DE1803947A1 - Camera for repeated phased operation - Google Patents

Description

DR-INQ. DlPL.-ΙΝβ. M.SC. OtPL. PH YS. OR. DtPL-PHVS.

HÖGER - LEGAL RIGHTS - GRIESSBACH - HAECKER

_# PATENT ANWÄUTe IN STUTTGART A Q f | QQ Λ

A 36 810 b

October 10, 196S

Texas Instruments Incorporated Dallas / Texas, USA

Camera for repeated step-by-step operation.

The invention relates to a camera for repeating stepwise Operation with a computer-controlled pilm table for producing a multiplicity of images along a multiplicity of vertical axes.

The invention particularly relates to a camera for producing a set of photolithographic diaphragms or lenses for Manufacture of large series of semiconductors. 3in cal lead, like

009823/0691

BATH ORIGINAL

October 10, 1968
ly-35

For example, a transistor, is usually fabricated using a number of diffusion operations. With everyone Diffusion stage is a layer of photosensitive polymer, known as a photo opacifier, on a silicon dioxide layer applied to the surface of the semiconductor carrier. A mask is pressed against the surface of the opaque material and this exposed. When the opacifying agent is photographically developed, certain areas of the opacifying agent are removed to protect the to expose underlying silicon dioxide layer. The free

vn T * ti

laid silicon dioxide layer / then removed by an etchant, which does not attack the opacifying agent to expose the underlying semiconductor material. The covering medium will then stripped from the silicon dioxide along with impurities which have penetrated into the surfaces of the semiconductor material which have been exposed through the openings in the silicon dioxide layer. Either is on the exposed Dividing the semiconductor material during the diffusion

Process a new silicon dioxide layer has grown or it such a layer is subsequently deposited, whereupon the process is repeated for the next diffusion stage.

Each subsequent diffusion will only be part of the previous one Diffusion or carried out in another part of the semiconductor die, so that one for each diffusion stage another mask is necessary. Each mask consists of a square, flat glass, on one surface of which one is photographed Fixed emulsion is applied with high image resolution, which has opaque and transparent areas. There the area of the luaske to which the fixed emulsion is applied is printed directly against the platelet, the patterns of the mask must be of actual size, with dimensions from a few millimeters down to a few, "millionths

009823/0691 - 3 -

1803S47

A 36 810 b - 3 -

October 10, 1968

Millimeters are included, with conductor widths in the order of magnitude of 1 / 1000mm is usually considered to be the ultimate limit when a silicon dioxide layer is used as a mask.

Semiconductor material can be more easily grown, handled and treated in the form of disc-shaped platelets that have a nominal diameter of about 38 mm and a thickness of about 0.25 mm. For this reason, a large number of semiconductors% on each disc is produced using the same process steps simultaneously. It is also customary to manufacture semiconductor diodes, resistors and capacitors for a complete circuit on the same semiconductor substrate and then to connect the individual elements with lines that are carved out of a metal film that is formed on the surface of a silicon dioxide layer by the same photolithographic process was knocked down. There are openings in the oxide layer, v / o the metallic conductors connect the individual active elements. The manufacture of integrated circuits usually requires a large number of diffusion steps and thus a large number of photographic glass for the diffusion steps and also an additional glass for the metal foil in order to create the pattern of the interconnection lines. It is also customary to simultaneously manufacture a large number of integrated circuits on each individual sheet of semiconductor material by the same process steps.

It is impractical, if not impossible, to make a mask for a large number of individual semiconductors or integrated circuits by drawing the entire mask in larger dimensions and then photocopying the whole mask. The main part of any mask that relates to a "particular semiconductor, group of semiconductors or an integrated circuit ,

D09823 / 069 1

BATH ORIGINAL

1803847

A 56 810 b - 4 -

October 10, 1968

however, it can first be manufactured on a much larger scale and then optically reduced to a real size photograph. The photograph can then gradually increase a photographic plate can be applied to create the complete mask. However, it is extremely important here that the light image is arranged with the greatest accuracy in each successive exposure position. Otherwise the successive masks do not completely match and the advantage is small.

One way of overcoming these difficulties is to use all the masks in a sentence at the same time in one Manufacture camera with multiple fits (multibarreled). The same local ones then appear in all the basques of the set Errors so that the masks or apertures match exactly. However, this method is not practical. Every mask is suitable only for a limited number of exposures, e.g. for 20 to 40. Because a relatively large number of discs or platelets prove to be defective at an early stage of the procedure proves, a significantly larger number of masks is necessary in the first stages of manufacture than in the last stages. In normal mass production, this process would quickly waste a large number of masks Lead time.

Integrated circuits are widely used as storage elements and used as logic gate circuits for digital computers and automatic control systems. This is often a large number of individually packaged integrated circuits connected by printed circuit boards or similar techniques to form a large system. In the last few years one has

- 5 009823/06 91

ORIGINAL

Ä 36 810 b - 5 - 1803847

October 10, 1Q68
ly-35

recognized the convenience of having a large number of integrated circuits on a single die of semiconductor material make the circuits in place on the platelets to check, and then only through the correct circuitry Thin film lines, which have been deposited on the platelets in one or more layers, to connect to form a row the. However, since a quarter to a third of the circuitry on a die can be faulty and these can be faulty Circuits on the platelet are randomly distributed, a very large number of different combinations can result result in error-free circuits. A customized mask or group of masks must therefore be made to apply the pattern of thin film lines to each individual die. This is when using the usual methods extremely impractical. The wiring mask can be generated by a system controlled by a computer. One such However, the system assumes that each element or each circuit is based on a predetermined location with great accuracy the disc is arranged. If this is not the case, it may be a shorted or open circuit at one point occur where the wiring pattern does not match the circuitry, causing the entire row to fail would.

There is, therefore, a need for a system for creating masks in which the location of each pattern on the mask is set with an accuracy of the order of a few microinches. The best systems for hiring the table of a camera or any other movable device, e.g. for the automatic setting of machine tools, currently available have a setting accuracy on the order of 0.001 mm (40 microinches), i.e. there is an improvement of about an order of magnitude

009823/0691 " ⁶ "

BATH

"A 36 810 b - 6 - -ι

October 10, 1968 '

conducive. In addition, only masks with a length and width of about 38 can now be produced with the previous cameras v / ground (1.5 inches), although dies are available that are approximately 7.55 cm (3 inches) in diameter. About 1000 individual integrated circuits can be accommodated on a small plate, whereby this small plate has a nominal diameter of about 2.5 cm (1 inch) and

e twa on a plate with a nominal diameter of / 7.5 cm

»I vv" H ρ τ * -

(3 inches) can ground approximately 10,000 circuits without reducing the size of the circuits. When using the previous cameras, such a large number of exposures would take an extremely long time.

The invention is therefore based on the object of creating an improved camera with which a photo mask is automatically created or a full set of photomasks to create a variety of integrated circuits or the like. Manufactured with dimensional accuracy on the order of a few microinches v / ground (1 inch = 25.4 mm).

According to the invention, this object is achieved by a table for placing a photographic plate and a projection system for projecting an image onto this plate, the table and the projection system moving relative to one another can run in the X and Y coordinate directions, also by a drive for moving the table and the .Projection system in the direction of the X-coordinate and the Y-coordinate, a detector for sensing the movement of the table and the Projection system in the coordinate directions that generates information corresponding to these movements by a control that responds to this information

and actuation of the drive to the table / projection system

optionally to move in relative coordinate positions, and around

009823/0691

BATH

A 36 810 b - 7 -

October 10, 1968

a relative coordinate position during the exposure of a to keep the photographic plate placed on the table.

The movement of the table in the X and Y directions is expediently carried out with the aid of a laser interferometer and an interference fringe counter, whereby the table is continuously adjusted in the X and Y directions by drive systems that are controlled in real time by a digital computer * controlled v / earth. The computer is programmed in such a way that it calculates the current position of the table from the number of strips, the desired sieving of the table and the force required for the drive system to move the table into the desired position and to move it in this position for the Hold duration of an exposure. The table is moved to a zero reference position by the computer in order to calibrate the fringe counters to calculate the interference numbers

are necessary to move the table to the desired exposure position and to continuously calculate the forces which are necessary to bring the table into the desired position and to keep it in this position. the desired tolerance and during the time necessary for the exposure J. This procedure is for everyone Exposure repeated.

The table is on air bearings in a precisely specified height movable and it is guided with the greatest precision in the X and Y directions by an air bearing guide these systems have very low friction.

The projection system according to the invention has vertically movable upper and outer frames, which have the corresponding I.Iutter transparencies and wearing lenses. The moving table and

- 8 009823/0691

BAD ORIGIN * ¹ -

A 36 810 b - 8 -

October 10, 1968

the upper frame have a common adjustment system, through which standardized photographic plates are easy, quick and accurate can be aligned in specified reference positions. This system allows the total number of exposures largely reduced by a two-step square root method.

The projection system simultaneously projects a large number of images onto the film that is placed on the table, after the stage is in any of a variety of predetermined exposure positions has been moved. A detector indicates when the table is in a Hull reference position and resets the counters. The computer calculates the coordinates of each exposure position and then calculates on the basis of the current barometric pressure, the number of stripes from the reference position to the first exposure position. the The position and the speed of the table are continuously monitored by the computer on the basis of the information provided by the strip counter calculated and the drive system operated accordingly. The table is then held in the exposure position during the exposure, its position being continuously changed by the Stripe counter is determined and the drive system is actuated to generate forces to correct the position errors.

Exemplary embodiments of the invention are provided below explained with reference to the drawing in which

Fig. 1 is a front view of a camera according to the invention.

Fig. 2 is a top plan view of the camera of Fig. 1 with the top frames removed for to show the movable table.

- 9 009823/0891

BATH

Oct 10 1 £ 68 ' ^w

Pig. 3 is a simplified perspective view of FIG

Bracket and drive for the movable table of the camera according to Pig. 1.

Pig. Figure 4 is a front view of part of the table of the camera of Figure 1 with parts removed are to show details of the construction.

Fig. 5 is a plan view of part of the table g according to Pig.4, with parts removed.

Pig. Figure 6 is a rear view of part of the table of Figure 5 with parts removed.

Figure 7 is a section along line 7-7 of Pig.5.

Pig. 8 is a partial section showing a detail of the table of FIG.

Fig. 9 is a plan view of a carrier for a

Variety of panels for the camera according to Fig.1, with parts removed to give details of the construction to show.

Fig.10 is a rear view of the carrier according to Pig.9

Pig. 11 is a side view of the carrier of FIG. 9,

with parts removed to show details.

Pig.12 is a front view of the carrier of Fig.9>

v / with parts removed to show details.

Fig. 13 is an enlarged plan view of a corner of the carrier according to Fig. 9.

Fig. 14 is an enlarged section taken along the line 14-14 of Fig. 9.

Figure 15 is a view of the outer surface of one of the Stand of the camera according to Pig.1.

- 10 -

009823/0691 _offlQlNA u

A 36 810 b - 10 -

October 10, 1968

Pig. 16 is a section along line 16-16 in Fig. 15.

Fig. 17 is a view of the inside surface of the stator according to Fig. 15.

Figure 18 is a section along line 18-18 of Figure 17.

Fig. 19 is a top plan view of the stand of Fig. 15. Fig.20 is a section along line £ 0-20 of Fig.15.

Fig.21 is a plan view of the upper stage of the camera of Fig.1.

Fig.22 is a front view of the upper step of Fig.21. . »

Fig. 23 is a bottom view of the upper step after Fig. 21.

FIG. 24 is an enlarged view of that broken off in FIG Part.

Figure 25 is a section taken along line 25-25 of Figure 21.

Fig.26 is a plan view of the lower frame of the Camera according to Fig. 1.

Figure 27 is a section taken along line 27-27 of Figure 26.

Fig. 28 is a schematic block diagram of the controller for the camera according to Fig. 1.

Figure 29 is a more detailed schematic block diagram a part of the control according to Fig.28.

Fig. 30 is a diagram showing the operation of the controller according to Fig.29 shows.

Fig.31 is a schematic diagram used for explanation of the operation according to Fig.1 is used.

- 11 -

009823/0691

A 36 81 Ob - 11 -

October 10, 1Q63

.Pig.32a and 32b together show a simplified flow diagram of the program of the computer shown in Pig.28, which is used to control the Camera according to Fig.1 is used.

Figs. 33 to 41 are flow sheets and show sub-routines of the program according to Fig. 32a and 32b.

Fig. 42 is a graph showing the

■ Indicates the forces that occur at a given ™ Positional errors applied v / earth to the table during 'exposure in a predetermined to hold onto.

