The Mobsby null test
Martin Trittelvitz 2001
Today, there are essentially two types of optical tests to determine the surface aberration of telescope mirrors:
1.) The shadow sample by Foucault and its derivatives such as Zone Test by Couder.
2.) The Ronchi, and the quantitative Ronchi by Mel Bartels.
http://www.efn.org/~mbartels/tm/software.html(See also Foucault and Caustiktest)
Both have a common major drawback: For perfect spherical mirrors, both zero tests, that the shade sample after Foucault and the lines in Ronchigram , are straight and parallel. There have since taken no deviation, to be corrected or recalculated, to a perfect sphere. The evaluation is done by simple rvisual observation.
If the image from Foucault test, just complete dimmed, then the mirror should be spherical. If reading from the lines at Ronchi are straight, it is a spherical mirror!
Everything changes when you want to make a parabolic mirror. The lines at Ronchi test are bent, for each level different for each position of the grid as well. It's like the expulsion from paradise!
Everything becomes unmanageable, complicated, difficult to interpret. It is no surprise that some mirror grinders have been thinking where is the key to paradise, how to get back to a zero test. Now when the mirror is parabolic, it is very simple: One measures the double passage of a plane plate which sends the light of the test equipment back through the optics, and thus reverses the light path.
Because the test is very sensitive, the light rays is indeed passed through twice. Once the path through the optics on the plane plate and then back again. Thus, the appearance go with their mistakes twice, and the test contains all optical aberrations very clearly. But we poor wretches do not possess any optical grade plane plate! And not so much space to make sense to such an experimental setup either.
H.E.Dall hit on the idea of bringing a lens in the beam path, which compensates for the deviation of the light rays at the Foucault test. But this method is not so simple.
Similarly, one can also setup the Ronchi test, but Eric G. H. Mobsby has come to a simpler solution: He went from the consideration, it is true that , the lines are bent when Ronchi in a specific and predictable manner. Now what if you were to construct a grid that compensates for these variations?
This is bent as it were, by the opposite side? Then one would just need to find the right place to bring into the beam path, and you get lovely straight lines again! The evaluation will be focused on the viewing and judging whether or not the lines are really straight and really parallel.
At the times of Mobsby of course there was no computer available, but now everyone has. The calculation, the presentation and the requirement of a useful template is achievable. Mobsby calculated the lines by hand and drawn, then the drawing is photographed and then used the negatives as grating as a Ronchi grating. This method was probably a bit intimidating, and thus the reason that this test was used by few and remained largely meaningless.
Now we have computers, and since "Mark D. Holm" also has a simple to use program that calculates for us these lines and prints. Only the photograph with a normal reflex on slide film remains. The program can be downloaded here: http://users.telerama.com/~mdholm/atm/invron/invron.html(not available anymore). The program provides us with extensive support available! In addition to the inverse Ronchi are for printing also prepared: a straight Ronchi grating, knife edge, data printouts to the slide afterwards, to be able to among other similar figure out the representation in negative format is possible, so that when photographing, a negative film can also be used.
It is the merit of Mark Holm that this test can be used as usual today.implementation: After downloading the program installed to it by calling the setup program. I had this problem since my unzip program cannot cope with long file names, and therefore some files for allegedly same filename is simply overwritten. Therefore, these files were missing during setup, which was then terminated. It was tedious to rename each file respectively and after repeated call for the archive program to save as a new file.
Once the program is installed, you call it up by clicking InvRonchi icon.One gets the following input window:It must have the following entries are made:
1. Select) Unit: I always use millimeters.
2.) radius of curvature, after entering the calculated focal length automatically. (And vice versa)
3.) diameter of the mirror
4.) It can also be tested other than parabolic mirrors, You must then change the conical constant.k = 0 would represent a spherical mirror, k = -1 it means is a parabola.
5.) offset of the light source from the optical axis, both horizontally and vertically enter right = + = + above
6) Do not forget: The entry at Moving source when it comes to a test device with a moving light source.
