[1]SFAiicfi ROOM Patented July 19..1949 21476@426 UNITED STATES PATENT OFFICE 2,476,426 APPARATUS FOR TEST]ING PORRO PRISMS USING NONPARALLEL LIGHT John R. McLeod, Rochester, N. Y., assignor to Eastman Kodak ComPany, Rochester, N. Y., a 6orporation of New Jersey Application May 2, 1945, Serial No. 591,501 7 Claims. (Cl. 88-14) 2 The present invention relates to a testing instrument, and more particularly to an Instrument for testing prisms in converging or nonparallel light in the same manner as when used iii binoculars to determine the presence of vari- 5 ous errors of such prisms. As is well known, a pair of Porro prisms are usecl to erect the image in a telescope or binocular. Each prism serves to deviate the light 180' Such a prism, referring to the drawings whi@h will 10 hereinafter be described in detail, has two internally reflecting faces 51 and 5 1 ' perpendicular to each other and to the triangular faces 64 ' and at an angle of 45' to the hypotenuse face 510. The light enters the hypotenuse 50 of each prism and 15 is totally reflected internally by each of the two reflecting faces 51 and 511 and then emerges from the other end of the hypotenuse 50. Such prisms ,errors may have errors of various kinds, and thes(, 20 may produce various errors in the teles(,ope or binoculars. These errors are as follows: I-Deviation errors 1. 90-degree angle error, that is the amount by which the angle between the faces 51 and 5 I' 25 differs from a right angle. 2. Pyramid error, that is, the tilt or angle by which the line B, constituting a 90-degree edge, differs from parallelism with the hypotenuse face 50. 30 II-Disl)lacement errors 1. Sharp dimension errors, which means the amount by which the prism is, as a whole, larger or smaller than the standard size. This maY be measured either by variation in the le,ngth A 35 of the hypotenuse face 50, or in the distance of the line R from the hypotenuse 50. 2. Lateral displacement of the 90-degree edge B. When the right angle is accurate, errors In the 45-degree angles, one being larger and the 40 other smaller than 45' by an equal amount, will result in the displacement of the edge B. 3. Inequalities of the bevels. That is instead of the bevels being of equal length and cutting off equal portions of the apex of the 45-degree 45 angles, one cuts off more than the other throwing the right angle B off center. III-Rotation errors 50 That is, errors caused by sides 64. of the prism not being normal to the reflecting surfaces, having the effect of rotating the system about an axis passing through th-e hypotenuse. If the 90-degree angle is in error, there wiR be N a deviation from 180'; but if a Pyrarnid error is present, the deviation wili be at an angle of 90, from that caused by the 90-degree-angle error. D!sPlacement errors are not detectable in parallel light so that non-parallel light must be used to.detect these errors. Accordingly, I mount the prism to be measured between the reticles and a collimating lens, that is, at a point where the rays are not parallel. If the sharp dimension a, Pig. 5, of the prism is in error, the length of the light path through the prism is changed and converging light will focus at a different point. Thus, - in a sense, a Porro prism has a focal length. If the two 45degree angles are unequal, a displacement of the converging light occurs but no deviation from 1800. The other displacement errors have to do with the mounting of the prism. In binoculars, each Porro prism is mounted in a recess that fits over the hypotenuse surface. The bases or parallel sides and the beveled 45-degree edges locate the Prism. If the two bevels of the 45-degree edges are unequal, the prism is displaced lengthwise and a displacement of the image occurs. The displacement may added to or counteract the displacement due to the unequal 45-degree angies. The two bases or parallel sides of the prism serve to prevent rotation of the prism in the plane of the hypotenuse. If the prism rotates, it serves to rotate the image and is somewhat serious In binoculars. This means that the bases or parallel sides must meet the hypotenuse in lines which are perpendicular to the 90-degree edge in order to eliminate rotational errors. Errors of deviation and errors of displacement are combined in a telescope. If errors of deviation and displacement are both present in a prism their effect, at the image plane, will add algebraically. If the Prism is far from the reticle, deviation errors are the more important. If, on the other hand, the prism is near the reticle, errors of displacement are of prime importance. In order to ascertain the suitability of the Porro prism for use in binoculars, the prism must be tested to determine wmeh of these errors are present, and if present, whether they are within the established tolerance limits. The invention flas, therefore, as its principal object the Provision of an apparatus for determining the various errors of a Porro Prism and indicating the amount of such errors. A further object is the provision of an apparatus that magnifies the error, in this case makes eacli twice its actual value.