A camera 10 is shown in FIG. The camera 10 is mounted on a solid concrete block 12, which is suspended from springs (not shown) in order to isolate it from vibrations from the earth. The springs are adjustable so that the concrete block can be adjusted in height, '..' If desired, a pneumatic, self-adjusting system can be used. A base plate 14 is attached to the concrete block 12 μ by feet 16. A lower granite block 18 is supported by the base plate 14 by three threaded bolts 20 and nuts 24 arranged in a triangle, which sit on the base plate 14. The upper surface 22 of the block 18 is exactly level, and by setting the buttern 24 on the threaded bolts 20, which sit on the base plate 14, exactly leveled. A table 26 for placing a film can be moved in the coordinate directions X and Y over the flat surface 22 of the lower granite block 18.

A pair of supports 28 and 30 are connected to the base plate 14, with a support on each side of the table 26 after

009823/0691

0RIQWAL

Ά 36 810 b - 12 - ionon / -7

extends above. Upper and lower frames 32 and 34 are attached to the supports 28 and 30 and are vertically movable and adjustable. The upper frame 32 supports two lamp housings 36 and 38, each of which contains nine light sources. Each light source has a lamp and a lens system to project light in the direction of eighteen separate optical axes. The upper frame has means for holding a Liutter transparency for each optical axis. The lower frame 34 carries a reduction lens for each of the optical axes and a bellows 40 for each optical axis extends between the upper and lower frames. The table has means for holding a photographic plate on each optical axis so that it is exposed to the image produced by directing light from the light source onto the photographic plate through the corresponding transparency body and the reduction lens. Each of the transparency bodies bears the pattern which is necessary for a photomask which is used for the various stages of semiconductor manufacture. When the table 26 is set to successive exposure positions, all of the plates placed on the table are exposed at the same time, so that the exposures have the same positional errors. All the patterns on the mask can be made so that they all match. The eighteen separate optical axes make it possible to manufacture a set of masks for an eighteen-step manufacturing process.

According to the invention, any exposure can be made on any film plate at any desired point within a square with a side length of several inches (1 inch = 25.4 ran) set with a setting tolerance of a few Llicrozolls. This doesn't just require adjustment of the table 26 within this tolerance, but it must also

009823/0691 - 13 -

A 36 810 b - 13 -

October 10th

the camera 10 can be housed in a room in which the temperature is kept constant within a fraction of a degree. The expansion of the mechanical parts of the camera would otherwise move the transparencies or the plates by an amount that is greater than the specified limits. In the same "/ else can vibrations of the camera that arise or in the camera itself out of the earth, extensions and reductions cause, so that the overall% desired tolerances v / ground not met. These problems arise from the size of the camera 10, which is necessary in order to achieve the great freedom of movement and the high photographic reduction ratio of more than 20: 1.

The table 26 consists of a granite block 42 which is a metal plate 44 carries. The granite block 42 is supported by four conventional constant pressure air bearings 46 mounted on the Sit on surface 22 of granite block 18. Each of the air bearings 46 has a flat bottom surface that is adjacent to the flat Surface 22 of the lower granite block 18 is arranged. A gas, usually nitrogen, is passed under constant pressure through the Pumped center of each air bearing 46 so that the table 26 ™ is continuously supported by a very thin layer of gas,

zw, ei whose thickness is in the order of magnitude of etv; a / I.ii er on. Of the As a result, table 26 can be moved on granite block 18 with a minimum of friction. The gas supply and that for each Pressure regulators provided for bearings are not shown.

The movement of the table 26 is precisely controlled by a Pührungs- and drive system that _{y is} a first guide means in the form of glass rods 48 and 50 HAFC which are mounted on the lower block of granite 18th The end faces 48a and 50a of the rods 48 and 50 are optically flat and precisely aligned,

- 14 -

009823/0691

_BA0

a 36 810 b -η- 1803847

October 10, 1963

and the opposite end faces 48b and 50b are practically flat and parallel to the optically flat surfaces. An intermediate frame is formed by granite slabs 52 and 54, which are firmly connected to one another by a third granite slab 56 educated. The granite plates 52 and 54 are arranged on opposite sides of the glass rods 48 and 50 and the third granite slab 56 bridges the glass rods 48 and the slabs 52 and 54 are grounded by pairs of air bearings 53 and 55 which rest on the surface 22 of the granite block 18. A pair of brackets 48 and 60 in the shape of an inverted one U are attached to the granite slab 56 and run down and stand from the opposite end faces of the Glass rods 48 and 50. The bracket 58 carries a fixed one Air bearing 62, which rests on the optically flat end face 48a, furthermore a pneumatically preloaded air bearing 64, which rests on the opposite surface 48b, and that fixed air bearing 62 presses continuously against the end face 48a with constant force. In the same way owns the bracket 58 is a stationary air bearing 66 which rests on the optically flat end face 50a and a pneumatically preloaded one Air bearing 68, which rests on the opposite end face 50b, around the air bearing 66 continuously to press against the reference surface with constant force. The intermediate frame can therefore only move freely in the X direction and it is in a predetermined Y-coordinate position during its entire barrel kept within the given tolerance of a few liikroinches (1 inch = 25.4 mm).

A second guide is formed by the glass rods 70 and 72, which are attached to the plates 52 and 54 and optically have flat surfaces 70a and 72a which are aligned exactly at right angles to the optically flat surfaces 48a and 50a.

- 15 009823/06 91

A 36 810 b - 15 -

October 10, 1968

Opposing surfaces 70b and 72b are practically flat and practically parallel to surfaces 70a and 72a. A second pair of brackets 74 and 76 in the shape of an inverted U are attached to opposite edges of the granite block 42 and are provided with fixed air bearings 70 and 80 which rest on the optically flat surfaces 70a and 72a. Furthermore, with pneumatically preloaded air bearings 82 and 84, which sit on the end faces 70b and 72b and press the stationary bearings against the reference surfaces with constant force. %

The granite block 42 and thus the table 26 can move in the X direction are moved along the glass rods -48 and 50, which form a guide track, with the aid of a drive consisting of a Iüotor 66 is made with a printed circuit, and which is attached to a column 28 and drives a wheel 88 which engages one side of a drive rod 90 with frictional engagement. The connecting rod 90 is' with the: intermediate gcntoll by a rod ° 2 tied together. A pair of idle rollers 94 are spring loaded against opposite sides of the drive rod 90 biased to be between the driving wheel 83 and the drive rod 90 to maintain a practically constant force. That opposite end of drive shaft of idotor c6 is' provided with a pneumatically operated disc brake 96.

UÜJirch The Granitclock 42 and thus the table 26 can be moved in the Y-direction / a second I-Iotor 98 with printed circuit moves Werder., that drives a wheel 100. The wheel 100 frictionally engages one side of a rod 102 which is connected to the bracket 76 is and thus by a rod 104 with the granite block 42. The motor 93 for the drive in the Y-direction and idling Rollers, 106 are attached to the granite slab 54 by means of a bracket 10S. On the ".velle of I.Iotors 98 is also a pneumatically operated brake 110 is provided. The Hotor 86

009823/0691 - 16 -

BATH

A. 36 810 b - 16 -

October 10, 1963

thus moves the table 26 in the X direction and the motor 98 moves him in the Y direction. As will be explained in detail below, the brakes 96 and 110 are only used when that System is not in operation and they are not used to adjust the table during exposure.

The plate 44 of the table 26 is shown in FIGS. 4-8 in detail shown. The plate 44 accommodates two carriers 120 for a plurality of plates, specifically in precisely predetermined positions relative to the optical axes. At least four carriers 120 are necessary for full operation of the camera. One of the carriers 120 is shown in detail in FIGS. 9 to H. Any film carrier has a base plate 122. A side wall 124 is made is integral with the base plate 122 and extends around the entire circumference of the base plate. A lid 126 is through Hinges 128 and 130 attached to side wall 124.

The carrier 120 has nine identical compartments formed by inner walls 132 which are also formed in one piece with the plate 122 and the side wall 124. Aligned square openings 122a and 126a are provided in the base 122 and lid 126 in resilient compartment to allow light to be projected through a pin plate located in the compartment. Each compartment can accommodate a standard, square, photographic glass plate 134 and hold the plate in a precisely aligned position relative to the support. The orientation in the longitudinal direction of the optical axis, ie the orientation in the direction uric as well as the division and orientation around the X- mm T-J: closer, is carried out by three pins 136 ^ which extend from the base plate Ί2- ' into each compartment The plate 134 extends in the Z-IIK-Y-direction: and in the twisting direction and the Z-axis through!

- 17 _

0 9 8 2 3/0891

^ßA 0 ORIGINAL

October 10th
ly-35

138 and 140 set, d ^:; . | seated pivotably on pins 142 and 144, furthermore by a third rabbit 146, which is seated pivotably on a pin 148. The edges of the lugs 138, 140 and 146 are straight along the length of the edge of the glass plate to capture the edge of the plate 134 for a considerable distance, but are rounded in the direction perpendicular to the plate so that they only putty the edge the plate 134 grasp. This curvature can best be seen in Fig. 12 f.

Two adjacent sides of the plate 134 are against the rabbits 138, 140 and 146 by springs 150 and 152 and an elbow-shaped Biased member 154 which includes that of the plate facing the edges against which the tabs abut. That When the plate 134 is removed, member 154 becomes in its position held by a pin 156 in a not shown Oversized hole in member 154 is received. The head of the peg 156 is larger than this hole to allow the To hold member 154 in place on the spigot.

Each plate 134 is urged against the three pins 136 by leaf springs 160, the leaf springs being carried by the cover 126 and engaging the plate 134 directly above each of the pins 136. The cover 126 is held by two brackets 161 against the combined force of the leaf springs 160. Each bracket consists of a web 162 which is attached to the cover 126. In an opening 164 in each web tian the rounded end of a bolt 166 aufgoriomirK-m, which is slidably arranged Ln the side wall 144 i.; F. and by ^w : h a leaf spring 163 is preloaded outwards.

Each of the supports 120 has a pair of ebcrurr FlH oh-: u IYO and 172, di j on an outer surface of one of the Ueiten .'- ln I ^ !.: ■] -

00 9823/069 1 ¹³

A 36 810 b '- 18 -

October 10, 1 968

and a third surface 174 formed on the adjacent side wall. The surfaces 170, 172 and 174 on each carrier 120 are in a precisely predetermined relationship to the noses in each compartment of the wearer. The base plate 122 has a number of precise reference surfaces 176, as in FIG Pig · 9 is indicated in dashed lines, which is in a common plane lie. The ends of all pins 136 in all compartments are also in a common plane that is parallel to the The plane of the reference surfaces 176 lies. A handle 130 is attached to the side wall 124 to facilitate manipulation of the carrier 120 facilitate.

The supports 120 are used to hold the unexposed film plates for the table 26 as well as to carry nut transparencies for the upper frame 32 through which the light is projected will. The supports 120 are easy to install since they are only brought onto a table with the reference surface 176 facing downwards, whereby the lock 161 is released and the lid 126 is lifted.

The plates 134 can then be placed in the appropriate compartments by manually moving the member 154 against the force of the springs 150 and 152, whereupon the plates are placed on the pegs 136 with the photosensitive emulsion or transparent surface facing down, whereupon member 154 is released. The springs 150 then press the plate against the rabbits 138, 140 and 146, whereby it is precisely adjusted relative to the surfaces 170, 172 and 174 after sänt-LLühö plates 134 have been placed in the appropriate compartments on these M-iiüii , if the Deakel 126 is closed and you: locking 161 scaffolded, so : i-i3 dU; B'edern 160 Ui.ij I'lutfcen 134 -richer against did fiapfen 136 dril-Kön, through wel- v '"ji diu L'latten exactly parallel:; u dor level - u- uterflachen

009823/0891 "-19-

^ß AD ORIG | _NAL

180384

A 36 810 b

October 11, 1968

Iy - 124

176 to be aligned. The carrier 121 can then simply be brought to the camera 10 and inserted, as follows is described.

The plate 44 of the table 26 of FIGS. 4 to 8 receives a pair of the carriers 120, although in the figures only one carrier ^ is shown. The plate 44 has a W-shaped top plate 200, which forms two openings 202a and 202b, each of which is slightly larger than the row of nine openings 122a in the base plate of the carrier 120 is. The supports 120 can be inserted into openings in the front of the panel 44 and into their positions slid under openings 202a and 202b on rails 208 and 210 on opposite sides of the Openings run. Two lugs 212 and 214 are seated pivotably on bolts 216 * and 218, on the right side of each of the openings 202a and 202b, and they are spaced to capture the reference surfaces 170 and 172 on a carrier 120. A third lug 220 is rotatably seated on a pin 222 at the rear from each opening 202a and 202b, and is "spaced" to capture the reference surface 174 on the carrier 120. The carrier 120 is biased against the lugs 212, 214 and 220 by a lever 224 which is pivotable on a pin 226 is seated against the corner of the beam 120 by a spring 228 is pressed, as FIG. 8 shows.