7) on the right side of the mask is entered the "Offset from Grating Mirror" above. It is made of a bar, which is horizontally mounted in front of the mirror. in the bar two nails are symmetrically taken to the center as a marker. The separation of the nails from the middle one carries into the o.g .. field. Input as Relative indication of the radius, that is, the mirror edge would 1.0, the mirror center theoretically 0, this would be to choose, however meaningless, since one can not determine here the position of the mirror. In our case, have I raised the preset by the program size 0.912.
8) After you choose the position of the grid on the selected starting position.
9) and 10) The number of grid lines shown can be selected, and the width of the grid lines relative to the total mirror diameter in percent. The default is 3 lines and 10% linewidth, finally,
11) is followed by the reduction factor. He indicates the factor by which the grid by the picture on a photographic film is reduced. The default is 100, which means that the film plane of a still camera with 50mm lens , the image of the grid in the form of a print on paper must have exactly 510 cm distance. I choose the size 50, then the grid is shown half the size. This is then exactly the same size on the film at a distance of 255 cm.When selecting "Display", the grid is displayed on the monitor, choose "Print" opens a new mask with the different options of the print menus are served.
The program behaves partly a bit strange, it replaced once entered sizes seemingly random by their own values, which need not always necessarily accepted as useful but. e.g. it is better to specify the radius of curvature than an even number, because this divided by two, the focal length results. With point and point the program does not properly handle, therefore becomes of 1992.5 or 1992.5 then quickly 19925, and the ROC, we suddenly 39850mm But these are absurdities which primarily on the different use of full stops and commas here and in the US result. It is advisable in any event to check each number again after entering!
Construction:
We needed to test a device that allows a light source with a fine shadow mask and the grid obtained accurately on the mirror to and transverse to lead to him. For this purpose it is enough a block of wood with a aluminum tube. In the aluminum tube, the light source is attached, best a 12 V bulb as it is used for lighting today. Before the lamp is a hole that lets the light fall on a perforated mask. This one imagines from strong aluminum foil ago (lid of a yoghurt cup) which are pierced with the finest needle, which can be found. The pad when drilling must be quite hard so the needle is not far penetrates the film, then the hole the picture of the grid is very thin, sharp. The measuring device must be capable of adjusting the height now still, this one turns around at least three machine screws (M8) in the corners of the carrier. so that the device can be precisely adjusted in height. Of course, many different versions of this measuring system are possible. The more effort you driving this regard, the easier the precise means to the mirror and the grid is possible
Execution:It is the mirror from the reflected light beam on the slide with the grid on. Now one can cover the light beam with the vertical search line by lateral displacement of the grating. the search line forms on the mirror a circle which is brought by displacement of the meter on the mirror or away from him to the position above, here 0.912% it mirror radius.One sees now on the mirror a concentric dark ring, exactly via the above-mentioned Nails the highlight bar (eng. Pinstick) runs.from this initial position is shifted to the grid, and in moving the light source and the light source by the amount entered to mirror out 2,93mm here (with dial is easily possible).By raising or lowering of the measuring arrangement of Reflected light beam is now brought into alignment with the grid. Looking at it through the grid mirror one must now look absolutely straight and parallel lines that run over the mirror.
A very decisive property is owned by the present test way before all anderenn procedures that are available to me: It is not necessary to implement the test structure directly on the optical axis. This requires a beam splitter, which can outgoing and bring the incoming light beam to cover. This program is similar to the lateral displacement of with, You are the horizontal and vertical displacement in the mask, the grating compensates for this. This of course requires that you use the grid in the correct orientation, that is, the printed page to be scanned in an upright position, the negative or slide must be used also upright and with the coated side to the mirror back. But then there is no astigmatic deformation of the image more, which usually is caused by the small angle is formed between the light source and grid or knife edge.
