US2872721A - Electron image multiplier apparatus - Google Patents

Description

Feb. 10, 1959 J. D. McGEE ELECTRON IMAGE MULTIPLIER APPARATUS Filed April 11, 195'! 3 Sheets-Sheet 1 5 Sheets-Sheet 2 FIG.

Feb. 10, 1959 J. D. M GEE ELECTRON IMAGE MULTIPLIER APPARATUS Filed April 11, 1957 3 Sheets-Sheet 3 FIG. 7.

FIG. 8.

assembled so that when out into slices, the alternate slices can be reversed to form a set of dynodes and the square cells of one dynode will accurately match the square cells of adjacent dynodes.

United States Patent '1 2,872,72l, ELECTRON IMAGE MULTIPLIER APPARATUS 5 James Dwyer McGee, London, England Application April 11, 1957, Serial No. 652,295 Claims priority, application Great Britain April 12, 1956 7 Claims. ('Cl. 29-2514) This invention relates to electron image multiplier exp paratus and the object of the invention is to provide improvements in multiplier apparatus made in accordance with the principles described in the specification of British patent application No. 15,508 of 1953. In said apparatus light rays forming the image to be multiplied are focussed on to a transparent photo-electric surface from which they liberate photo-electrons which are in turn focussed to form an image on the first of a series of dynodes, i. e. electrodes from which secondary electrons are liberated in successively increasing numbers, each dynode comprising a set of inclined plates parallel to each other, transverse'to the path of electron flow and inclined so as to intercept all, or nearly all, electrons and another set of plates at right angles to the inclined plates and dividing the spaces between the inclined plates into cells and minimising lateral dispersion of the electrons.

Adjacent dynodes are set in reverse position to each other so that the inclined plates of one dynode are in side view at angles which are equal and opposite to the angles of the inclined plates of the other dynode, thereby forming a zig-zag path for the electrons through the pack of dynodes.

The dynodes may be mounted in a highly evacuated glass envelope having a flat window at one end on the inner surface of which is the photo-electric surface, while the dynodes are mounted in the envelope, means being provided between two ends for accelerating and focussing the electrons. A large potential dilference will be provided between the photo-electric surface and the first dynode, and a further increase of potential on successive dynodes, and between the last dynode and a fiuorescent screen on which the final multiplied image is viewed.

According to the present invention a dynode is made The tubes will be accurately pre-formed and accurately The above and other features of the invention will now be described by way of example with reference to the accompanying diagrammatic drawings wherein:

Figure 1 shows an electron image multiplier apparatus made in accordance with the invention;

Figures 2 to 5 show stages in the method of making dynodes;

Figure 6 shows the front view of a dynode;

electric surface or cathode 11 at one end and a screen 12 at the other end. The light rays forming the image to be multiplied are focussed on to the surface 11 from which photo-electrons are liberated which are in turn focussed by electro-magnetic means (not shown) to form an image on the first of a series of dynodes 13.

The dynodes are made by assembling a stack of nickel tubes 14 as shown in Figure 2. The tubes 14 are about 0.002 to 0.003 inch wall thickness and rectangular crosssection, say at x 20! very precisely. The dimension d for the individual tube may be 1 mm. and the dimension D for the stack may be 1 inch. The tubes may be made A by forming cylindrical tubes to a rectangular cross-section cooling. The stack is then removed from the jig 16 and, clamped on to the bed of a milling machine and sawn into slices with a Carborundum cutting wheel having a thickness of 0.01 to 0.02 inch. The planes 22, 23 of cutting these slices in planes perpendicular to the surfaces of the narrow sides and at approximately 45 to. the surfaces of the wide sides so that the cells formed by the tube sections are of approximately square crosssection. Thus the sides 22 of the tubes and stack are the same length as the sides 23 i. e. the cells made by the sectioned tubes are square and the slices are also square. Each slice will form a dynode. If desired the assembly may be embedded solidly in say Perspex" (Registered Trademark) which fills the tubes and so prevents distortion while it is being sawn into slices; the Perspex being then dissolved out.

Each slice is clamped to a plate 25 (Figure 4) notched,

at its edges to receive wire 27 (e. g. copper wire of 0.001 inch diameter) which is to form a screen between each pair of adjacent dynodes. The wire is wound in any suitable manner e. g. to form one strand across each cell in each direction. Alternatively if desired the wire may be arranged diagonally as shown in Figure 12.

If desired the wire 27 may be passed across the mouth of a cell, through the cell and around the plate structure, across the mouth of the next cell, and so on, as shown in Figure 5. The wire may be looped around the plates at their intersecting positions e. g. at every junction or at each second, third or fourth junction. It is not easy to attach a conventional metal mesh to such a structure in such a way that it cannot lift or buckle suificiently during processing to cause short-circuits between dynodes, but by the improved method above described the mesh can be tightly fixed so as not to lift from the dynode.