Four adjustable bolts 206 are around each of the openings 202a and 202b, although only three are shown in FIG. 5 around the opening 202, and they extend down through the Upper plate 200 and end in a common level that lies parallel to the surface 22 of the granite block 18. A manually operated mechanism is provided on each of the carriers 120 from the rails 208 and 210 to bring it snugly against the respective groups of downwardly extending bolts 2Co

009823/0691 - 20 -

BAD ORlQlNAl-

180394

A 36 810 b ^

October 11, 1968
Iy - 124

to keep. This lifting mechanism is best seen in the right hand portion of Figs. 4 and 5 and in Figs. Two shafts 230 and 232 are rotatably mounted in side walls 23 ¹ J and and in a rib 236 of the plate 44 therebetween. Levers 244 and 246 are bolted to shafts 230 and 232, and a box girder 238 is pivotally connected to levers 244, 246 by bolts 240 and 242. Two spring loaded bolsters 248 and 250 extend upwardly from the box girder 238. A second carrier 252 is identical to the box girder 238 and runs between crank pins (not shown) on lifting plates 254 and 256 which are fastened to the shafts 230 and 232. A second pair of spring loaded bolsters 258 and 260 extend upwardly from the bracket 252. A connecting rod 264 is pivotally coupled to lift plates 244, 246 by pins 266 and 268, and an identical connecting rod 270 is pivotally connected to lift plates 254 and 256 so that shafts 230 and 232 can be rotated simultaneously by operating a hand lever 262, on the shaft 230 by a wedge or a feather key or the like. is attached. The lifting mechanism associated with opening 202a is identical, except that hand lever 262 is seated on shaft 232. When the hand lever 262 is moved into the position shown in the right half of the figures, the stools 248, 250, 258 and 260 are lowered so that the carrier 120 rests on the rails 208 and 210. When the lever 262 is moved counterclockwise to its left position, the stools 243, 250, 258 and 260 are raised to grip the cover of the carrier 120 and to raise the reference surfaces I76 on the base plate against the downwardly protruding bolts 206 .

By an image reduction and projection system comprising the upper frame 32 and the lower frame 34, which by the supports 28 and 30! are held, images are placed on the

■ j ■: · ;. BATH

A 36 810 b

October 11, 1968

Iy - 4

photographic plates 134 are projected in the carriers 120 and which are set in the table 26. The supports 28 and 30 are identical, which is why only the support 28 is described will. The support 28 consists essentially of a metal part, as shown in FIGS. 15 to 20, which is attached to the base plate 14 is screwed on. The support 28 must be very stable and be vibration-free in order to achieve the desired tolerances. At the support 28 are upper and lower cast brackets 300 and screwed on. Front and rear guide rods 304 and 304 which are arranged vertically extend between the brackets 300 and 302 306, a ball screw 308 and a calibrated rod 310, these parts all being attached to brackets 300 and 302 are attached. Two brake rods 312 are within the Support 28 and extend from a bracket 314 near the top of the support to a block 316 near the top Bracket 302, which in turn sits on two feet 3l8. The feet 318 are screwed to flanges 320, which are made in one piece with the support 28 are cast. The support 30 similarly carries front and rear guide rods 322 and 324, one central ball screw 326, a calibrated rod 328 and two brake rods 330, such as from Figs emerges.

The upper frame 32 has a base plate 350 which extends between the supports 28 and 30. The base plate 350 is through Sprockets 352 and 354 held, which are rotatably mounted in the base plate 350 and on the stationary ball screws 308 and 326 are screwed on. The base plate 350 is at two diagonally opposite corners by straight, elongated Ball bearings 356 and 358 are held on guide rods 304 and 324 run and are arranged around them * The ball bearings 356 and 358 are quite long so make a good one

009123/0691 "** V

α · 36 810 b 180394 ^.

October 11, 1968

Stability against movement about an axis that runs diagonally between guide rods 304 and 324 ^

The upper frame 32 can optionally be raised and lowered by an electric motor 36Ο, which drives a flexible steel belt 362 with the aid of a chain wheel 364. Like on can best be seen in Fig. 2, the steel belt 362 is perforated, and the sprocket 364 has corresponding protrusions around one Ensure positive drive. The flexible steel belt 362 runs around an idle wheel 3β5, around the chain wheel 352, an idle one Wheel 366, a spring loaded wheel 367, an idle wheel 368, the chain wheel 354 and an idle wheel 369. The Sprockets 352 and 354 also have corresponding protrusions that engage the holes in the steel belt 362 for a Ensure positive drive.

Individually controllable pneumatic brakes 370 and 372 are attached to opposite ends of the base plate 350 and capture the Brensstangen 312 and 33 * 0. Each brake has a pair of plates on opposite sides of the corresponding one Brake rods are attached. The rod of a pneumatic cylinder runs between the two brake rods around the press the two brake disks onto opposite sides of the rods when the cylinder is actuated. A set of counterweights 374 is connected to each end of the upper frame 32 by a chain 376 which is connected by a shoulder with a * See sleeve 377 (Fig. 20) up through the set of counterweights and over a sprocket 378 attached to the support is stored, up to a bracket 379 on the upper frame.

Two optical devices 38Ο and 382 are attached to opposite ends of the front edge of the base plate 350 for the exact

00982370691 _ÖAn " **

A 36 810 b

October 11, 1968

iy - 121 21 1803947

Determine the location of the ends of the frame relative to the calibrated rods 310 and 328. The upper frame 32 can be adjusted to any desired height above the table 26 with the aid of the motor 36O until one end has reached the selected height, which is determined, for example, by the optical device 38O. The brake 370 is then applied in order to lock this end securely at the desired height. The motor 360 then receives a pulse λ in the appropriate direction until the other end is at the desired level, whereupon the other brake 372 is applied. The play found in even the best ball screw assemblies is sufficient to allow the opposite ends of the frame to be individually adjusted to the desired height within the required tolerance.

The base plate 350 of the upper frame 32 can also accommodate two downwardly directed supports 120 in the positions indicated by dashed lines in FIG. 21. Each of the beams 120 is adjusted precisely in the X and Y directions by the same adjustment system as described above in connection with FIGS set, with three rotatably mounted lugs 390, 391 and and a spring-loaded pivotable lever 39 ¹ I are provided. The supports 120 are arranged so that the reference surfaces 176 on the base plate 122 look downwards, the reference surfaces 176 normally resting on four upstanding stops 395, which are shown in detail in FIG. The carriers 120 can be raised from the stops 395 on rails 418 and 420 by a hand-operated mechanism, which is shown enlarged in section in FIG. 24. Two shafts 400 and 402 are rotatably supported in the base plate 350. Shaft 400 carries four cams 404 and shaft 402 carries four cams 406. Each of cams 404 and 406 is practically the shape shown in FIG. The rails 418 and 420 are supported by pushrods 422 and 424 which extend downward and onto the cams

009823/0691

BATH ORIGINAL

1803247

A 36 810 b - -atf -

October 11, 1968

and ¹ JOo sit on. Two connecting rods ¹ IOB and 410 are rotatably connected to the cam 404 and, as shown in Fig. 24 4O6 ¹ IOO and positioned 402 adjacent to the ends of the shafts connected by eccentrically arranged pivot pin 412, so that both shafts in Rotate synchronously. A hand lever ¹ II ¹ I is ^{attached to the shaft 1} 100 on the left side of the frame, and a hand lever 416 is attached to the shaft 402 on the right side of the frame to rotate the levers and the shafts, respectively. Thus, when the hand lever ¹ IIo is operated, the rails can be lifted 4l8 and 420 to the carrier 120 withdraw from the stops 395 can be moved after which the carrier on the rails 418 and 420th

The upper frame 32 has two identical lamp housings 36 and 38 on. Each lamp house has nine common light sources, each of which is an electric lamp and a suitable lens and may have a filter if necessary. It can be a forced cooling system that works with air Cooling the light sources provided. Each light source is individually controlled by the computer that follows is described. If necessary, a suitable system can be provided to keep the intensity of each light source constant to keep.

The lower stage 34, shown in detail in Figures 26 and 27, carries the image reduction lens of the projection system. The lower stage 3'I has a base plate 460 which is practical is designed in the same way as the base plate 35Ο of the upper frame 32. The base plate 460 is chain wheels 462 and 464 carried on the stationary ball screws 308 and are screwed on. The lower frame is held in place by linear ball bearings 466 and 468 that surround the guide rods and 322 are arranged. The upper step 34 can be raised and

009823/0691 - 25 -

^ß AD ORIGINAL

1803S47

A 36 810 b

11 October 1--68

Iy - 124

be lowered with the aid of a drive system identical to that used for raising and lowering of the upper frame 32 is used. The drive system includes a drive sprocket 470 which is not through a The electric motor shown is operated, a perforated flexible steel belt 472 and idle wheels 473 to 477. The ones used to move The power required for propulsion of the lower frame is adjusted to one desired by two sets of counterweights 478 Value reduced, these counterweights being connected to the opposite ends of the rack by chains 479 that run over Sprockets 481 run, which are mounted in the supports and which are coupled to the frame by brackets 483. The lower Frame 34 also has pneumatically operated brakes 480 and 482 at each end that have brake rods 312 and 330 therein Detect in the same manner as for brakes 370 and 372, as described above in connection with upper stage 32. The location from each end of the lower stage 3 ^ is also by optical Devices 484 and 486 determined that allow precise alignment with ■ the calibration marks on the rods 310 and 328.

The base plate 460 is provided with eighteen sleeves 488 which serve to accommodate the lens mounts 490. As shown in Fig. 27, each lens barrel 490 can have any desired lens arrangement take up. A typical twenty power arrangement is shown at 492, a typical tenfold arrangement Power at 494 and a typical arrangement with fivefold Performance at 496. All lens arrangements that are simultaneous are used, of course, have the same performance, i.e. reducing and enlarging ability, respectively, the Lens systems with different performance have only been shown for the sake of explanation.

Each lens mount 490 has a rib 493 which is pressed by a leaf spring 500 fiddling the ^ Iimn ^ nvana Λβώ sleeve 493, whereby

^BATHROOM ORIGINAL "7ft"

A 36 810 b -

October 11, 1968

the leaf spring is arranged on the opposite side of the socket. The axial alignment of the lens barrel 490 can be adjusted by two sets of clamping screws 502 and 504. As can best be seen from the right-hand part of FIG. 27, the upper clamping screw 502 runs through a collar 506 on the socket 490, and it is screwed through a ring 508. The lower ends of the clamping screws 502 abut ball bearings 510, which are arranged in bores in the upper ends of the sleeves 488. The clamping screws 504 are screwed into the clamping screws 502 and press the springs 512 against the collar 506. The collar 506 is supported on the ring 508 by two ball bearings 514 which are shown in connection with the central lens in FIG. The ball bearings 514 ^N lie in an axis which forms an angle of 90 ° to the axis of the ball bearings 510, and these two axes form an angle of 45 ° to the axis formed by the rib 498 and the leaf spring 500. The lens mounts can thus be adjusted so that the axis of each lens is aligned with the optical axis of the projected light.

Each bellows 40 depends from the base 350 of the upper frame 32. The lower ends of nine bellows 40 are connected to a plate 501, and the upper ends of the other nine bellows 40 are connected to a plate 503. Plate 501 has nine downwardly depending tubular sleeves 505 and plate 503 has nine like downwardly depending sleeves 507. Panels 501 and 503 can be raised or lowered simultaneously by a winch, shafts 511, 513 and 515, which are coupled to one another by suitable gears and which are rotatably mounted in the base plate 350. An electric motor 509 drives the shaft 511 and thus the shafts 513 and 515 . Flexible ropes 517 and 519 are cornered

- 27 009823/0691 «_

A 36 810 b

October 11, 1960 Iy - 12 * 1

of plates 501 and 503 connected accordingly and around the shafts 511 and 515 wrapped. When the motor 509 is in one direction of rotation runs, the shafts 511 and 515 are rotated in the direction that the flexible ropes 517 and 519 are wound up and the plates 501 and 503 are raised. The engine 509 runs in the other

the Tlattt-n 501, 503 lowered.

f ..