Interpretation:Quite simple: If the lines are parallel then the mirror is corrected all right. If the lines are gathered in the form of ears of corn sheaf, then the mirror is under-corrected, they are barrel-shaped, then the mirror is over-corrected. It is crucial that the grid the precise position assumes relative to the mirror and the light source, because otherwise the lines may be bent.Also Zone errors and TDE one would now be able to see very clearly.That sounds now quite complicated, one has built a stable unit, a strong light source and a very fine needle hole in front, then this structure is not that complicated. Good offices can afford here the cross support of a disused lathe.It remains to be investigated whether the method is a telescope mirror to produce diffraction-limited accurate and sensitive enough or determine the maximum size and focal length, a mirror can be tested with this method. Desirable would be a line density of about 10 lines / mm to obtain a resolution of approximately 1/4 Lambda. (Analogy from Ronchi). Then the image of the grating on the film would however / be 2mm wide around 1, so it is quite difficult to align the grid accurate enough.In any case, this method is to use a mirror only once to a threshold to get where then even finer measurement methods such as wire test (that later) or the Caustic test must accompany the last nanometers of the distance.The not to be underestimated advantage of this method is the fact that to get along until then without any scrap paper, pen and data confusion. Once you've made the mask for a particular level, then you do not need to process any more measurement results
Practical experience:
I have carried out this test on a 17.5-inch f4.5 Mirror.In addition I have described above, an inverse Ronchi grating manufactured. The offset I have adjusted to a value of 0.812, because I have to fit a mark on the Pinstick to this value. The zero position is 7 mm for back mirror. In the print menu you can also choose the expression as a negative that I have used since I wanted to photograph the template with SW negative film. One can just as well a slide film such as Kodak Elite Chrome 100 use, but we must take the delay of two days to develop in purchasing. The black and white film can easily develop in the kitchen itself. Additionally, I have next to the inverse Ronchi grating even asked a straight Ronchi grating, to a black cross with a square in the middle. This can use the conventional shadow sample after Foucault perform in different diameters. It is possible largely to rule astigmatism. Additionally, I have to put a sheet on which an identification number is printed as large that the slide can be identified later. Another sheet is provided with black lines of different widths, the lines shown on the slide can then be used to test wire. So I get a slide which enables me to perform very different tests without great technical effort, a single test I would rely on no account!
I am generally an sailors, and sailors do not trust their own butt!The slide I made apx25 film with an Agfa and developed with Documol. The templates I have to simply attached to a door with tape, before provided me with the camera at the correct distance and flash exposed. I have a bracketing made of F = 2.8 to 5.6, but it turned out that this was not necessary because all the negatives well exposed, were sharp and contrasty.The performance of the assay designed to easily. It provides the slide so in the light beam, that it falls on the vertical search line above or below the inverse lattice. This line you see now when you look through the slide on the mirror, projected as a circle with a vertical line on the mirror. It represents the distance of the grid from the mirror now so that this circuit therefore passes through the selected zone by the markings on the Pinstick that has been chosen for the calculation of the grid.
Now you move the grid to the mirror back to the amount one has intended (here 2.93 mm). I can control this route exactly with the gauge.By moving the grid up (or down) is pushed, the inverse Ronchi grating in the beam path. If everything was done properly then the grid fits into the size of the mirror. the grid must be accurately centered, suddenly the lines appear straight and to be parallel. In my mirror the central area looks shirred something that shows that the mirror is under corrected something here. Just as I had already noted with the wire test.It remains the mirror now to correct and see whether this deviation is then also seen.
In any case I give the Mobsby null test in preference to the "quantitative Ronchi". In this, the Ronchigram are calculated a plurality of grid positions and compared with the actually present Ronchigram. The process is the exploration for several positions of the grid significantly more complex. In addition, I can be much safer to decide whether a line parallel to another, and whether it is just as if several lines so run exactly as predicted on a computer printout. The preparation of a suitable grid is certainly not an obstacle, even if it has to be redone for each mirror. With the new program, you can create the template in 10 minutes, to photograph and develop takes maximum 2 days if you develop yourself about 1 hour.
