摘自astromart"
Testing Your Telescope Indoors
By Darren Drake - 3/29/2007
Have you ever received a new prized telescope only to have cloudy skies for what seems to be weeks at a time? We've all felt the desire to get out with our new scopes and put it through its paces only to have miserable weather night after night. Often when it is clear the seeing makes stars look like bloated balls and star testing is all but impossible. Well it's possible that you may already have the materials necessary to star test your new scope indoors under controlled conditions in perfect seeing with star images at infinity. It's also possible that you can with a similar test but with the use of a knife-edge test the optics in your prized scope with a very sensitive test that you may have thought only possible in an optical shop with very expensive equipment.
The key to this sensitive test is the use of an artificial star but with a twist. Ordinarily we use our telescopes to bring parallel light from a point source to converge to a perfect point at focus. If however the reverse is done, that is to place a point source at the focus then light will bounce off the primary mirror or pass through the lens and come out being perfectly parallel provided the main optic is of high quality. The result is an artificial star is produced at infinity for anyone literally looking into the objective but with essentially perfect seeing. This allows some level of optical evaluation to be conducted indoors under controlled conditions. The key to being able to do this requires that we already have another scope known to be of high optical quality, or a control telescope, and the ability to make a very high quality artificial star that can be inserted into your scope's focuser.
How to make a focuser insertable artificial star
Most commonly available artificial stars such as the one from EZ Telescopes alone are not good enough as the light source is too large. If however some optical elements such as an eyepiece or two are placed in front of the light source the effective diameter can be significantly reduced.
In my case I used 2 inexpensive .965-inch eyepieces. One eyepiece had it's barrel removed and the optics were taped right next to the hole opening of the artificial star and the other was placed about 3 inches away on the other end of the tele-extender. Duct tape was used to hold everything in place to assemble a focuser insertable artificial star. I found that the best compromise between brightness and sensitivity is an artificial star that is about 3-5 arcseconds in angular diameter. If the resulting star image is too large, try inserting a Barlow lens in the optical path. It will further reduce the size of the star in increase sensitivity. It's important to
make sure that the eyepieces used are clean and free of dust or imperfections or blemishes will interfere with the resulting test images.
Star test your scope indoors at infinity
To star test a telescope indoors, the tester needs to locate the exact position the focuser needs to be at to simulate infinity for one of the two scopes being used in the test. To do this, take a casual look through the other scope and focus at infinity at a distant object. Then literally aim that scope right into the objective of the other scope that has the artificial star placed into it and an out of focus star should be seen once aimed properly. It's then necessary to focus the scope with the artificial star in it so that the image through the scope with the eyepiece is focused or nearly focused. This assures that the star is EFFECTIVELY at infinity. It's best to somehow record the approximate position of the focuser so that this procedure only needs to be done once. Then a normal star test can be conducted just as outside but with no thermal issues. If the diffraction patterns appear fuzzy or ill defined then the artificial star is still too large and a different combination of lenses should be used. Either smaller focal length eyepieces or a greater number of eyepieces should work. The smaller the airy disk produced, the more sensitive the test but at the cost of having a dimmer light source. It needs to be mentioned here that it is essential that the control telescope needs to be of high optical quality AND properly collimated for this or the outgoing light may not be perfectly parallel.
Optics 101 and the null test
For most telescope enthusiasts, this might be as far as one would desire to go but it is possible to take the optical testing to the next level. To explain further, I need to digress to another type of optical test that very few amateurs actually do on their scopes: the null test. When an optician is making a parabaloidal telescope mirror one of the steps is to make the glass a sphere. At this point, the Foucault image shows that the glass has no correction and looks essentially flat with no apparent hills or valleys. The image shows a perfect null and deviations from this can easily be seen. Then the optician goes on to add correction until the measurements show it to have all the right deviations from sphere at the right places, which can be very difficult to verify unless the tester is experienced.