·. "Ii

" ¹ ^« r

A 36 8I0 b
Iy - 137
lo. October 1968 1803947

The position of the table 26 in the X and Y directions is measured by sensing and counting the interference fringes of a laser interferometer which are produced by the movement of the table from a reference position. The interferometer consists of a helium-neon laser 516 which directs a light beam 517a onto a first partially silver-plated mirror, as shown in Pig. 28. Part of the beam 517a travels along the path 519a to a second partially silver-plated mirror 52o. Part of the light beam that strikes the mirror 52o is reflected along the path 523 onto an adjustable mirror 52 * 1 which is mounted on the table 26. The light is reflected from the mirror 52 ¹ I practically along the path 523 to the mirror 52 ¹ I, with part of the light continuing along the path 525. The other part of the light of the beam 519a which strikes the partially silver-plated mirror 52o continues its path along the path 526 to a stationary mirror 527, where it is reflected back onto the partially silver-plated mirror 52o. Some of the light returned by mirror 527 will be along the path

525 and a portion returns generally along path 519a. The mirror 527 is exactly perpendicular to the beam

526, while the mirror 52 ^ is set at an angle slightly deviating from 9o °. This combines the portion of the coherent light projected along beam 525 from mirror 527 with the coherent light from mirror 52Ί to produce interference fringes that have a practically sinusoidal change in illuminance along the axis shown in FIG Direction in which the mirror 52 ¹ I is inclined. As table 26 moves in either a positive or negative direction along the X-axis, ie away from or towards mirror 52o, the level of illuminance at any point in the interference pattern goes through a complete cycle each time the Tis: h moves by half a wavelength.

"29 009 123/0691

A 36 8lo b
• ly-137
lo.October JL968 r> q. 1803947

A portion of the light beam 517a becomes a third partially silvered or semi-transparent mirror along the path 521 522 is reflected, where it is again divided and a part is reflected back along the path 528 to a mirror 529, which is attached to the table 26, while the other part runs along the path 53o to a stationary mirror 532. That of that Mirror 529 reflected light that passed through mirror 522 and the light reflected back from mirror 532 that came from | reflected by mirror 522 are combined to create interference fringes in beam 534, so that the movement of the table 26 along the Y-axis shows a change in the illuminance at any fixed point determined by the Fringe pattern is illuminated over a full sinusoidal cycle every time the table moves by half a wavelength.

The interference pattern in beam 525 is ^{pulled apart by an optical system 5 1} Jo consisting of lenses 5 * 12 and 544. Two mirrors 546, at an angle to the beam and spaced along the beam, split the beam into two beams, spaced apart and each directed onto a photodiode 548. The beam 53 ² I is drawn apart and separated by an identical system, with the separated beams being directed onto two diodes 552. The outputs of the diodes 548 are given to an amplifier / X-stripe detector 554 and the outputs from the diodes 552 are given to an identical amplifier and Y-stripe detector 556.

The photodiodes 548 and the amplifier and detector 554 are in FIG Fig. 29 shown in detail. The outputs of photodiodes 548a and 548b are connected to the inputs of amplifiers 560a and 560b The output of the amplifier 560a is connected to an input of a high-speed comparator I and connected to an input of a

BATH ORIGINAL

A 36 8I0 b
Iy - 137 '
lo. October 1968 1803947

the. The output of amplifier 560b is similarly connected to an input of a high speed comparator III and to an input of a high speed comparator IV. A first reference voltage T ₁ is connected to the reference inputs of the comparators I and III and a second reference voltage T ₂ is connected to the reference inputs of the comparators II and IV.

The outputs of the photodiodes 548a and 548b and thus the outputs of the amplifiers 560a and 560b form a sinusoidal signal, as shown in FIG. The output from each of the comparators I-IV is a digital yes. Level when the voltage of the corresponding amplifier 560 a or 560b, the threshold reference voltage or T. Tp fiv _ _he progresses and it is a digital No- level when the threshold reference voltage exceeds the voltage signal of the corresponding amplifier. The output voltages of amplifiers 560a and 560b are indicated by curves 56oa and 560b in FIG. The inclination of mirror 524 or 529 on table 26 is adjusted so that the output from amplifier 560b lags the output of amplifier 560a by 90 degrees. The threshold voltages T, and Tp are set so that the outputs of the comparators I - IV have the form shown in Fig.3o, where each strip cycle is divided into 8 equal parts, each of the 8 parts being clearly identifiable by the 4 digital outputs is. Since the light from a helium-neon laser has a wavelength of about 62o millimicrons (24.8 micro "), a table movement of about 39-4o millimicrons (1.55 micro") can be measured by measuring the change in the digital outputs of the 4 comparators is measured. Furthermore, the direction of movement of the table can be determined from the digital outputs of the 4 channels using standard decoding techniques. ''

009823/0691 -31-

A 36 8lo b
Iy - 137
lo. October 1968 1803847

A decoder 562 for the position and the direction in the X-axis and a decoder 564 for the position and the direction in the Y-axis each generate a pulse for a respective movement of the table 26 by approximately 38 millimicrons (1.5 micro " ) in the X and Y directions, and they also generate signals which indicate whether the table in the positive or negative X and Y directions to move. the pulse signals of the decoder 562, and accordingly to be 56Ί fringe counter 566 and 568 for the X and Y directions, each of which has 2k bits. The stripe counters 566 and 568 increase by one count for each pulse coming from the respective decoders when the table moves in the positive coordinate direction, and they decrease by one count when moving the table in the negative coordinate direction, as indicated by the direction signal from the respective decoders 562 and 56 ¹ I.

The count of the strip counters 566 and 568 can be read accordingly by a digital computer 57o via an X-axis connection and control circuit 572 and a Y-axis connection and control circuit 57 ¹ J, and also via input and output networks 576 and 578. Comparators 580 and 582 for counting stripes in the X and Y directions with 12 bits each can be set by the computer via the control circuits 572 and 57 * f 582, X and Y comparator interrupts are generated and applied to the priority interrupt unit 588 of the computer 57o via channels 584 and 586. For each time that the coaparator is set by the computer, only one interrupt signal can be generated in order to prevent the generation of false interrupt signals due to the noise of the counters.

Di · O Besugsstellung of the table 26 in the X and Y directions wtrd "n by electronic calibration systems 585 and $ 87 from

00 ** 23/0691

. . ■ '. BAt)

A 36 8lo b

Brown & Sharpe eaten. The electronic gauges generate a semi-rectified Io KHz signal representing a DC level is embossed, which varies from -o.2 volts to + o.2 volts for a table position from -75o to + 75o Millikron (-30 to + 3o micro ") relative to the zero reference point. The outputs of gauges 585 and 587 v / ground first amplified by amplifiers 589 and 591 and then on X and Y detectors 593 and 595 given. These origin detectors each have a high-speed comparator, which converts the analog signal to a digital signal, and zv / ei individual recording circuits that generate an interrupt signal when the table has the zero reference point in any Direction Exceeds. The zero point interrupt signals in the X and Y directions are on the priority interrupt unit 588 of the computer on channels 597 and 599 The zero point detectors also have a flip-flop circuit, which is set by the output of the individual recording circuits and is reset to generate a status signal indicating whether the table is on the positive or negative side of the zero reference position, and these status signals are on the control circuits 572 and 574 and to the input network 576 given.

The wavelength of light in air changes with pressure, temperature and humidity. To set the table with the desired To position tolerance, the camera Io should be set up in a room where the temperature is kept practically constant is, for example, within + o.3 ° Celsius (♦ o.5 ° F), and the humidity should be kept within a certain range. As it is inconvenient, the barometric pressure To regulate the pressure within the room, a barometer 600 is used in space and this pressure is made digital by an encoder 6c ~ 2 coded. The digital information is continuously available to the computer 57o via a control circuit 6o4 and the network 576 to disposal · . .

nnqtr3 / 0891 - 33 -

BAD ORfGJNAL

A 36 8lo b *

Iy -129

lo. Oct 1968

The motors 86 and 98 for the drive in the X and Y directions exert a force which is proportional to the voltage which is applied to the amplifiers 6o3 and 605 by a digital-to-analog converter 6o7. The output of the converter for controlling the drive motor in the X and Y directions is set by the computer 57o f via a connection circuit 578 and a control circuit 6o9 for the converter. The exposure lamps * J3o in the lamp housings, which are built into the upper frame, are individually controllable by a light relay register 6I0, which in turn is controlled by the computer 57o via the connection circuit 578 and the relay register control 612. 31 shows the geometric range of motion of the table 26 relative to the granite block 18 and thus relative to the projection system. The lines Y _Q and X _Q indicate the position of the table 26 at which the zero detectors emit zero point interrupt signals for the X and Y directions to the computer. The lines X ₁ . and

L

Y- indicate the positions at which the X and Y state lines g change from a positive to a negative value, and they are about 75o millimicrons (3o microinches) from the X ₀ and Y _Q lines. Lines X _g and Y _g indicate the points at which the table encounters mechanical stops to limit the movement of the table, actuating limit switches which automatically shut down the operation of the system. Limit switches and mechanical stops (not shown) are provided at the other two limits of the range of motion.

The computer 57 ° is programmed as in the basic ones 32a and 32b and shown in the subroutines of Figs. 33-41. If the computer is one of the receives four interrupt signals, as / Fig. 28 is shown

0 09823/0 6 91 _ ₃ i | ..

BAD ORKäiNAL

A 36 8lo b

Iy - 129

lo. Oct 1968

he immediately switches to a predetermined subroutine and returns then return to the starting point of the program when the subroutine is complete. All other inputs on the computer have the shape. of state lines drawn by the computer on a particular Point of the program must be read. The first time it is started up, the computer 57o cycles through a test subroutine 7oo, during which it carries out various self-monitoring to ensure that it is working properly. After that, all interruptions, with the exception of the lack of energy, are made ineffective at 7o2. After that, the typewriter interrupts built up at 7o4 and the codes and words introduced at 7o6.

The computer then operates the typewriter for the numeric display at 7o8 to indicate the program to be followed, and then at 71o it prints out the instruction to insert the format punch tape. The format tape is then read in the computer by the tape reader and stored at 712. The computer then cycles through the test subroutine at 71 ¹ Y and tests all of the status lines at 716. If the EXIT flag is set at 718, when the EXIT button is manually pressed, the computer exits into an EXIT subroutine 72o to exit the program. If the EXIT flag is not set, the status of the HOME ^ button is checked at 722. If the HOME button (HOME = home position) has not been pressed, the program returns to the test subroutine 71 ¹ *. If the HOME button has been pressed manually, the computer switches through the HOME

shown in subroutine 726 detailed in FIG. At the beginning of the HOME subroutine shown in Fig. 35 can be the table be in any position within the field that passes through

009823/0691

1803S47

A 36 8lo b

Iy - 129

Io. Oct 1968

the stops Χ _β1 , X _g2 , Y ₃₁ and Y _{g2 is} limited, so that the limit switches X, and Y. can either be switched on or off. The table is therefore first moved into the first or positive quadrant. The state of limit switch X is thus checked at 728. When the limit switch is switched on, the speed of the table is compared with Ιοί the target speed by a subroutine 73o for measuring the speed. In subroutine 73o, the X counter is read, a predetermined number of microseconds waited and the counter read again. If the difference in the readings is greater than a predetermined number, the speed in the X direction is greater than Ιοί the design speed. If the speed is greater than + Ιοί the design speed at 712, the state of the limit switch X ^ is checked again at 728. If, however, the speed in the X direction is less than Ιοί the design speed or negative, the motor for the drive in the X direction is set to + Ιοοί the maximum target force at 73 ¹ I, whereupon the computer sets a predetermined number of microseconds at 736 waits, whereupon J the motor is zeroed at 738. The state of the limit switch, X, is then checked again at 728. When limit switch X. is off, indicating that the table is in positive X range, the X motor is zeroed at 7 ^ o.

The computer then switches to check the state of limit switch Y _T at 7 * 12. If the limit switch Y. is switched on, the Y-speed is measured at 7 ¹ IH and compared with + Ιοί the design speed at 7M6 . If the speed is greater than + Ιοί, the state of the limit switch Y _{r is} checked again at 7 * 12. If the! Speed

ti

is not greater than + Ιοί or negative, the Y-motor is on

O09823 / Q69T - 36 - "

BATH ORIGINAL

1803847

Α-3β 81ο fc

Iy - 129

Ιο. Oct. 19-8

+ loo? the force is set at 7 ^ 8 _s where the computer waits a predetermined number of microseconds at 75 © _s whereupon the Y-motor is set to zero at 752 before the state of the limit switch Y ^ is checked again at 742. If the limit switch Y _{x is} switched off, the motor for the drive in the direction of the Y-axis is set to zero at 744. The table is then in the positive quadrant delimited by the axes X- and Y. in FIG.