[2]3 A sti ' ii further object of the Invention is the provision of such an apparatus which affords a ready, easy and reliable testing of such prisms. And still another object of the invention is the provision of an instrument of this type which is of a rugged construction suitable for use in production control. StiU another object of the invention is the provision of an apparatus of this type which enables the operator to determine whether the errors present are within the permissible tolerance limits. To these and other ends, the invention resides in certain improvements and combinations of parts, all as will be hereinafter more fully described, the novel features being pointed out in the claims at the end of the speciflcation. In the &awings: Fig. 1 is a vertical sectional view through a portion of the testing apparatus of the present invention, showing the position of the various optical members and the relation thereto of the prism to be tested; Fig. 2 is a sectional view through the testing apparatus, on line 2-2 of Figure 1 and at right angles to that shown in Fig. 1; Fig. 3 is a fragmentary view of the outside of the apparatus, showing the handle for moving the reticle, and the limiting stops therefor; Mg. 4 is a schematic arrangement of the testing apparatus showing the relation of the prism tG be tested, and Mgs. 5 to 8, inclusive, are the diagrams referred to in the discussion of the prism errors under observation. The same reference niimerals throughout the various views indicate the same parts. The &awings show a hollow housing which has sides I I and a top 12. The latter is provided with a recessed portion 13 and an opening 14 positioned in optical alignment with a lamp 15 mounted in the housing. The under side 16 of the top 12 has secured thereto, by bolts 17, a depending ring or sleeve 18, the lower end of which is internally threaded with a steep pitch thread tO receive a similarly threaded sleeve 19 which supPorts the lower end of an axiauy movable mount 20 which carries the reticle Ri. An opal or frosted glass 21 is positioned in the mount 20 below the reticle Ri to provide uniform illumination for the latter. To permit axial movement of the mount 20 and reticle Ri, for reasons to be later described, the sleeve 19 has secured thereto an arm 22 which projects laterauy through an opening 23 in one of the sides I 1, as shown in Fig. 2. Thus by moving the arm 22 in one direction, the sleeve 19 .will be rotated to move upwardly along the threads of the sleeve 18. This wfll serve to move the mount 20 and reticle Ri upwardly against the action of the coil springs 24 positioned between the upper face 25 of the mount 20 and the under surface 26 of the recess 13. When, however, the airm 22 is moved in the opposite direction, the sleeve 19 moves downwardly, and the mount 20 is siinilarly moved under the action of the springs 24 to move the reticle Ri downwardly. Thus the arm 22 and springs 24 serve to move the reticle Ri up and down alon.- the optical axis of the apparatus, for a purpose to be later described. In order to adjust the azimuth of the reticle Ri, ari opening 27 in the t3p 12 is threaded to receive a slotted adjusting nut.L8 on which is eccentrically mounted a depending pin 29 the lower end of which engages the end of the radially extending rod 30 which projects through an opening 31, 2,476,426 4 formed In the sleeve I 8 and extends into mount 20, all as shown in Fig. 1. Thus by rotating the slotted nut 28, the inount 20 may be slightly rotated to adjust the azimuth of the reticle Ri. 5 The top 12 bas mounted thereon a supporting member 35 formed with a vertical opehing 36 in optical alignment with the lamp 15 and aperture 14, as ibustrated in Fig. 1. The lower end of the opening is closed by an inclined semi-transparent 10 mirror 37 mounted in a frame 38 secured to the inclined bottom 39 of the member 35, as shown in klg. 1. The top 40 of the member 35 carries a plate 41 formed with a recess 42 arranged in optical align15 ment with the opening 36 and Providing a seat for a Porro prism 43 to be tested. By means of this arrangement, the prism 43 is positioned in the path of the light rays emitted by the lamp I 5 and passing through the reticle Ri and semi20 transparent mirror 37. Thus, the prism 43 is positioned in the system between the reticle Ri and the collimating lens, that is, in a beam the rays of which are not parallel. The prism 43 thus constitutes part of the optical system through which 25 the light beam Passes by a double internal reflection from the sides 51 and 51' of the prism, as is apparent. An eye-piece or microscope 44 is carried by the member 35 and is arranged in alignment with the opening 45 connecting the device 30 44 with the mirror 37, as shown in Fig. 1, so that the image reflected by the mirror may be viewed. As such eye-pieces are well known, details thereof are not deemed necessary. A plan inirror 48 serves to direct the beam reflected by the prism 43 to a 35 codimatirig lens 49 from which the light rays pass in parallel relation, as is well known. The mirror .48 is used merely as a convenience in arranging the parts. The light rays thus pass from the lamp I 5 40 through the reticle Ri and semi-transparent mirror 37 and into one side of the hypotenuse 50 of the prism 43; and, after a double reflection from the sides 5 I and 5 1', again emerge from the other side of the hypotenuse 50, and are then reflected 45 by the mirror 48 to the collimating lens 49. Such a device thus constitutes, in effect, a