Other testing methods can be set up in the optical shop so as to show a perfect parabaloid as a perfect null, which is highly desirable because any deviations from optical perfection should ideally show up clearly in a 3D shadowgram. But large aperture precision optical flats and/or precision null lenses are needed for this and the costs and difficulties in setup are very high. So most opticians rely on their ability to interpret Foucault images they see before deciding that a mirror is done and for many inexperienced glass pushers they may not have a very good optic but will have to unfortunately wait until they see the questionable images before knowing for sure what they have. If a knife-edge is placed in the focuser of a completed scope while focused on a bright star then, because the point source is at infinity, a perfect parabaloidal mirror will show that perfect null that is so much more easily interpreted than a with a Foucault test in the optical shop. To set up this test, an accurate knife-edge must first be placed in the focuser.
How to make a focuser insertable knife-edge
It is important to have an accurate and very straight knife-edge that cuts right through the halfway point of the focuser. To do this, I simply used a typical collimating device with a small hole in the center and a slice of aluminum from an old floppy disk. The two can be permanently combined with two-sided tape. It is important that the knife-edge be perfectly straight with no jagged edges as seen with a strong magnifying glass. If it isn't, the accuracy of the test will be affected. It's also important for the knife-edge to travel right through the halfway point of the hole so as to cut right through the center of focus.
Conducting the null test
To conduct a null test, simply place the knife-edge into the focuser of any scope being tested. Then while aiming at Polaris or another brighter star if the scope has tracking, the observer will see half of the mirror illuminated while the other appears dark. The edge or boundary between light and dark will appear sharp or blurry. The scope should then be placed so that the light/dark boundary goes EXACTLY through the center of the mirror. As the focuser is adjusted, the intent is to see the edge get increasingly blurry until the light and dark side becomes flipped. At this point, the focuser has gone to far and needs to be reversed just a bit until the mirror appears equally illuminated but somewhat dimmed.
For this to work, the scope needs to be precisely aimed at the star and the knife-edge precisely cutting through the airy disk which requires very high pointing and adjusting accuracy to the point where it may require patience and experience to get it right. To verify that the desired result is being achieved, the tester can place his or her hand in front of the scope and look for very prominent 3 dimensional heat waves emanating from the silhouette. Once the knife-edge is properly placed, any deviations from a perfect parabaloid should in theory be clearly visible in 3 dimensions as brighter or darker zones. If there is some difficulty achieving the null, try rotating the knife-edge slightly so as to make fine adjustments as the knife-edge slowly approaches the airy disk and allows for precise positioning.
While this test can be very revealing, it does have one major flaw: it requires nearly perfect seeing to achieve the desired result. Tube currents and a turbulent atmosphere can overwhelm the shadow patterns and reduce the sensitivity considerably.
Combining the indoor artificial star concept and the null test
By using both the artificial star concept along with the null test, it becomes possible to set up a genuine null test that can be used to test a large aperture scope or smaller scopes depending on which one is KNOWN to be of high quality. In my particular instance, I have an 8-inch f/6 Meade Starfinder that shows a VERY good null on Polaris. Also, when examining the mirror with a Foucault tester and putting the numbers in Figure XP, the mirror is found to have a .998-strehl ratio with a rating of 1/16 wave at the wavefront and a very smooth surface.
Given that the optics are essentially perfect, I can use this as the control scope to "look" at the mirror in my bigger or smaller scopes. I happen to have an 18-inch dob along with a back up 17.5-inch mirror of the same focal length. I know both mirrors have tested in the .95-strehl range and give terrific images but now there is a way to examine the mirrors using a null test under controlled conditions using the 8-inch. To do this I simply placed the artificial star in the focuser of the larger scope and the knife-edge in the focuser of the 8-inch and looking at various parts of the bigger mirror while looking through the knife-edge. Even though the control mirror is the smaller one it can be used to test the larger one because the very edge of the secondary shadow and the edge of the big mirror is visible or nearly so. The result is that a full profile can be seen of all the zones in the big mirror. Since both surfaces are contributing to the null test images equally, it is actually possible to "reverse the roles" and actually look through the focuser of the big scope with the artificial star in the smaller scope.