The state of the signal from the limit switch X _L is checked again at. When the limit switch X. is off, which is normally the case, the X-speed is measured at 758 and entered with -Ιο? compared to the design speed at 760. If the X speed is not less than -Ιοί, the X motor is set to -loo # at 762. The computer then waits a predetermined number of microseconds at 764 and zeros the motor to drive in the X direction bed 766. The computer then switches to the state of limit switch Y ₁ . to check at 768. On the other hand, if the limit _{switch X L is switched} on at 756, the X-motor is set to zero at 77o before switching to the state of the limit switch Yj. to check at 768. If the X-speed is already less than -lo £ at 760, the computer also switches on to check the state of limit switch Y at 768 without giving the X-motor an impulse for a force of -loo? . If the limit switch Y _Q is not switched on at 768, then the Y speed is measured in the subroutine 772 and with; Ίο% compared at 77 ^. If the Y rate is not ifeniger as -lo% of design speed, so aar Y motor is set at a force of at -loo £., The computer waits a predetermined period of time at 778 and sets the Y-Piotor at 780 to again NuIl ₅ before he returns. * to measure the addition of the limit switch X _T at "56.

00982 3/069 1 ^L

^BATH ORIGINAL

λ ^ C Οίο h Iy ■ - 129
Oct 7, 1968

Vsenn die! '- speed less than -lo? at 77 *, the Y-motor receives no pulse. For this purpose, the state of the Gronz switch X _L is checked at 756. This routine, in which the state of the limit switch X, is checked at 756, in which the X-motor is also pulsed if necessary to maintain the speed at which the state of the limit switch Yq is checked at 768 and in which the Y-motor receives a pulse if this is necessary to maintain the speed, it is repeated until the limit switch Y _{Q is} switched on. Then the Y motor is zeroed at 782 and the state of the limit switch _{X Q} is checked again at JQk. If the limit switch X ₀ is not yet on, the X speed is measured again at 786 and -lo $ 5 at the design speed 783 compared. If the X speed is not less than -Ιοί, will the X motor revert to -loo? Set at 79o, the computer waits a predetermined amount of time at 792 and then ^{zeros the X motor at 79 1} J before returning to check the state of the X _Q limit switch at 784. If the X-speed is less than -lo £ the design speed at λ 788, the X-motor receives no pulse. To do this, the state of the limit switch X _Q is immediately checked again at 784. If the limit switch X _{Q has been} turned on, the brakes of the X-motor and the Y-motor are set and the computer returns to the main program shown in FIG. 32a and writes out the information "specified format" at 796.

The state lines are then checked at 798 and if the EXIT button has been pressed at 800, the EXIT routine 72o follows, to bring the system into a standby state. If the EXIT button has not been pressed, all interruptions are canceled withheld at 8o2 except for the interruption

gets by in the absence of energy. A style number with four digits / typewriter entered at 84o in ^e 8omputer what

009823/0691

BATH ORIGINAL

A 3Ci 8lo

Iy - 129

insert the interruptions again at 8οβ «. The four-digit format number is checked at 808 to determine if si® is legal. If it is not legal, the number is checked by Bio ₉ to determine whether it represents a special format. If the number does not represent a special format, the typewriter writes an error at 812 and returns to the indication © "special format" at 796 »menu, however, is a special number, the special identifier is set at 8ΐή, whereupon the computer switches on to write out the information "ready to run" at SlS ₀ If the number is legal, the format specification is converted to the appropriate address M, N, XSPAC, YSPAC in the memory and the information "ready to run" ( ready to run).

The computer then switches on through the test subroutine 818 and checks the state of the run button at 82o. When the run button is not depressed, the state of the special token is checked at 822. If the special mark is not is set, the program returns to the test subroutine 818 and this cycle is repeated until the run key is pressed. If the special flag is set, the status of the highlight key is checked at 824. If the highlight button is not depressed, the program returns to test subroutine 818 until the highlight button is depressed is.

If the scroll button has been pressed in normal format, or if the highlight key has been pressed for a special format, the computer increments the run number command (run numbe: tally) and writes out the run number at 826. The barometric pressure measured by the barometer 600 is read and written out at 828 before being redirected to

009823/0691 _ ₃₉ >

BATH ORIGINAL

1803847

A 36 8lo b

Iy - 129

lo. Oct 1968

the table by the subroutine 830, which is shown in Fig. 3 * 1 in detail is shown in the initial position to bring.

The HQME subroutine 726 has set the table approximately near the zero point calibrations 585 and 587, represented in FIG. 33 by X _Q and Y ₀ , with the drive motor brakes 96 and Ho applied. The table is initialized by the subroutine shown in Fig. 3 ^ by first releasing the brakes on the X-motor and Y-motor at 832. This requires about a second while the computer waits Then the X _Q break is established at 834 and the status of the X _Q calibration is checked at 836. If the status of the X _Q calibration is positive, indicating that the table is to the right of the line X ₀ in FIG 33, the X motor is set to -2οί the design force at 838 and the X speed is measured at 84o and compared to -I0J6 the design speed at 842. If the X speed is less than -Ιοί the design speed, then the X-motor zeroed at 814 and the speed measured again at 84o. However, if the X-speed is not less than -Ιοί the design speed at 842, the X-motor will again be -2οί at 838 set, eh e the speed is measured at 84o. If the state of the X _Q calibration at 836 is not positive, the same procedure is repeated, except that the X-Kotor is set to 4 2οί the design force at 846 and the speed is measured at 848 and measured with lo? the design speed at 850 is compared. If the X speed is greater than • Ho% , the X motor is set to NuU at 852. If the X speed is not greater than + Ιοί, the X motor is kept at 42οϊ.

In any case, the table is referenced to the zero reference point X _n

009823/0681 °

- UO BAD ORIQiNAL

A 36 8lo b Iy - 129 Io. Oct 1968

moved about Ιοί the design speed too. When the table moves past the zero point X _{Q in} either a positive or negative direction, the X interrupt occurs by detector 593 and it is passed to priority interrupt unit 588 of computer 57o. The X-axis strip counter 566 is reset immediately at 85U upon receipt of the X interrupt signal. As explained above, all bits of the counter can be reset, except for the

characteristic. . _ .. ..,,. . _Grandpa . . last three bits read and stored at 856

will. The direction of travel of the table in the X direction is determined at 858. If the X-speed is positive, the X-Mptor is set to -5o £ the design force at 860 to stop the table. The X-speed is then measured at 862 and the measured speed compared to + 2JS of the design speed at 862. If the X speed is less than + 2% of the design speed, the X motor is zeroed at 866. If the X speed is not less than + 2JS the design speed, the speed is measured again at 862 and the cycle is repeated until the X speed is less than + 2% of the design speed. On the other hand, if the X speed at 858 is negative, the X motor is set to + 5055 the design force at 868 and the speed is measured at 870 and compared to -2% of the design speed at 872. If the X-speed is less than -2% of the design speed, the speed is measured again at 870. However, if the X-speed is less than -2% of the draft. ^ Speed at 872, the X-motor is turned on at 866 aiii * Hull. When the X-speed has reached a value of less than 2% of the design speed, it goes through 4i · ;; Frictions quickly return to gauze when there is no power from the X-motor; exercised wiroL

nu 9 g 2 3/069 1. ■; -M-

BATH ORIGINAL

18C3947

A 36 8lo b,

Iy - 129

Io. Oct 1968

Thereafter, the Y interruption or the Y zero point interruption is established at 874. The same process is then repeated to reset the Y-axis strip counter 568. First, at 876, the condition of the Y-zero calibration is determined. When the table is in a positive Y position, the Y motor is set to -2o £ to slowly move the table in the negative direction with about lo? the design speed until the Y _Q interrupt is effected by the Y origin detector. If the table assumes a negative position relative to the zero point axis Y _Q , the Y-motor is set to + 2o £ and the table is moved at approximately + lo £ the design speed until the Y _Q interruption occurs. If the YQ interrupt occurs, the Y counter is reset at 878 and the last three flagging bits at 880 are read and stored.

Thereafter, the table is brought to a stop in the Y direction by first determining at 882 whether the Y speed is positive. If the Y-speed is positive, the Y-motor is set to ~ 50% at 884 until the Y-speed has dropped to less than + 2% of the design speed. Thereafter, the Y motor is zeroed at 886 to stop the table. On the other hand, if the Y-speed is negative at 882, the Y-motor is ramped up to + 5o? the force is set at 888 and this force is maintained until the Y-speed becomes greater than -2% of the design speed. The Y motor is then set to zero at 886. The strip counters for the X and Y directions are then reset, with which the table has been brought into the initial position and the computer returns from the subroutine of Fig. 3 ^ to the main program of Fig. 32a when the brakes in X- and Y-direction are attracted.

-42-

^009823/0681 BAD0B1Q1NAU

A 36 8lo b

Iy - 129

Io. Oct 1968

The special license plate is then checked at 89 ©. If no special format has been specified, the format number is written out at 892. Then write out the width of the row (X-axis) in inches (1 inch = 25 * »mm) and the height of the row (Y-axis) in inches at 89 ² I, whereupon the distance between the centers of the exposures is in follows the X and Y directions at 896. Thereafter, the format is calculated at 898, at which point the number of exposures required, the number of rows and columns required, and the location of the blank exposures in which the marks for alignment will be sequentially exposed are determined. The computer then writes out the number of exposures at 900 and calculates the first exposures at 902. The brakes are released at 9o4 and the EXPOD (Cobol counter) exposure command is set to -2% since an exposure is made in the HOME position before switching to the row, followed by a final exposure in the HOME position is made after all exposures in the series have been taken so that an offset between the two HOME exposures indicates the cumulative error of the run. Then the values XOLD, YOLD, XNEW and YNEW are set to zero at 9o7 to 9I0. XOLD and YOLD are the coordinates of the last exposure, given in milli-inches (1 inch = 25.4 mm) from the HOME setting (which is given by X _Q and Y _Q ), while XNEW and YNEW are the coordinates of the next exposure, expressed in milli-inches from the HOME setting. XTAL is then set to "zero end YTAL is set to" one "to define the HOME exposure position.

The computer then goes to subroutine 912 for the strip number to calculate, to position the table and to carry out the exposure, this subroutine in Fig. 35 in detail is shown. According to FIG. 35, the state lines and the

00 9 82 3/0691 -43-

BATH

A 36 8lo b # * »

Iy - 129 ^T

Io. Oct 1968

License plate checked at 91 ¹ »and 916 and then XAIM and YAIM at 918 calculated. Since XAIM and YAIM represent the number of interference fringes from the zero point in the X and Y directions, the barometric pressure read at 828 is included in the calculation. For the first exposure in the HOME position, XAIM and YAIM are the numbers read and stored from the last three digits of the counters. Then XNEW is compared to % XOLD at 92o. Since XNEW and XOLD are both set to zero, they are both equal to the fact that the program proceeds directly to point POSY along line 922. YNEW is then compared to YOLD at 926. Since YNEW and YOLD are both set to zero, the program continues directly with positioning during exposure subroutine 93o along line 928. Subroutine 93o is described in detail below. ·

If on the other hand at 92o XNEW larger than XOLD, which is normally the case for all exposures except the first HOME-BeIichtung, the computer goes to the subroutine 932 shown in g Fig. 36, positive to the table in the X direction to move. First, the distance the table is to be moved is ^{compared at 93 1} * with the minimum practical distance to move the table at high speed. If this distance is less than the minimum practical high speed distance, the computer enters subroutine 936, shown in detail in Figure 37 and labeled SLOW-X. In the SLOVi-X subroutine, the XrComparator 580 is fed with XAIM and the X counter 566 is read at 938, whereupon XAIM is compared with the current X coordinate at 9'Io.

009823/0 (191

BATH ORIGINAL

1S03947

A 36 8lo b Ψ * -JMf "-

Iy - 129

lo. Oct 1968

When XAIM is greater than the current X coordinate »becomes the X-motor is set to + 2ojf the target force at 9 ^ 2 to the Move the table very slowly in the positive X-direction. Then the X-speed is measured at 9M and the measured X-speed compared to Ιοί the design speed at 9 ^ 6. When the X speed is equal to or less than Ιοί is the design speed, the X motor turns on at 9 ^ 8 O-force set, but if the table is not yet + lo? the design speed, the X-motor is set again to ♦ 2oX at 9 ^ 2 and the process is repeated at the X-speed towards about + lojt of the design speed keep. On the other hand, if XAIM is less than the current X coordinate, the X motor is set 7-2o £ at 95o, to move the table in the negative X direction. Then the X-speed is measured at 951 and the measured speed with -Ιο? compared to design speed at 952. When the table -lojt has reached the design speed, the X-motor is set to 0 at 95 ^ and the speed measured again at 95o to repeat the cycle. But if the table lo? has not yet reached or below lo at 952? is the X-motor on -2o? set at 95o and the cycle repeated. The table is thus with about -lo? the design speed towards XAIM.