The dynode is then coated with wax except at positions where the wire crosses the edges of the cells. The dynode is then subjected to a metal plating process so that the deposited metal plating fastens the wire to the edges of the cells. The wire strands are then severed between the dynode and the plate 25.

The solder melts and runs down between the tubes which are thus soldered together on The dynode now consists of a set of cells which as shown in Figures 6, 7 and 8 consist of a set of plates 28 parallel to each other and inclined to the general plane of the dynode and another set of plates 29 at right angles to the inclined plates and dividing the spaces between the inclined plates into cells. With this technique such dynodes may be made with 25 or more apertures per inch and hence for a 100 line picture (10 picture points) which represents very useful image definition, dynodes 4 x 4" would be required. The construction of dynodes of large cross-section may become diflicult. The dynodes may therefore be made in convenient cross-section, e. g. l" x 1", and then assembled to form larger areas such as 4" X 4", which would be formed from 16 sections each 1" x1".

The dynodes are now mounted with alternate dynodes reversed in position as compared with their original positions in the stack of tubes. That is to say in the original stackthe planes 31, 33 in Figure 9 were adjacent each other whereas after reversal the planes 31, 32 are adjacent each other so that the cells in these dynodes are at equal but opposite angles to the general planes of the dynodes so that zig-zag channels extend through a pack of dynodes. The aforesaid insulating cement may be filled into some cells e. g. the four corner cells to hold the pack together.

If desired the adjacent edges of the plates of adjacent dynodes e. g. edges 31, 32, Figure 9,, may be provided with a coating 34 (Figure 10) of electrically insulating material. The dynodes are then mounted in contact with each other, or in contact with a wire mesh between them. This arrangement ensures that secondary electrons originating in one cell of one dynode will pass into one cell only of the next dynode and not stray laterally into neighbouring cells. A clear image definition is accordingly maintained. The insulating material may be provided on the edges on the side of the dynode from which secondary electrons emerge and may form a thin ridge. The material may be an enamel such as can be made from a mikture of potassium silicate and lumina. powder. This may be made of the'consistency of thin cream and smeared on a flat surface of a fairly soft rubber pad. The surface of the rubber surface, which causes the liquid enamel mirrture to gravitate towards the metal edges. 7

On removing the dynode vertically from the rubber pad it will take away a narrow rim of cement along each edge which on drying and baking becomes a hard insul'ating enamel. It now becomes practicable to place each dynode directly on its neighbour, align the respective apertures accurately and then cement it solidly in place by means of the enamel. The rim of insulating enamel or cement will preserve insulation between dynodes while at the same time the walls of; successive dynode apertures are substantially in contact.

The enamel may be slightly conductive to provide a gradual change in the electric field. The dynodes may be arranged so that the cell edges of one dynode meet the corresponding edges of the cells in the adiacent dynode so that looking in the direction of the electron beam one cannot see through the pack, alternatively the adjacent inclined walls of the cells may be spaced sutficiently apart from each other as to leave small gaps through the pack of about 0.005 to 0.02 inch to receive feeler gauges 40 through say four sets of cells, as shown in Figure 11 which may be left until after the cement has set. In this way the feeler gauges may be used to achieve accurate alignment of all the dynodes in a stack. Alternatively the small gaps make it easy to align the dynodes by optical means.

I claim:

1. A method of forming a dynode which comprises assembling rectangular section thin walled tubes into a stack, said tubes having two opposite sides of equal length and longer than the other sides which are also of equal length, cutting a slice from said stack in a direction parallel to the long sides and at such an angle to the short sides that the cross-section of the cells formed by the tube sections is approximately square.

2. A method as claimed in claim 1, wherein the stack is secured in a jig, the open ends of the tubes at one end are plugged, solder is placed on that end of the stack with that end upwards, the stack is heated to cause the solder to flow down between the tubes, and the stack is then removed from the jig and cut into slices.

3. A method of making a set of dynodes which comprises cutting slices from a stack to form a series of dynodes by the method claimed in claim 1, reversing alternate slices and mounting them in a stack to form zig-zag paths of square cross-section for the electrons,

4. A method as claimed in claim 3, wherein the opposed edges of adjacent slices in the set are coated with electrically insulating material and a wire screen is provided between them, the coated edges being in contact with the screen.

5. A method as claimed in claim 4, wherein the screen is provided by a wire that is passed across the mouth of each cell and through some of the cells.

6. A method as claimed in claim 4, wherein the wire screen is attached to the cells by metal plating.

7. A method as claimed in claim 1 wherein a stack of dynodes is assembled into a rigid structure by filling a series of corresponding cells with a cement which sets to form an electrically insulating solid; material.

References Cited in the file of this patent UNITEDv STATES PATENTS 2,619,438 Varian et a1 Nov. 25, 1952.,

ara;

Images