Fortunately and not unexpectedly, both big mirrors I have show a virtually perfect null and therefore I feel confident in using the big scope as the control scope for all tests as the mirrors are more than good enough to reveal problems in smaller scopes. Since there is some uncertainty in what "profile " a given optic may reveal if there is a deviation from a perfect null it may be unclear as to which mirror that imperfection is in. All one has to do is observe the shadow pattern observed and determine which circumference the deviation appears to involve. For example if one is examining a portion of a big mirror with a smaller aperture scope and the shadow pattern appears to follow along the circumference of the big mirror it can be concluded that the error is in the bigger mirror. If the errors appear to follow along the smaller mirror's circumference the error is in the smaller scope. If a perfect null is seen it can be concluded pretty much for certain that BOTH surfaces are good. Using this procedure, it should be possible to see such optical errors as spherical aberration, surface roughness, lap marks, astigmatism, and turned edges. If there is a question of a mirrors edge one can specifically look at that part of the mirror and see if there is a fuzzy or darkened ring near the edge, which would indicate a turned edge. Since the light source can be quite dim, it is best to do this test under darkened conditions.
It should be mentioned here that in order to do this type of testing it is essential to have no floor vibrations or air currents in the room. The focuser that has the knife-edge in place needs to be adjusted to within about 1/1000 of an inch or so of the focal point in order to work. This again may take some time and patience to be able to achieve consistently. The test if set up properly will reveal sharp and contrasty heat currents from any heat source such as a persons hand. If heat currents are not visible when a heat source is put into the optical path, the knife-edge is not properly positioned.
What the null test can reveal
If an overall null cannot be achieved, then some order of spherical aberration may be in the system. If while moving the focuser inward the center darkens first, then there is undercorrection. If the edge darkens first, then there is overcorrection. It is also possible that since the entire telescope is being tested that a bad or miscollimated secondary can affect the results. If a refractor is being tested and the results are uncertain or questionable, the secondary should be removed so that only the main optic is being tested. If an achromatic refractor is being tested, it's best to place a green filter in the optical path so as to reduce confusing chromatic effects.
One scope I recently tested for example was my Orion 80 ED. Even though images through this scope are quite good, star testing reveals that intra and extra focal images are not quite identical. Using the knife-edge test with the 18-inch as the control scope, I see that there is a slight zone appearing as a raised hill at around the 75-85 % radius that could explain the imperfection in the star test.
Using this method can reveal other things as well. For example it is possible to experiment with ways to control the boundary layer above the mirror of a large Newtonian. It is possible to see the exact effects various fans and heating elements can have on the optical path. Blowing air from a house fan for example causes considerable turbulence even though the temperature is constant because the moving air has varying densities as it moves turbulently though the environment. The effects of the secondary heater can also be seen as a calm rising plume of air as it moves straight up over the secondary.
Using the null test on an unfinished mirror
For mirror makers, it is possible to take this yet another step further. If there is some uncertainty in exactly what state an unfinished mirror is in, or the optician simply wants to verify that the mirror is done, he/she can place the Foucault tester's light source at EXACTLY the halfway point of the radius of curve, which is of course the focal point. By placing a high quality telescope right behind the Foucault tester, the optician should then be able to achieve a null test with the knife-edge in the scope's focuser but with the silhouette of the testing devise in the light path. This in effect achieves the ultimate test for mirror makers but without the need for expensive optical flats or null lenses.
Concluding remarks
With some patience and experience, using this method makes it possible to accurately evaluate a telescope's optics by using the star test so many of us know well. In addition, the use of the indoor null test allows observers to evaluate their prized optics with a sensitivity and accuracy once thought to only be found in high-end optical shops. Good luck and happy testing. |
置顶卡
变色卡
千斤顶
显身卡
全部都是外文?
用金山就很没意思了