When the table has reached XAIM, the X comparator 580 generates the X interrupt signal which immediately switches the computer to the SLOW-X subroutine 956 which first determines at 958 whether the X speed is positive or negative. If the X-speed is positive, does the X-motor go to -50? the design force set at 960 and the X-speed measured at 962 and compared to + _ 2% of the design speed at 96k . If the speed is not less than 2% , it is measured again. When the X speed

009823-0691.

ÖAD ORIGINAL

A 36 8lo b, ".

Iy - 129 T>

lo. Oct 1968

is less than 2% , the X motor is set to 0 at 966. On the other hand, if the X-speed is negative, the X-motor is set to + 50 * at 968 and the speed is measured at 97o and the measured speed is compared to -2Jf the design speed at 972. If the speed has not slowed to a value greater than -25i, the X-speed is measured again at 97o. When the X- (| speed has slowed to a value greater than -2% , the X motor is set to 0 at 966 and the computer returns to the point POSY in FIG.

In Fig. 36, if the distance the table is to be moved is greater than the minimum practical distance for high speed movement at 93 * 1 practical minimum distance, then X comparator 580 at 968 is inputted XAIM minus the braking distance necessary to stop the table when it is moving at lo2Jf the design speed. A comparator interrupt occurs approximately every 0.162 mm (o, oo64 ^w ) due to the comparator capacity of 12 bits. Since the piece to be traversed is of the order of about 2.5 to 5.0 mm ^{(0.1 to 0.2 If} ), the number of X interruptions that must be ignored (XIGNR) is 97o If the X interrupt occurs, the program goes to the X interrupt subroutine at 972. The positive X panel light is then turned on at 97 to indicate that the table is in the positive X direction moves and the X-motor is set at 976 to the maximum acceleration force of + looji.

The X-speed is measured at 978 and the measured speed is compared to 99% of the design speed at 980. If the X-speed has not yet exceeded 993 »the design speed, the speed is

- l \ 6 009 8 2 3/0691

BATH ORIGINAL

18Ö3947

A % 8lo b

Iy - 129

Io. Oct 1968

measured again at 978 and compared at 980 until the X-See speed exceeds 99Sf the design speed »Then the X-MötQ3? set to O at 982. The program switches üann in a pulse subroutine 984 'in which the digital-to-analog converter ■ it the desired power is fed in a & r further predetermined waiting time, the converter is supplied with / force a post-, followed by a further predetermined waiting time ansehliessfe

The! Speed is then measured at 986 and the time that has elapsed since the last! -Underroufcine. 930 "which shuts down the operation of the device and the Bremsen'anzieht vtm to stop the table. If the time since the last interruption was not too long, the X-speed set at 986 is compared with the selected tolerances, eg with ± 2% of the design speed. If the speed is within tolerance, the X motor is zeroed at 982 for the next pulse subroutine 984. If the! Speed is not within tolerance, it is determined at 992 whether the speed is slow or fast. If the X speed is low, then at 994 it is determined whether the! Speed was low during the last pulse. If the speed was not low during the last pulse, the motor power is set to + 50% at 996 before returning to the pulse subroutine 984 via reference point XPR4. If the speed was low during the last pulse, the motor power is increased in the positive direction by a predetermined amount at 998 before returning to the pulse subroutine 984. If the X speed is low at 992, it is determined at looo whether the speed before the last pulse

009823/0691 - 47 -

BATH ORIGINAL

A 36 8lo b

Iy - 129

Io. Oct 1968

was great . If the X-speed was not great before the last pulse, the X- motor is set to -5ojt at Ioo2, but if the! predetermined value set more negative, namely 984 before returning to the pulse subroutine Soeit the table is practical with the similar design speed in the positive X direction moving. If the speed of the table falls below or exceeds the design speed by a specified tolerance, appropriate positive or negative pulses are emitted to bring the speed back within the tolerance limits and the size of successive pulses in the same direction is increased until the desired speed is reached.

When the table in the positive X direction is moved, an X interrupt occurs approximately every o, l62 mm (6 ¹ I mils). On each comparator interrupt, the computer exits subroutine 932 and switches to the X interrupt subroutine lolo as shown in FIG. In the subroutine lolo, which is shown in detail in FIG. 39I, it is first determined at Iol2 whether all additional interruptions (XIGNR) previously calculated at 97o have been ignored. If the number of interruptions to be ignored has not yet been reached, the register keeping track of the number of interruptions that have already occurred is increased at lol ^, the X comparator is fed with XAIM minus the braking distance at Iol6 and the subroutine 932 turned back on at the point the X comparator interrupt occurred.

At the first interruption after the correct number of interruptions has been ignored, the point at which a retarding force is applied to the table to make it

0 09823/0691 -48-

BAO ORIGINAL

A 36 8lo b

Iy - 129

Io. Oct 1968

hesitate, at I0I8 in the form of the number of interference fringes calculated from the HOME setting. This calculation is based on the speed of the table at the time of the calculation. The number of interrupts that must be ignored in order to get to the delay point is also calculated and stored. The X comparator 580 is then stored at Io2o with the calculated value XAIM minus XRETARD, whereupon the computer switches to the X interrupt subroutine Io22 when each comparator interrupt occurs. Then, when the next X interrupt occurs, the computer enters the X interrupt subroutine Io24, which is shown in detail in Fig. 42 and to which reference is made. Subroutine 932 to move the table in the positive X direction continues. However, every time a ROmparator interrupt occurs, the computer will now go into the 1ο2Ί subroutine rather than the lolo subroutine. For each interruption, the number of interruptions in the register holding path that have occurred is compared at Io26 with the number of interruptions to be ignored. If all of the interruptions to be ignored have not yet occurred, the register keeping track of the interruptions is increased and the X comparator at Io28 is fed again. After all of the new interrupts have been ignored, the X-motor is set to -Ι00Ϊ at Io3o after the next interrupt and the X-speed is measured repeatedly at Io32 and compared with + 2% of the design speed at I034 until the speed of the table falls below + 2% falls. The X motor is then set to 0 at I036, whereupon the computer switches to POSY in FIG.

If XNEW at 92o is not greater than XOLD, the computer goes to subroutine I038 to move the table in the negative X direction

.009123 / 0611 " ⁴⁹ *

BATH ORIGINAL

Ä 35 8lo b

Iy -129

Io . Oct 1968

move sy. The subroutine Io3?> Is identical to the sub-routine 932, except that the forces act in reverse to cause a vision movement in the negative direction. Every time sr receives an X interruption, the computer goes into the X- Unb ^ rbrcChungs subroutine lo ^ o that the Ünterroutine lolo is identical, except as regards the Bewegungsi ^ ichtung, specifically ™ Ti8 aind all necessary Kornpsirator-Unterbrechungftn occurs the computer then switches to the X-Unterbrechungd-sub. " i '^ utine 10 * 42, which is identical with the sub-Croatia Io2k , oyster hurries back in the direction of the delay force, and awar ΐ ": .- i.:-i receipt of successive comparator interruptions hang up until the table is longer than -2% of the design speed j η dfi>"has slowed near XAIM. From the point POSY from Iv'ihc the program in the part of the subroutine for moving the

rxsehes from YOLD to YNEW ₁ is identical / with the program on moving the table from XOLD to XNEW, the subroutines ioi>'5jlo5Os Io52 and Io5 ^ are identical to the subroutines 532> 1ο1ο ₄ 1ο24,1ο38,1ο4ο and lo ^ 2 except that they are for the moving | along the Y-axis.

- 50 .0 09823/0691

in .; ι, ϊ tv ¹

nil! "Π" Γ ~.! ' ti '"" ¹

Ί C

II "DIQ

J ¹ I * mJ1

⁷ I *, T

I 1 »

II »

τ ^r ϊ

ti J ¹

UL tier i ^ sc iv / inr) Jg ¹ Ze ^ c Ί " ¹ ^" tdb bexifCineG, immediately before it reaches the last point at which the force can be applied and the table can still be stopped at XAIM if it is Moved at the expected maximum speed, that is, at 102 % of the design speed.

If the table has been decelerated to less than 2% of the design speed in the X and Y directions, and the drive motors for the X and Y directions have been set to zero, the computer goes to the sequence shown in FIG .Al shown subroutine 93Q ″ for positioning during exposure. First, the "state lines and indicators at ΙΟβΟ and. 10.62 are checked. Then

00982 3/0691

-51-BATH ORiQiiMÄL

A 36 810 b

10.10.1968 £ j

checked the state line for the pause button at 1064. If the pause button is pressed, the program is delayed until the pause button is released. Then the positive and negative X and Y tolerances necessary to achieve a good exposure are set, whereupon the computer is set to use the center position of the table in calculating the necessary restoring force at IO66, if this is desired '. The status lines and the labels are checked again at IO66 and at 1070. If the item flag is not set at 1072, the computer turns on, around the X- and Y-Zäh- μ ler to read and store the numbers in the X and Y registers 107 at ¹ J. If position averages are used, the averages at 107 ** are also updated.

Then at IO76 XAIM is compared with XAVERAQE in order to determine the central position of the error E and a restoring force XFORC, which is calculated according to curve 1075 in FIG. 2. If XAIM is greater than XAVERAGE, a positive restoring force F is calculated at IO78 using the equation F = k (E) ⁿ + F _f , in which k is a constant, E is the absolute value of the difference between XAIM and XAVER-AGE, η a positive number and F _f the force necessary to overcome the friction. This force is represented by the part of the curve labeled 1075a. If XAIM is less than XAVERAGE f, the force is calculated in the same way at 1080, but with a negative sign. This force is represented by the part of the curve labeled 1075b. This is followed by the same process for the Y-axis by comparing XAIK with XAVSRAGE at 1082 and then calculating the positive or negative force at 1084 and IO86 that is necessary to move the table in the direction of YAIM accordingly. When the error E exceeds a predetermined maximum value of -E, the force is decreased to the constant value sufficient to overcome the friction F. to move the table at low speed regardless of the size of the error E This force is represented by the part labeled 1075c and 1075d

009123/0101 - 52 .- ..

V * '■ = ■

■ ^{x v} _BAD ORIGINAL

1 36 810 b
10.10.1968

Iy - 153 53

the curve shown «It is necessary., a sharply rising Provide force when the error increases to the positioned Keep the table within the desired tolerance «Usually the value of η should be increased if the response time of the Drive system increases to within a given tolerance positioned table. However, if the force is held at a high maximum value over long distances by XAIM, then the high restoring force would accelerate the table to such speeds that a relatively slow responding drive ™ system does not stop the table over a considerable distance after passing XAIM, each time the excess tends to increase, resulting in an unstable condition. if however the restoring force is reduced to a low value after the error has exceeded the selected limits, the table will never be accelerated to such a high speed that the restoring force in the areas 1075a and 1075b cannot reverse the course of the table before the limit is exceeded again ten and the restoring force has been reduced.

After the restoring forces have been calculated, at 1037 the X-motor is set to the X-force and the Y-motor is set to the Y-force represents. The position of the table on the X-axis is then compared with XAIM at 1088 ■ to determine whether the table is within the selected tolerance. If the table is out of tolerance, the register that records the number of errors is increased at 1090. If the table is within tolerance, then the position of the table on the Y-axis is compared to YAIM at 1092. If the table on the Y-axis is outside the tolerance, the error tally is increased to 1090. Lies the table on the Y-axis is within tolerance, the COBOL counter (tally) is incremented "within tolerance" at 1094 and this number is compared with a selected minimum number at IO96. If the tolerances have not hit the selected number, order indicates that the table has been stabilized within the desired tolerances at XAIM and YAIM, becomes the condition of the line

009823/0691 -53-

BATH ORIGINAL

A 3β 81.0 b
10.10.1968

"Exposure ended" checked at 1098. If the exposure has not ended and has not even started at that point, it will switch the computer returns to check the condition at 1068 and then it repeats the program.

After the tolerances are the number required to indicate that the table at XAIIl and YAIM; has stabilized, the error tally is initiated at 1100 and the space flag is checked at 1102. If the blank flag is set, the status line of the manual exposure button is checked at 1104. If this button is not pressed, the computer returns to check the condition at 1068 using the condition check 1098 for the completed exposure. However, if the blank flag is not set, or if the exposure button is depressed when the blank flag is set, the lamps 1106 are turned on and the status of the exposure ended is checked at 1098. The exposure positioning subroutine 930 continues to read the X and Y counters, calculate the restoring forces in the X uηi / direction necessary to that set to X-JCAIM and / ΑΙ. "! Hold table ~ u, compare the position of the table with XAIH. And YAI ₁ I to see if the table is within the desired tolerances and keep a tally of the times when the table is inside or outside the The duration of the exposure is controlled by a mechanical timer and depends on the intensity of the light source and the emulsion used, and it can vary from a few seconds to a small fraction of a second Exposure is complete, the X and Y motors are zeroed at 1108 and the number of errors calculated at 1110 before returning to the subroutine shown in Fig. 35 and there with in the main program shown in Fig. 32b.

