to follow the required curved path, while ions collide with the chamber walls and are discharged.

Referring in detail to Fig. 4, an X-ray tube comprises a cathode 2, an anode 3, a cathode tube-portion 32, an anode tube-portion 31 and supporting structures therein similar to those of-similar reference-numeral in Fig. 1. These two tube-portions are interconnected by -metallic wall-portions 14, 15 and a glass wall mid-portion 16 like those similarly numbered in Fig. 1. Supported from the metallic wall-portions 14 and 15 by metal partitions 51, 52 are two cylindrical metallic chambers 53, 54, having apertures 55, 56 and 57, 58 in their endwalls. Midway of the respective chambers 53, 54 are partitions having apertures 62, 63 positioned to one side of the straight lines’ between the apertures 55, 56 and 57, 58, respectively. A magnetic field is provided, by means not shown, in the region of aperture 55 so oriented as to deflect an electron passing from cathode 2 through aperture 55 to the proper side of the straight line between apertures 55 and 56 so that it travels toward aperture 62; arid a similar magnetic field of opposite polarity is provided in the region of aperture 62 to reverse the curvature thus introduced into the- electron path. These magnetic fields are given the strength required to make the reversely-curved path pass through aperture 62. Similarly, a pair of magnetic fields at opposite polarities are provided on the other side of aperture 62 and in the region of aperture 56 and given such strength that the aforesaid electron after passing aperture 62 follows a reverse-curve path through aperture 56 traveling directly toward aperture 57 in chamber 54. By reason of their much greater ratio of mass to electric charge neither positive nor negative ions passing through aperture 55 of aperture 56 will be deflected from a straight-line path sufficiently by the said magnetic fields to pass through aperture 62.

A similar arrangement of magnetic fields adjacent the apertures 57, 63 and 58 will insure that electrons will follow a curved path passing through these three apertures, but that no ions will be able to do so. Electrons emitted by cathode 2 can therefore follow the above-described curved paths through apertures 55, 62, 56, 57, 63 and 58

3,871,402

S but any ions generated by collision with residual gas molecules along the path between cathode 2 and anode 3 will collide with the metallic walls of chambers 53 and 54 and be discharged. Resistors 19 and 21 similar to those described in Fig. 1 may be applied to the Fig. 4 arrangement if desired; however, conventional methods of fixing the potentials of sleeves 53 and 54 such as tapping them to intermediate voltage points on potentiometers or the like which have this voltage between anode 3 and cathode 2 impressed across them are an alternative to the resistor 19—21 structure.

The tube of Fig. 5 comprises an envelope 1 which may be quite similar to that in Fig. 1, containing a cathode 2, anode 3 and other elements 4 through 16 like those of similar numeral in that figure and so needing no additional description here. The metal annulus 14 supports a metallic sleeve 71, having an arcuate projection 72 in the lower portion of its wall, and is connected to ground through an impedance 73. Metal annulus 15 similarly supports a sleeve 74 having a similar projection 75 in the upper portion of its wall, the projections 72 and 75 overlapping somewhat along the tube axis. Sleeve 74 is connected to ground through an impedance 76. Fig. 6 shows a schematic view on lines VI—VI of projections 72 and 75.

During normal operation the electron path from cathode 2 to anode 3 is along the axis of sleeves 71 and 74 and no substantial current flows in impedances 73 and 76. The supply-voltage source will have its midpoint grounded, making 2 negative and 3 positive. Should an arc-over occur between either sleeve 71 and cathode 2 or sleeve 74 and anode 3, (suppose it is the former) current will flow to ground through resistor 73 giving sleeve 71 a substantial negative voltage relative to sleeve 74, and setting up an electric field component normal to the tube axis from projection 75 to projection 72. This field component will deflect electrons and any ions which come into this region from the normal axial path between cathode 2 to anode 3 so that ions from the discharge between cathode 2 and sleeve 71 cannot move into the space inside sleeve 74 and initiate a discharge between sleeve 74 and anode 3. The discharge is therefore confined to flow between cathode 2 and sleeve 71, and is limited to low value by resistor 73 so it is quickly extinguished, without ever initiating a short-circuit path of low impedance direct from cathode 2 to anode 3.

Fig. 7 shows a modification of the shielding sleeves 71 and 74 of Figs. 5 and 6 in which the projections 77 and 78 are plane instead of arcuate in section.

Fig. 8 shows a modification of these shielding sleeves in which the adjacent ends of the sleeves are slanted at an acute angle to the central axis.

Fig. 9 shows a tube in which the split sleeves 71 and 74 of Fig. 5 are replaced by a single shielding sleeve 91 supported on a single metallic annulus 92 sealed into the mid-point of the tube walls. The sleeve 91 is connected to ground through a magnetizing winding 93 which sets up a magnetic field transverse to the axis of sleeve 91. As long as current flows directly from cathode 2 to anode 3 in normal fashion, substantially no current flows in winding 93. If however an internal breakdown occurs inside the tube so that current flow starts from sleeve 91 to either cathode 2 or anode 3, current flows to the grounded mid-terminal of source 22—23 through winding 93, and the magnetic field thus produced deflects electrons and ions from the direct path between cathode 2 and anode 3 and prevents a discharge following that path. The winding 93 may be given any desired impedance which will limit the current-value of the discharge.

Under stable conditions, the electrical potential of the various shields and sleeves in the above-described embodiments of my invention is determined in large part by the fraction of the electrons of the anode-to-cathode discharge which they intercept and by the number of secondary electrons emanating from the anode which reaches them, such electrons flowing away through their various grounding and ofi-take resistors and producing IR drops therein.