collimator with the prism 43 to be tested in position in the path of the converging light rays of the telescope between the reticle Ri and the collimating lens r)o 49. In order to reflect the light rays back through the collimating lens 49 toward the reticle Ri, I provide a reflecting means positioned beyond the lens 49 and in the path of the parallel rays pass55 ing therethrough. This reflecting means comprises, in the present embodiment, a front coated semi-transparent mirror 55 behind which is positioned a prism 56 arranged in the manner best shown in Figs. 2 and 4. A color filter 57, such as 60 a green Wratten Rlter No. 56 is positioned between the mirror 55 and the prism 56 for a purpose to be later described. These reflecting members direct the rays back through the system to the semi-transparent mirror 37 where they are 05 reflected and brought to a focus at the reticle R2 which is Positioned at the virtual image of Ri formed by the mirror 37. However, when looking through the eye-piece 44 at the reticle R2 the real Image of Ri will be seen. We thus have an auto70 collimating system. The reticle R2 preferably comprises dark lines on a transparent background such as glass, while the reticle Ri comprises transparent lines on an opaque background, although other forms and designs of reticles may be used. ya As the parts of the apparatus have now been de-
[3]2,476,426 scribed, their operation for making the various tests will now be discussed in the same order in which the tests were defined above. 901 angle error a U there is an error of the 90-degree angle between the faces 51 and 51', the rays deflected from the second side of the incorrect prism will be deviated and will not be parallel to the reflected rays of the correct prism, as shown in 10 broken lines in Pig. 7. These errors will appear at R2 -by having the image of Ri th the right or left of the center line of R2, the actual direction and amount depending, of course, on the direction and amount of displacement of 90-degree 15 error. Pyramid error Pyramid error gives rise to a deviation which is at right angles to that caused by the 90-degree error, or at right angles to the plane of the dia- 20 gram sh(ywn in Fig. 7. No other error in the prism compensates for pyramid error, and its value is read as the actual vertical displacement of the itnage Ri and R2. In binoculars, due to the fact that the two pi@isms are positioned at 25 right angles to each other, the deviation of the emerging ray Oue to pyraniid error in one prism may be compensated for by displacing the other prism in a plane normal to that of Fig. 7 so that the final image will be located @correctly. 30 Sharp dimension error Fig. 5 shows a correct prism in solid line, and a prism, as shown dotted, ha-oing a sharp dimensional error, designated -by the letter a. It will 35 be a,,pparent that the light path through the prism having such an error, shown in broken line, is longer than through the correct prism, as shown in solid line. The result of such an error is to cause the image of Ri to be brought to a 40 focus and appear between R2 and the mirror 37. If, however, the prism having the error was smaller than the correct prism, then the path therethrough will be shorter than through the correct prism, and the IfoGused image of Ri 45 would fall between R2 and the viewing device 44. In order to measure this er-ror, the reti-cle Ri is moved axially, -by means of the sleeve 19 and arm 22 or springs 24 so that the image of Ri is@ brought to a sharp focus at R2. During this 50 movement, small protuberances 60 on the arm 22 engage spaced stops 61. Each of these stops 61 is formed on the end of a spring member 62 carried by a sliding member 63 adjustably mounted on the sid-e i I of the apparatus. The 55 position of the stops 61 may be adjusted for various tolerances. If the focus 4Df Ri at R2 is secured with the protuberances 60 positioned somewhere between the stops 61, then the sharp dimension error is within the esta)bhshed tolerance limits. 60 I,f, on the other hand,, it is necessary to move a protuberance past one of the stops in order to secure -the focus of Ri at R2, then the error is too g-reat and the prisni is rejected. As a focal point is best detern-dned if the fgcus is passed through, 05 it is preferred to use spring stops so that the protuberances may pass thereover during thle focus-: ing operation. The resolution may be checked by o:bserving at R2 the iniage of a resolution chart plac,ed at Ri. 70 45-degree angle and bevel errors Fig. 6 shows the result of the lateral d-isplacement of the 90-degree edge B due to unequal 45-degreeangles. Suchdisplacenientcausestheiy&; SEARCH .@ROOlVi emerging rays to be displaced from but to be parallel to that of the correct prism, as indicated at C. Such error,% are not distinguishable from centering errors of the 90- degree edge B caused by unequal bevels at the ends of the hypotenuse, as shown In Fig. 8. Such centering errors are due to the fact that one bevel cuts off @more material than the other. Referring to Mg. 8, the correct dimensions of prism and the path of the light rays are shown in full line. If one of the bevels is higher than the other, the width of the prism being the same, the outl-ine of the prism and the path of the -rays will be as shown ia dotted line. It is to be understood that Figs. 5 to 7, which are used to ex-plain the errors, are not to be taken as indicating that bevels are not ordinarily used. Both the 45-degree errors and the bevel error cause dis-placement