- 54 -009823/0691

I> UIC L
10, IC. 19üi.

After the first exposure in the ΗΟίίΕ position, XTAL is "incremented" j at 1120 so that XTAL and ITAL are equal to 1 ₉ whereby the first exposure of the row and column shown in Pig, 35 is defined as "It becomes" a check is made to see if this is one of the four exceptions by examining the space flag at 1122. If it is not an exception, the exposure coordinates are in mils (1 inch = 25.4 mm) at 1124 compared to the HOI-IE setting, namely before switching at 1126 'into the subroutine shown in FIG. 34 and described above ,,

After exposure is complete, XTIL is compared to XUNIT at 1128 ', where XMIT is defined as the total number of exposures in each row or row. If IS ¹ AL is not equal to XUIIIT, then ITAL is increased by 1 at 1120. This' process is continued until all exposures in the row XTAL equal to "1 are completed, at which time XTAL is equal to UNIT = If the exposure defined by XTAL and YTAL is one of the four exceptions and is not exposed, the subroutines 1124 and 1125 omitted, Heim at 1128 XTAL equals XUHlT, indicating that all exposures in the series are completed, yTAL Eilt ΙΜΪΪ is compared at 1130. If YTAL is not equal to YUNIT, ITAL is incremented if i * I> £ iss 1 before examining the exception at 1134. Then the coordinates of the new exposure are calculated at 1136, before the exposure subroutine continues at 1138. If after exposure XI ¹ AL at HkQ is not equal to 1, becomes. XTAL at 1142 decreased by 1. The further exposures along the second row back are carried out in this way until XTAL at 1140 is finally equal to 1. If at 1134 any If exceptions are encountered, subroutines 1136 and 1138 are omitted. After the exposure, in which XTAL is equal to 1, YTAL is compared with YUHIT at 1146 and if the two values are not equal, ITAL is increased by 1 at 1144 before returning to checking the blank indicator at 1122 "

.0 09823/0691 _ '- 55 -

BATH ORIGINAL

A 36 810 b

10. 10, 1968 _r

Based on this program, the table is moved over the first row from left to right, then over the second row back, then over the third row, etc., until either XTAL = XUNIT at 1128 and YTAL = YUNIT at 1130 or XTAL = 1 at 1140 and YTAL - YUNIT are at 1146. Thereafter, the vacancy flag arrives at 1148 and the table is returned to the HOME position and exposed by subroutine 1150 which is virtually the same as the subroutine for calculating strip count, position and exposure shown in Fig. 35 except for _a XAIM and YAIH are just the stored values of X _Q and Y _Q. ™ After the second exposure in the HOiME position, any major cumulative error is indicated by the fact that the two HOME exposures do not match.

After the second HOME message, the computer writes out the "ready for exceptions" information at 1152 and waits for the run key at 1154 to be pressed. This gives the staff time to replace the mother transparencies used for exposing the row by the mother transparencies for the reference patterns, which are arranged at the four exceptional locations. Thereafter, the table is again brought to the initial position by the subroutine shown in FIG. 34 at 1156. Then M XTAL and YTAL are set at II58 to indicate the column and row of the first exposure and the coordinates of the first exception are calculated at II60 in mils (1 inch) from the HOMB position. Thereafter, at 1162, exposure is performed using the subroutine shown in Fig. 35 for the calculation of the number of stripes, the position and the exposure. At 1164, XTAL is then flagged for the second exception, the coordinates of the second exception are calculated at II66 and the exposure is carried out at II68. Then YTAL is set to the third exposure column at 1170, the coordinates are calculated at 1172, and the third exception exposure is performed at 1174. At 1176, XTAL is then set to the fourth exception and the coordinates are calculated at II78

009823/0 6SJ - 55 -

BATH

A 36 810 b

10.10.1968
iy - 133

and made the fourth exposure for the fiducial mark at II80. The space indicator is then reset at 1182 and the table ^{is returned to zero at 118 1} I before the information "run ended" is written out at II86. The program then returns to point cc of Figure 32a and begins the next programmed run, unless the EXIT flag is set.

If a special format is programmed, the special flag is set at 89O in Fig. 32a. The special format allows the table to be set to a number of coordinates that are specified on a punched tape. The first step of the routine for the special format is thus to switch on the strip scanner at 1200, to set the EPOD (Cobol counter) exposure command to zero at 1202 and to set XOLD = YOLD s 6.35 now (250 mils) at 1204. As a result, the first exposure is related to a point about 6.35 mm (250 million) from the zero point, ie X ₀ and Y _Q. The computer then writes out the special size and column and row of exposure at 1206 before checking the stripe scanner at 1208. An eight bit character indicating the X coordinate of the exposure is then entered into XNEXT at 1210. A check is then made at 1212 to see if the end of the row has been reached before an eight-bit character from the strip indicating the Y coordinate of the exposure is entered into YNEXT at 1214. YNEXT is then converted to thousandths of an inch (1 inch = 25.4 mm) and stored in XNEV / at 1206, and YNEXT is converted to thousandths of an inch (1 inch = 25.4 mm) and stored in YNEV / at 1218. Exposure is performed at 1220 by the exposure subroutine shown in FIG.

Each 'further exposure is made by recreation of the program starting at 1210 until the end of the row is sensed at 1212 on the strip. Then the space indicator again set at 1222, the table zeroed at 1224

00982 3/0691 - 56 -

BATH ORIGINAL

Ti '

A 36 810 b _w

10. 10. ' 1968 3 if

and the computer writes out "run complete" information at 1226 before returning to point cc of the program .

A summarized explanation of the operation of the device according to the invention is given below.

In order to prepare the camera 10 for operation, suitable lenses are accommodated in the holders of the lower frame 34. Usually reduction lenses from five to twenty times be used \ performance. The bellows 40 can be raised to the positions shown in Figure 22 by driving the motor 509 to raise the plates 501 and 503. The lower frame 34 can be adjusted to the proper height above the table 26 by actuating the drive motor which drives the perforated steel belt 472 and thereby the sprockets 462 and 464. First, one end of the lower frame 34 is adjusted to the correct height using, for example, the optical viewing device 484 and calibrated rod 310, and the brake 480 at that end is applied. The other end of the lower frame is then adjusted to the appropriate height in that the drive motor receives a pulse, whereupon the g brake 482 is applied in order to lock the frame 34 exactly horizontally. The straight bearings (466 and 468, which sit on guide rods 306 and 322 at diagonally opposite corners of the frame, ensure that the frame does not tilt about an axis that runs between the two sprockets 462 and 464. The upper frame 32 is then in The movable frames 32 and 34, together with the easily interchangeable lenses, provide a wide reduction range over a large area without the images being reduced by more than the basic tolerance of about 40 millimicrons (1, The bellows 40 can then be lowered by actuating the motor 509 until the sleeves 505 and 507 are lowered onto the appropriate lens mounts.

009823/0691 - 57 -

SAD

A 36 810 B 10. 10. 1968 ^ly " ¹³³

The nut transparencies are inserted into the supports 120 for the upper frame 32. Every mother transparency body is made up from a square glass plate with high resolution, each plate having two contiguous reference surfaces which are arranged at right angles to each other. The image to be projected is in the form of opaque and translucent Areas formed by developing a photosensitive material coated on a surface of the glass plate. The mother transparency bodies are inserted into the corresponding compartments of the carrier 120, whereby the surface on which the transparency layer is designed to move downwards towards the pegs 136 looks, and further the reference sides are biased against the lugs 138, 140 and 146 by the spring-loaded member 154. When the lid 126 is closed, the leaf springs 16O hold each panel firmly in the corresponding compartment, with the surface of the plate carrying the pattern is aligned exactly parallel to the reference surfaces 176 on the bottom of the base plate 122, wherein furthermore, the reference surfaces of each plate are parallel to surfaces 172 and 174 on the side wall of the base plate.

Each of the carriers 120 containing the mother transparency bodies can then be inserted into the upper frame 32 in which the Hand levers 414 and 416 are rotated so that the rails 418 and 420 are raised. The carrier 120 can face down be used by placing the reference surface 176 over the rails 418 and 420 are pushed until the surfaces 170, 172 and 174 abut the lugs 391, 390 and 392. The spring-loaded Levers 394 are then placed in position to urge the carrier against the lugs while hand levers 414 and 40 are rotated to lower reference surfaces 176 onto upward stops 395. Each · of the motherboards is then precisely aligned to the correct point along the corresponding optical axis, it also becomes exactly perpendicular to the corresponding optical axis optical axis and aligned exactly in the correct rotational position relative to the corresponding optical axis.

009823/0691 -58-

BATH ORIGINAL

A 36 810 b

10. 10. 1968 ^ly " ¹³³

The photosensitive plates to be exposed are identical to the plates on which the mother transparencies are formed. Medium-sized mother transparencies can be produced with the camera 10, as a result of which the total number of exposures which is necessary to produce a finished mask can be largely reduced. The photo-sensitive plates are inserted into the carriers 120 in the manner described above. This places the carriers in the table 26 with the reference surfaces 176 facing upward by rotating the hand levers 262, lowering the stools 248 and 250 below the rails 208 and 210. The carriers 120 are then over the rails 208 and 210 pushed until the surfaces 170, 172 and 17 ¹ J abut the lugs 212, 21Ί and 220. Then the spring-loaded lever 224 is in its position ^ ⁸⁰ * ¹ * ⁶¹¹ ^, to press the carrier against the lugs, while the hand levers 262 are rotated to raise the bolsters 248 and 250 and the reference surfaces 176 on the support against the stop pins 206. Thereafter, the photosensitive material is deposited on the surface of each of the plates at exactly the correct point on the corresponding optical axis, also aligned exactly at right angles to the axis and in a precise rotational position about the axis.

A typical photomask made by the camera 10 is shown in Figure 32 with each of the rectangular patterns representing the image that has been projected from the mother transparency body. However, the mother transparency is usually five- to twenty times larger than the picture on the finished mask. Leg operation the camera 10, the table is adjusted so that the image projected along the optical axis onto the film plate in one of the shown Exposure positions occurs. The film plate is then exposed and the table 26 is moved to the next exposure position, whereupon the next exposure is carried out. This process is repeated until all exposures are for the photomask have been carried out. For a mask with a large area, e.g. a square with a side of 7.5 cm

009823/0691

- 59 -

BADORIO1NAL

A 36 810 b

10.10.1968

Iv - 133 *

(3 inches) and small pictures, 10,000 exposures may be required be. Although the stage can be moved in stages and the exposure done in two seconds, the manufacture would such a mask take about five hours.

However, the total number of exposures and thus the total time required by the camera can be greatly reduced because the plates used for the mother transparencies, which are attached to the upper frame 32, and the photosensitive plates, which are arranged on the table 26, are of the same size and have been positioned on both table 26 and frame 32 by the same system. It is assumed, for example, that the finished mask should consist of an arrangement of 100 × 100 or 10,000 individual patterns. It is also assumed that the size of these basic patterns should be reduced by several times. The lens located in the lower frame 3 ^ is then selected and the upper and lower frames 32 and 34 adjusted to produce a reduction equal to the square root of 16, ie, a reduction by the factor k. The basic pattern is then pushed stepwise over the photographic plates which are arranged on the table 26 in order to produce an intermediate pattern with an exposure series of 10 × 10. Thereafter, the photographic plates placed on the table 26 are removed and developed to produce intermediate mother transparency bodies which are then installed in the upper frame 32. The same program is then repeated using a reduction factor of k and the camera is shifted through the same bracketing of 10 χ 10.

The finished product is then an arrangement of 100 χ 100 exposures with a reduction of 16. However, only 200 single exposures are necessary compared to 10,000, which means the exposure time is reduced from more than 5 hours to less than 10 minutes. The reduction values can be any desired value as long as the product of the values is 16, and the number of loading

0 0 9 8 2 3/0691 - 60 - "

BATH ORIGINAL

A 36 81.0 b
10.10.1968

Clearances in the intermediate pattern can be an arrangement of 20 χ 20 or 5x5 and the final exposures using the intermediate pattern can be arrangements of 5x5 or 20 χ 20, as desired. However, the fourth of the square roots indicate the minimum total loan-to-value ratio.