I have now tested the sensitivity: In addition I have only once observed, as the focal length must change, so that a deviation of 1/4 Lambda happens. In the mirror I just are editing the 1900 mm instead of 1992.5 mm, the difference I find surprisingly high!I then two identical lattice calculated one for the actual focal length of the wrong to short focal length. I have made it two pairs, one for a intrafocal position, one for a extrafocal position. I have compared against each other, both pairs in the test. In both grids could be seen a fine but visible difference in the test. Both intra- and extrafocally could recognize "false screens" a very slight sheaf-like deformation of the lines I in. The difference is not large, but clearly visible.Another feature is also clearly visible: (extrafocally clearer) The grid produced by the program fits with a very small overhang exactly into the mirror. The calculated for different focal length grating is at the exact position clearly too big!This is precisely the reason why the relatively elaborate positioning proceed must be taken into account. If one fits a grid simply by its size in the cone of light, then one and the same grid at very similar focal lengths as well just give lines without the mirror is properly corrected.I can say with my 17.5 "f / 4.5 mirror am I detect a deviation of less than 1/4 lambda able. For a good mirror, this method is absolutely sufficient. At the same time I can not in my mirror detect visible deviation more and therefore declare him for it! But as I said, I would never leave me alone on a test and to check with zone test or wire test.addendum:I have the above-mentioned Experiment performed with grids of different position. The sensitivity of the test increases significantly when the exact grid position is very close to the center of curvature. As is known indeed from the Ronchi for spherical mirrors, in which the sensitivity is in the immediate vicinity of the passage is the greatest. Here you can accept the fact that the lines in the center of the mirror just disappear (corresponding to the above passage, finds the parabolic nature in each zone to a different location instead.) One can with a second grid illustrating this game just yet (1mm on to mirror back) close the gap. My printouts I do now so that I on the front side, the grid data prints the grid and Identification number on the back. which reduces the amount of paper produced considerably. The photographs are made on a black background, left a straight grid, right my Astigmatism cross. I have time to experiment with different patterns. For example, you could test a grid with vertical and horizontally intersecting lines on astigmatism.
Sources:
Ingals; "Amateur Telescope Making", 3 volumes, Wilman Bell
Mark D. Holm, "InvRon - Inverse Ronchi Grating; Program" http://users.telerama.com/~mdholm/atm/invron/invron.html. "InvRonHelpText.htm" help File for the above-mentioned Program is downloaded with the program.
Eric G.H.Mobsby / R.E.Cox / Roger Sinnott; "Gleanings for ATM's, Photographing the grating," from "A Ronchi zero test for paraboloid". Sky and Telescope, November 1974
original Martys website in german
Martin Trittelvitz 2001
Today, there are essentially two types of optical tests to determine the surface aberration of telescope mirrors:
1.) The shadow sample by Foucault and its derivatives such as Zone Test by Couder.
2.) The Ronchi, and the quantitative Ronchi by Mel Bartels.
http://www.efn.org/~mbartels/tm/software.html(See also Foucault and Caustiktest)
Both have a common major drawback: For perfect spherical mirrors, both zero tests, that the shade sample after Foucault and the lines in Ronchigram , are straight and parallel. There have since taken no deviation, to be corrected or recalculated, to a perfect sphere. The evaluation is done by simple rvisual observation.
If the image from Foucault test, just complete dimmed, then the mirror should be spherical. If reading from the lines at Ronchi are straight, it is a spherical mirror!
Everything changes when you want to make a parabolic mirror. The lines at Ronchi test are bent, for each level different for each position of the grid as well. It's like the expulsion from paradise!
Everything becomes unmanageable, complicated, difficult to interpret. It is no surprise that some mirror grinders have been thinking where is the key to paradise, how to get back to a zero test. Now when the mirror is parabolic, it is very simple: One measures the double passage of a plane plate which sends the light of the test equipment back through the optics, and thus reverses the light path.
Because the test is very sensitive, the light rays is indeed passed through twice. Once the path through the optics on the plane plate and then back again. Thus, the appearance go with their mistakes twice, and the test contains all optical aberrations very clearly. But we poor wretches do not possess any optical grade plane plate! And not so much space to make sense to such an experimental setup either.
H.E.Dall hit on the idea of bringing a lens in the beam path, which compensates for the deviation of the light rays at the Foucault test. But this method is not so simple.
- The lens is required to be accurately positioned in the beam path.
- Also to get the right interpretation, the distance from the light source is to be accurately calculated and adjusted.
- The alignment to the mirror is crucial.
- Finally, you need a monochromatic light, otherwise will disturb judging chromatic aberration of the lens ( mirror?).
Similarly, one can also setup the Ronchi test, but Eric G. H. Mobsby has come to a simpler solution: He went from the consideration, it is true that , the lines are bent when Ronchi in a specific and predictable manner. Now what if you were to construct a grid that compensates for these variations?