of the emergMg beam. In addition, as mentioned above, the 90-degree error,causes a deviation. The final result of the displacement and deviation errors is indicated by the failure of Ri to line up with R2. These errors may, however, compensate each other to bring Ri into proper relation with R2 so that it is not known whether such errors are lackiiig or whether they merely egmpensate each other. If, however, Ri and R2 do not line up when theprism is moved to the limit of its movement in seat 42, the prism 43 may be turned around and tested. This reversal serves to reverse the displacement, not the deviation error. if , after reversal, Ri and R2 continue in proper relation, the prism is satisfactory. If not, it is rejected, or possibly marked for special assembly, In the assembly of sueh prisrns in -binoculars, however, such marked prisms -niay be assembled with prisms having suitab le,compensating pyramid errors. This is possible because the pyranlid errgr is at right angles to the other errors mentioned above and theprincipal planes of the two Prisms are at Tight -angles. In deternuning the deviation and displacement errors, as above described, only the mirror 55 is required, and if only these errors are to be deterniined, the semi-tran@parent mirror 55 could be replacedby a fully reflecting -plane mirror and the prism 56 could be eliminated. Also, the deviation and displacement ertors, as indicated iby the abovedescribed apparatus, are twice the value of those which such a prism -would give when used in a,binocular. Rotation errors In addition to these deviations and displacemen@t errors, the prism 43 may have rotational errors caused by the fact that the intersections of the hypotenuse and the bases 64 are not normal to the 90-degree edge B. Such a defect will result in a rotation of the prism in its seat about, the axis X-X of Fig. 4, or about an axis parallel to face 50. In deterraining such r(ytational errors, the 90-degree prism 56 is required. If only the mirror 55 is used, itwill be foulid that the image Will be the same whether or not the pris-m has rotational errors. However, -with the addition jof the prism 56 it will be forand that the final image will have a twist which is four tknes that of the twist or rotation of the prism 43. In other words, if the prism 43 is rotated through an angle Odue to its rotational error, the final iinage will be rotated 4 O', the latter being twice the rotation which would occur if the defective prism were used in a binocular. Thus the prism 56 enables the detection of the rotational error in the ptism 43. If such errors are found to exist, the prisnl
[4]@43 is manually rotated, by a permitted amount, in its seat by the observer, and if such a movement, within the prescribed tolerance, is sufficient to correct the error, then the latter is within pern-iissible tolerance limits and the prism may be suitable for use inbinoculars. If, however, such manual rotation does not correct the error, the prism is Unsuitable and is rejected. In order that such rotational errors can be more; readily distinguished from deviation and displacement errors, a color filter 57 is positioned between the mirror 55 and the prism 56 in the position illustrated in Figs. 2 and 4. This filter is preferably a Wratten No. 56 green filter, and serves to color the image which goes through the prism 56 green, this colored image is the one that shows the rotational error of the prism 43. Thus if tlie prism has rotational error, the green image will be rotated or twisted relative to the reticle R2 to indicate clearly that such error is present, and the amount of such rotation or twist will be four times the rotational error of the prism. The above described apparatus thus enables an operator to ascertain readily, quickly, and easily whether the prism under test has deviation, displacement, or rotational errors, and also the amount of such errors. Also the operator can check quickly to see whether these errors are within the permissible tolerance limits set up and thereby determine the suitability of the priSM for use in binoculars. While the above apparatus has been specifically described in connection with the testing of Porro prisms in converging light to determine the different prism errors, it is to be understood and it is contemplated that such an apparatus is also adapted for testing various other types of prisms such, for example, as penta prisms, rhomboid prisms, roof prisms, etc. in converging light to ascertain the various errors of such other prisms as require converging light for their deterrnination. In each case, the prism to be tested is positioned in a converging light beam rather than in a beam of parallel rays. It is apparent, however, that the specific positions of the various elements of the testing apparatus may have to be slightly rearranged to suit the particular p-rism being tested, but such rearrangements will readily suggest themselves to those in the art. Also the term "Prism" when used in the specification and claims is intended to cover a single prism member or a plurality of members which are arranged, either cemented or otherwise, to form in effect a single prism. While one embodiment of the invention has been disclosed, it is to be understood that the inventive idea may be carried out in a number of ways. This invention is, therefore, not to be limited to the precise details described but is intended to cover all variations and modifications thereof falling within the scope of the appended claims. I