It is important to keep the entire camera 10 at a constant temperature. Otherwise, expansions and shortenings of the various elements of the camera would cause displacements of the mother transparency body, the lenses and the photographic plates that are greater than the permissible tolerances. This difficulty is reduced to a minimum in that the base plates for the table 26, the upper frame 32 and the lower frame J> k are made as identical as possible, so that the expansions and shortenings are the same. However, it is important to keep the temperature in the room in which the camera is installed as constant as possible, preferably within a tolerance of 0.05 C (0.1 F). The individual base plates should also be designed in such a way that stretching or shortening as a result of vibrations or shocks is reduced to a minimum.

The table 26 (Fig. 33) is moved to the successive Beiich- ä processing positions by the computer which operates according to the program described above. It is useful to express the movement of the table in terms of X and Y of the coordinate directions of the last series of exposures. It will be noted that the table 26 is actually moving in the opposite direction. The table is first pushed into the positive quadrant, which _{is delimited by the limit switches X r} and Y _r , if it is not already in these quadrants. This is achieved by moving the table first in the positive X direction and then in the positive Y direction until the limit switches X. and Y _{r are} switched off.

Ju II

Then the table becomes both in the negative X-direction and more negative at the same time Moved in the Y direction until the limit switches X and Y are switched on again. As soon as X_ is switched on, the X motor becomes

009823/0691 - 61 -

BATH ORIGINAL

A 36 81 Ob

10.10.1968.

is set to zero and once Y. turned on, the Y-Mo- tor is gestellt- to zero, so that the table in the vicinity of the starting point X ₀ and Y holding _0th The brakes are then applied so that personnel can insert the carriers 120 into the camera.

When the program is resumed, the brakes will be the Drive motor released and the table very slowly in! "Direction

he the timing

moves until X index crosses 0, to which. an interrupt signal is generated. On the basis of the interrupt signal, the computer resets the most characteristic 21 bits of the interference fringe counter and reads and stores the last three characteristic bits. Then the X-motor is set to zero. The table is then slowly moved in the Y direction ₅ until it _{crosses the Y Q} axis, at which point an interrupt signal is given to the computer ₉ which resets the most characteristic 21 bits of the Y axis strip counter and the numbers of the last three identifying bits are sampled and stored »

The computer then activates the X and Y motors to continuously set the table to the zero point (X ₀ Y _Q ) in order to iaach the first exposure at the zero point. Then the .Computer scans the barometer pressure r, b and calculates the distance in expressions of X and Y stripes for the first exposure. XTAL = 1 and YTAL = 1. The computer then operates the drive motors for the X and Y directions in such a way that the table is accelerated in X direction with the maximum design force to the masciraal design speed. When the table has reached the design speed , the X-Ho gate is set to Mull "While the table its orbit continues _s measures, the computer periodically the speed of the table and operates the X-motor to generate a series of pulses ;, which is either in the positive or negative X-direction act to keep the design speed within a specified tolerance »When the table has reached a point that is a predetermined distance from the first

0 0 9823/0891 - 62 -

BATH ORIGINAL

A 36 810 b

October 10, 1968 (J3

Exposure, an interrupt signal is generated due to whose computer calculates the point on the basis of the table speed at which the drive motor should be set to the maximum design force or to the maximum setpoint strength, to stop the table in the first exposure. When the table has reached the point at which the deceleration force is applied must be generated, another interrupt signal is generated on- because of which the computer switches the X-motor to deceleration until the speed of the table has been reduced to a predetermined minimum. Then the X-motor is set to zero. The J The same process is then repeated with the Y-motor in order to bring the table into the Y-coordinate for the first exposure.

Once the table is positioned near the first exposure, the computer begins to run a positioning program to sense the position of the table in the X and Y directions, with the strip counters are scanned, the Y and Y graduations are compared with the exposure position to determine an error value and further actuating the X and Y motors to move the stage back to the correct exposure position will. The computer keeps a record of the number of times the table is within specified tolerances and after the Table has been within the tolerance a predetermined number, the exposure lamps are automatically turned on to to begin exposure, the duration of which is controlled by a mechanical timer. The positioning program is run during the exposure is continued and it is recorded how often the position error was measured and how often the table is outside the tolerance is.

Moving between exposure positions only takes a fraction of a second, while exposure time can be up to one or two seconds. However, if a high intensity light source is used, the exposure time can be so short be that the table is in the exposure position during the exposure is effectively not moved, but only on XAIM and YAIM

0 0 9 8 2 3/0691 - 63 -

A 36 810 b
10.10.1968

^ly " ¹³³ " ⁷ . . 1803847

passes, at what point in time the light source is triggered. The stage is then incrementally moved over the first row, back along the second row, and over the third row, and so on until all exposures are made. However, no exposures are made at the four vacancies. After the exposures have been carried out, the table is returned to the zero point X ₀ and Y ₀ and a second HOME exposure is carried out. The offset between the first and second HOME exposures is a measure of the cumulative error of the run.

The system then goes into standby position while the carriers 120 in the upper frame 32 are removed and carriers 120 are inserted which contain mother transparencies for the fiducial marks which are exposed at the four exceptional positions. Thereafter, the computer is again operated to _{reset the counters to X 0} and Y _Q , and the table is moved in the same manner to successively the fiducial marks in the first, second, third and fourth exception locations before returning to to expose the HOME position.

The invention is not limited to those shown and described Limited execution forms, but changes are possible ^ even without deviating from the claims »

0 0 9 8 2 3/0691 bad original

Claims (1)

Ä 3.6 810 b
10. "10. 1968 ^ "* 1" 8O3947 Claims 1. A camera for producing a plurality of images along a plurality of vertical optical axes, characterized by a table for holding a photographic plate and a projection system for projecting an image onto dies.e photographic plate, further characterized in that the table and the projection system are relatively movable in the direction of the X and Y coordinates, further by a drive { for moving the table and the projection system in the X and Y directions, a detector for sensing the movement of the table and the projection system in the X. - and the Y-direction, which generates information corresponding to the X- and Y-position and, by a controller, which is based on this. Information responds to actuate the drive so that the table and the projection system are movable into selected relative coordinate positions and that a relative coordinate position is maintained during the exposure of a photographic plate arranged on the table. 2. A camera according to claim 1, characterized in that the detector eist ge characterizing λ net an interferometer for the generation of interference fringes and a counter for counting the interference fringes on v /, which by the relative movement of the table and of the projection system were generated / and the controller has a digital computer for calculating the relative positions of the table and the projection system for counting the counter, and for actuating the table mechanism in full according to the calculated position of the information. 3. Camera according to claim 2, characterized in that that the detector has at least two detectors for sensing separate parts of the interference line and a logical one 0 0 9 8 23/0681 ^{BAD 0RiQ1NAL} A "36 310 b 10.10.1968. . iy- £ 133 * "^ tSO3947 Circuitry responsive to the detectors for generation of data "indicating the direction of table movement. A camera according to claim 1, characterized in that the photographic plate is approximately aligned with each of a plurality of perpendicular optical axes, and in that the projection system further comprises a support which extends over the table and a lens frame which extends over the table is held by the support and is movable and adjustable in the vertical direction, and that this lens frame has a lens system for each optical axis, and that above the lens frame, an upper frame is held by the support, which is movable and adjustable in the vertical direction _s and that finally the upper frame has a light source for each optical axis. Camera according to claim 1, characterized by an X comparator for generating an X comparator interrupt signal when the count of the interference line counter for the X direction is the number set in the X comparator is the same, further in that the number of the X comparator is adjustable by the computer and by a Y comparator to generate a Y comparator interrupt signal if the count of the interference line counter for the Y direction equals the number set in the comparator, furthermore in that the number of the Y comparator is adjustable by the computer and that the computer is programmed to respond to the X comparator interrupt signal and the movement in the X direction, and un to the Y comparator interrupt signal address to decelerate the movement of the table in the Y direction. 009823/0891 Iy - 133 Of 6. A camera according to claim 2, characterized in that the "controller comprises a computer which is programmed to operate the drive mechanism to move the table in a reference division, un the number of interference lines in the X and in the To calculate the Y-direction from the reference position up to a coordinate offset, furthermore to operate the drive mechanism to move the table into the selected coordinate position and to operate the drive mechanism, and the table in the selected μ coordinate position within given tolerances during a To keep the period of time that is necessary to expose photographic plates with appropriate Bildörn. 7. Camera r ^ ch claim 2, characterized in that the detector is used to sense the interference lines comprises at least two potentiometer detectors located in the interference pattern are arranged at a distance, un two output signals see below that are out of phase, and that further a threshold mitigating device is provided to convert each of the output signals in logic II level3, and that a logic circuit is provided for decoding the logic levels, un first and generate second signals. (| 8. Camera according to claim 4, characterized e t that this upper frame has a light source for each optical axis and a mount sun cold of a photographic Has transparency body on each optical axis, and that a Uechanisrjus is also provided, which on each frame is arranged to raise and lower the corresponding racks and that a 3ren-.se at each ignition of the racks ^^ rachts which engages in a 3rense adjacent to the corresponding ends of the racks, and optionally the corresponding ignitions of the racks in the dug-up mowing across the table. .0 09823/0691 BATH ORIGINAL a 36 31Ob 1803847 10. 10. 19o3 / j> Iy - 133 ** -X- 9. Camera according to claim 3, characterized in that e t that each frame is stabilized in a plane perpendicular to the optical axes by two perpendicular guide rods, which are in diagonally opposite one another Corners of the frame are arranged and attached to the top, and that the frame is provided with elongated bearings is slidably arranged in each of the guide rods are. 10. Camera according to claim 1, characterized e t that the table has an adjustment mechanism to to precisely define standardized photographic plates in a standardized, predetermined position relative to each of the optical axes, and that: read the upper frame an identical adjustment mechanism up / down to specify a transparency body exactly, the photographic plate standardized by developing the same in the same position relative to the corresponding optical plate Axis was made. 11. A camera according to claim 10, characterized in that the adjustment mechanism has a controller for positioning a face of each photographic plate in a plane perpendicular to the corresponding optical plate Axis, further three adjustment points which are arranged to be two adjacent sides of the photographic plate capture un the photographic plate in a predetermined Relationship in the Z- and in the! -Direction and in a given one Adjust the rotational position relative to the optical axis. 12. Camera according to claim 1, characterized by one
a reference circuit for generating / II -'- Tullpunkts-ünterbrechunss- . signals when the table has the zero reference point in the X direction happens and to generate a Y-zero point interruption 009823/0691 - 5 - BATH ORIGINAL A 36 810 b
October 10, 1963
Iy- 133 Chungssignales when the table has the cheesecloth reference point in the Y-direction happens, further by a scrap detector for generating a hull signal when the table is the scrap reference point happened. 13. Camera according to claim 1, characterized that the drive mechanism for moving the table in the X and Y directions has a first guide rail has, which is anchored in the support surface and runs in one of the coordinate directions, furthermore an intermediate " Stell, which is supported by air bearings on this support surface and carries a second guide rail, which is located in the other coordinate direction extends, furthermore a guide on the intermediate frame in order to be slidable in the first guide rail to intervene in order to restrict the movement of the intermediate frame to this one coordinate direction, and that a guide is provided which is connected to the table in order to capture the second guide rail to the movement of the To restrict the table relative to the intermediate frame to this other coordinate direction. 14. A camera according to claim 12, characterized marked ■ Λ is characterized in that each of the guide rails has a flat surface, and that each of the guides is provided with a plurality of air bearings, which rest on the planar guide surfaces. 15. Xamera according to claim 1, characterized by a coding system for measuring the barometric Pressure on the interferometer and for generating a digital coded signal corresponding to this pressure point. 16. Camera according to claim 2, characterized e t that this interferometer is a laser interferometer is and that it generates a first interference pattern whose .069823 / 0891 -6- BAD onou *. ^{A 3δ 8l} ° ^b Λ οηοη /. '7 Iy- 133 p * Movement relative to the movement of the table in the X direction and which generates a second interference line pattern whose movement is relative to the movement of the table in the Y-direction is, and that further two photodetectors are arranged at a distance in each of these interference fringes, and that each pair of the photodetectors into the corresponding interference pattern is spaced to produce two output signals that are 90 degrees out of phase and also two Threshold detectors connected to the output of each photodetector and that the threshold level of the group of four threshold detectors connected to the outputs of each of the Threshold detector pairs is connected, is chosen so that the output of one of the four threshold detectors the logical Level changes each time a corresponding ' Interference line pattern moves one eighth of a cycle. 009823/0681. ^{BATHROOM 0RIGINAL} Blank page