This is bent as it were, by the opposite side? Then one would just need to find the right place to bring into the beam path, and you get lovely straight lines again! The evaluation will be focused on the viewing and judging whether or not the lines are really straight and really parallel.
At the times of Mobsby of course there was no computer available, but now everyone has. The calculation, the presentation and the requirement of a useful template is achievable. Mobsby calculated the lines by hand and drawn, then the drawing is photographed and then used the negatives as grating as a Ronchi grating. This method was probably a bit intimidating, and thus the reason that this test was used by few and remained largely meaningless.
Now we have computers, and since "Mark D. Holm" also has a simple to use program that calculates for us these lines and prints. Only the photograph with a normal reflex on slide film remains. The program can be downloaded here: http://users.telerama.com/~mdholm/atm/invron/invron.html(not available anymore). The program provides us with extensive support available! In addition to the inverse Ronchi are for printing also prepared: a straight Ronchi grating, knife edge, data printouts to the slide afterwards, to be able to among other similar figure out the representation in negative format is possible, so that when photographing, a negative film can also be used.
It is the merit of Mark Holm that this test can be used as usual today.implementation: After downloading the program installed to it by calling the setup program. I had this problem since my unzip program cannot cope with long file names, and therefore some files for allegedly same filename is simply overwritten. Therefore, these files were missing during setup, which was then terminated. It was tedious to rename each file respectively and after repeated call for the archive program to save as a new file.
Once the program is installed, you call it up by clicking InvRonchi icon.One gets the following input window:It must have the following entries are made:
1. Select) Unit: I always use millimeters.
2.) radius of curvature, after entering the calculated focal length automatically. (And vice versa)
3.) diameter of the mirror
4.) It can also be tested other than parabolic mirrors, You must then change the conical constant.k = 0 would represent a spherical mirror, k = -1 it means is a parabola.
5.) offset of the light source from the optical axis, both horizontally and vertically enter right = + = + above
6) Do not forget: The entry at Moving source when it comes to a test device with a moving light source.
7) on the right side of the mask is entered the "Offset from Grating Mirror" above. It is made of a bar, which is horizontally mounted in front of the mirror. in the bar two nails are symmetrically taken to the center as a marker. The separation of the nails from the middle one carries into the o.g .. field. Input as Relative indication of the radius, that is, the mirror edge would 1.0, the mirror center theoretically 0, this would be to choose, however meaningless, since one can not determine here the position of the mirror. In our case, have I raised the preset by the program size 0.912.
8) After you choose the position of the grid on the selected starting position.
9) and 10) The number of grid lines shown can be selected, and the width of the grid lines relative to the total mirror diameter in percent. The default is 3 lines and 10% linewidth, finally,
11) is followed by the reduction factor. He indicates the factor by which the grid by the picture on a photographic film is reduced. The default is 100, which means that the film plane of a still camera with 50mm lens , the image of the grid in the form of a print on paper must have exactly 510 cm distance. I choose the size 50, then the grid is shown half the size. This is then exactly the same size on the film at a distance of 255 cm.When selecting "Display", the grid is displayed on the monitor, choose "Print" opens a new mask with the different options of the print menus are served.
The program behaves partly a bit strange, it replaced once entered sizes seemingly random by their own values, which need not always necessarily accepted as useful but. e.g. it is better to specify the radius of curvature than an even number, because this divided by two, the focal length results. With point and point the program does not properly handle, therefore becomes of 1992.5 or 1992.5 then quickly 19925, and the ROC, we suddenly 39850mm But these are absurdities which primarily on the different use of full stops and commas here and in the US result. It is advisable in any event to check each number again after entering!
Construction:
We needed to test a device that allows a light source with a fine shadow mask and the grid obtained accurately on the mirror to and transverse to lead to him. For this purpose it is enough a block of wood with a aluminum tube. In the aluminum tube, the light source is attached, best a 12 V bulb as it is used for lighting today. Before the lamp is a hole that lets the light fall on a perforated mask. This one imagines from strong aluminum foil ago (lid of a yoghurt cup) which are pierced with the finest needle, which can be found. The pad when drilling must be quite hard so the needle is not far penetrates the film, then the hole the picture of the grid is very thin, sharp. The measuring device must be capable of adjusting the height now still, this one turns around at least three machine screws (M8) in the corners of the carrier. so that the device can be precisely adjusted in height. Of course, many different versions of this measuring system are possible. The more effort you driving this regard, the easier the precise means to the mirror and the grid is possible
Execution:It is the mirror from the reflected light beam on the slide with the grid on. Now one can cover the light beam with the vertical search line by lateral displacement of the grating. the search line forms on the mirror a circle which is brought by displacement of the meter on the mirror or away from him to the position above, here 0.912% it mirror radius.One sees now on the mirror a concentric dark ring, exactly via the above-mentioned Nails the highlight bar (eng. Pinstick) runs.from this initial position is shifted to the grid, and in moving the light source and the light source by the amount entered to mirror out 2,93mm here (with dial is easily possible).By raising or lowering of the measuring arrangement of Reflected light beam is now brought into alignment with the grid. Looking at it through the grid mirror one must now look absolutely straight and parallel lines that run over the mirror.
A very decisive property is owned by the present test way before all anderenn procedures that are available to me: It is not necessary to implement the test structure directly on the optical axis. This requires a beam splitter, which can outgoing and bring the incoming light beam to cover. This program is similar to the lateral displacement of with, You are the horizontal and vertical displacement in the mask, the grating compensates for this. This of course requires that you use the grid in the correct orientation, that is, the printed page to be scanned in an upright position, the negative or slide must be used also upright and with the coated side to the mirror back. But then there is no astigmatic deformation of the image more, which usually is caused by the small angle is formed between the light source and grid or knife edge.
Interpretation:Quite simple: If the lines are parallel then the mirror is corrected all right. If the lines are gathered in the form of ears of corn sheaf, then the mirror is under-corrected, they are barrel-shaped, then the mirror is over-corrected. It is crucial that the grid the precise position assumes relative to the mirror and the light source, because otherwise the lines may be bent.Also Zone errors and TDE one would now be able to see very clearly.That sounds now quite complicated, one has built a stable unit, a strong light source and a very fine needle hole in front, then this structure is not that complicated. Good offices can afford here the cross support of a disused lathe.It remains to be investigated whether the method is a telescope mirror to produce diffraction-limited accurate and sensitive enough or determine the maximum size and focal length, a mirror can be tested with this method. Desirable would be a line density of about 10 lines / mm to obtain a resolution of approximately 1/4 Lambda. (Analogy from Ronchi). Then the image of the grating on the film would however / be 2mm wide around 1, so it is quite difficult to align the grid accurate enough.In any case, this method is to use a mirror only once to a threshold to get where then even finer measurement methods such as wire test (that later) or the Caustic test must accompany the last nanometers of the distance.The not to be underestimated advantage of this method is the fact that to get along until then without any scrap paper, pen and data confusion. Once you've made the mask for a particular level, then you do not need to process any more measurement results
Practical experience:
I have carried out this test on a 17.5-inch f4.5 Mirror.In addition I have described above, an inverse Ronchi grating manufactured. The offset I have adjusted to a value of 0.812, because I have to fit a mark on the Pinstick to this value. The zero position is 7 mm for back mirror. In the print menu you can also choose the expression as a negative that I have used since I wanted to photograph the template with SW negative film. One can just as well a slide film such as Kodak Elite Chrome 100 use, but we must take the delay of two days to develop in purchasing. The black and white film can easily develop in the kitchen itself. Additionally, I have next to the inverse Ronchi grating even asked a straight Ronchi grating, to a black cross with a square in the middle. This can use the conventional shadow sample after Foucault perform in different diameters. It is possible largely to rule astigmatism. Additionally, I have to put a sheet on which an identification number is printed as large that the slide can be identified later. Another sheet is provided with black lines of different widths, the lines shown on the slide can then be used to test wire. So I get a slide which enables me to perform very different tests without great technical effort, a single test I would rely on no account!
I am generally an sailors, and sailors do not trust their own butt!The slide I made apx25 film with an Agfa and developed with Documol. The templates I have to simply attached to a door with tape, before provided me with the camera at the correct distance and flash exposed. I have a bracketing made of F = 2.8 to 5.6, but it turned out that this was not necessary because all the negatives well exposed, were sharp and contrasty.The performance of the assay designed to easily. It provides the slide so in the light beam, that it falls on the vertical search line above or below the inverse lattice. This line you see now when you look through the slide on the mirror, projected as a circle with a vertical line on the mirror. It represents the distance of the grid from the mirror now so that this circuit therefore passes through the selected zone by the markings on the Pinstick that has been chosen for the calculation of the grid.
Now you move the grid to the mirror back to the amount one has intended (here 2.93 mm). I can control this route exactly with the gauge.By moving the grid up (or down) is pushed, the inverse Ronchi grating in the beam path. If everything was done properly then the grid fits into the size of the mirror. the grid must be accurately centered, suddenly the lines appear straight and to be parallel. In my mirror the central area looks shirred something that shows that the mirror is under corrected something here. Just as I had already noted with the wire test.It remains the mirror now to correct and see whether this deviation is then also seen.
In any case I give the Mobsby null test in preference to the "quantitative Ronchi". In this, the Ronchigram are calculated a plurality of grid positions and compared with the actually present Ronchigram. The process is the exploration for several positions of the grid significantly more complex. In addition, I can be much safer to decide whether a line parallel to another, and whether it is just as if several lines so run exactly as predicted on a computer printout. The preparation of a suitable grid is certainly not an obstacle, even if it has to be redone for each mirror. With the new program, you can create the template in 10 minutes, to photograph and develop takes maximum 2 days if you develop yourself about 1 hour.
I have now tested the sensitivity: In addition I have only once observed, as the focal length must change, so that a deviation of 1/4 Lambda happens. In the mirror I just are editing the 1900 mm instead of 1992.5 mm, the difference I find surprisingly high!I then two identical lattice calculated one for the actual focal length of the wrong to short focal length. I have made it two pairs, one for a intrafocal position, one for a extrafocal position. I have compared against each other, both pairs in the test. In both grids could be seen a fine but visible difference in the test. Both intra- and extrafocally could recognize "false screens" a very slight sheaf-like deformation of the lines I in. The difference is not large, but clearly visible.Another feature is also clearly visible: (extrafocally clearer) The grid produced by the program fits with a very small overhang exactly into the mirror. The calculated for different focal length grating is at the exact position clearly too big!This is precisely the reason why the relatively elaborate positioning proceed must be taken into account. If one fits a grid simply by its size in the cone of light, then one and the same grid at very similar focal lengths as well just give lines without the mirror is properly corrected.I can say with my 17.5 "f / 4.5 mirror am I detect a deviation of less than 1/4 lambda able. For a good mirror, this method is absolutely sufficient. At the same time I can not in my mirror detect visible deviation more and therefore declare him for it! But as I said, I would never leave me alone on a test and to check with zone test or wire test.addendum:I have the above-mentioned Experiment performed with grids of different position. The sensitivity of the test increases significantly when the exact grid position is very close to the center of curvature. As is known indeed from the Ronchi for spherical mirrors, in which the sensitivity is in the immediate vicinity of the passage is the greatest. Here you can accept the fact that the lines in the center of the mirror just disappear (corresponding to the above passage, finds the parabolic nature in each zone to a different location instead.) One can with a second grid illustrating this game just yet (1mm on to mirror back) close the gap. My printouts I do now so that I on the front side, the grid data prints the grid and Identification number on the back. which reduces the amount of paper produced considerably. The photographs are made on a black background, left a straight grid, right my Astigmatism cross. I have time to experiment with different patterns. For example, you could test a grid with vertical and horizontally intersecting lines on astigmatism.
Sources:
Ingals; "Amateur Telescope Making", 3 volumes, Wilman Bell
Mark D. Holm, "InvRon - Inverse Ronchi Grating; Program" http://users.telerama.com/~mdholm/atm/invron/invron.html. "InvRonHelpText.htm" help File for the above-mentioned Program is downloaded with the program.
Eric G.H.Mobsby / R.E.Cox / Roger Sinnott; "Gleanings for ATM's, Photographing the grating," from "A Ronchi zero test for paraboloid". Sky and Telescope, November 1974
original Martys website in german
留言
張貼留言