Home Science and Industry Astronomy Contact Us



Like many aspects of reflecting telescope construction the exact nature and design of the secondary mirror support system has caused much controversy and debate over the years. Many designs have come forward and these may basically be divided into two types: those that attempt to minimize the diffraction produced by the supporting members and those that attempt to make the diffraction unnoticeable. The most perfect optical system is one that has no obstructions. The very existence of an "edge" causes diffraction, but if this edge is clean and circular the deleterious effects of diffraction can be kept to a minimum; hence one of the reasons for the excellent performance of refractors. Reflectors however, require that their secondary mirrors be supported in some manner and the most common way of accomplishing this is to suspend the secondary mirror from metal strips attached to the sides of the tube. 

The most common form of secondary mirror support is the standard four-vane spider. This design is so universal that it really needs no explanation. Suffice it to say it is generally conceded that the thinner the vanes the less diffraction produced. The amount of diffraction produced appears to be governed by two factors: the overall thickness of the obstruction as well as the linear amount of obstruction. So, it stands to reason that the supporting vanes should be both short as well as thin. This discussion is somewhat complicated by the fact that it has been found that curving the supporting vanes in certain special ways can reduce the apparent impact of diffraction; in other words, make it invisible. Plans for such "diffractionless" spider supports can be found as early as Amateur Telescope Making Book II. In there will be found an article by Andre Couder on constructing a secondary spider having a cross-section so designed as to smear the diffraction over the visible field and make it appear to vanish. This design actually produces more diffraction than the ordinary straight vane type but so scatters it that it apparently becomes invisible. However, it also has the effect of not only brightening the background but actually diffusing fine planetary detail more than would be the case with a conventional spider. It's single virtue appears to be in removing the "objectionable" spider spikes. The fact remains that unless one supports the secondary mirror on a plane parallel window or corrector plate, diffraction will be present in one form or another. The question now remains as to whether it is significant and obtrusive or only an aesthetic annoyance. For a thorough analysis of the subject of spider diffraction, I recommend and article that appeared in Amateur Telescope Making Journal, Volume 11, by H. Richard Suiter with diffraction graphics created William Zmek.

Actually, I have always considered the problem of spider vane diffraction to be somewhat over rated. It is not noticeable in lunar work nor in planetary work and only appears distinctly as fine spikes emanating radially from very bright stars, otherwise one simply cannot find it. In fact, stars less than third or fourth magnitude do not show diffraction spikes at all. Most optical engineers who study the problem are in agreement that as long as the spider vanes are thin and well aligned the impact on fine planetary observing is simply not detectable. As regards the dividing of close binary stars, most of the doubles are far too faint to produce visible diffraction spikes. The only problem appears to occur when one is separating a double in which one of the components is magnitude two or greater in brightness and the companion is dim and might be hidden by an unfortunately placed spike.

Certain extreme diffraction limiting designs have come forward over the years. Back in 1963 when as a young member of the Central Connecticut Amateur Astronomers one of our members, R.R. Willy (one of the first optical engineers to extensively use computer ray racing methods) did some work on designing spider supports made of fine piano wire. As I remember, he concluded that while there were some benefits to be had it simply was not worth the effort involved in producing such a complex support system. I have seen such piano wire spiders at Stellafane. They appear to show up in extreme planetary reflectors where the maker is attempting to draw every ounce of performance out of the optical system.

Perhaps the most simple form of secondary support is a single stalk design. This was favored by telescope manufacturers during the 1960s as a simple and inexpensive way of avoiding the complexities of the conventional spiders support system. In telescopes of small aperture it can actually work rather well, but in larger instruments the single vane needs to be to thick in order to avoid vibration or sagging of the secondary mirror in certain attitudes. Interestingly, Sir William Herschel was fond of the single stalk secondary support in his smaller instruments. The first to use it may have been John Hadley, the inventor of the first truly practical reflector. In the present day, when simplicity, no matter how crude, is more appreciated than ever, single stalk supports made of thick pieces of wood only roughly sanded have actually been recommended in construction books. I assure you, that this form of single stalk support will cause a lot of diffraction.

I once saw spider support vanes on a Stellafane telescope made of wood covered with (I think) flocked paper. No... not a good idea.

As to the material from which spider vanes should be constructed, smooth metal is probably the best. Brass is an excellent material and suitable brass strips of exactly the correct thickness and width can be found in hobby shops just waiting to be cut to length. The use of wood is to be avoided. I mention this only because I have seen it used so many times. Brass is a soft and malleable material that works and solders easily. When drilling into brass extreme caution should be exercised inasmuch as twist drills can grab and tear the brass out of your hands - and possibly a good deal of flesh along with it. If you have to drill a large hole drill a smaller hole first and then carefully drill the larger hole. Use a drill press if possible and machinist's vice. Aluminum can behave in much the same way.

My own studies into the subject have suggested to me several things that are important in spider vane construction. One, that the spider vanes be very carefully aligned and not be twisted so as not to increase the apparent width of the vane. This can be extremely detrimental to optical performance. A 32 thousandths strip of metal can easily twist to become 1/8 of an inch. Two, that if a four vane spider is used that care be taken to insure that the opposing vanes be carefully aligned so as not to create a double vane. It is for this reason that I tend to prefer a three vane support over the four vane support.

My own favorite secondary support holder has come to be a two vane design which I came up with in 1990 when designing a compound reflector having both Newtonian and Gregorian functions. I wanted the secondary mirrors supported on plates or panels that could be easily attached and detached from the telescope tube; thus allowing easy conversion from one optical system to another. It became apparent that the only way this could be accomplished was to support the secondary mirrors on two vanes so the secondary mirror could be withdrawn through a large hole in the side of the tube. The design is pictured at the top of this page and in additional pictures below. Construction is extremely simple and is essentially a triangular design with the apex of the triangle formed by a "long nut" or joining nut. This item is a standard nut which has simply been elongated so as to join two threaded rods. These are found in most hardware stores. The base of the triangle is formed by a strip of brass. This base plate (or strip) is drilled so as to allow mounting to the tube. The sides of a triangle are the actual supporting vanes running from the base plate to the long nut. The long nut is convenient in that it can receive a threaded rod from a conventional diagonal mirror older. This is an extremely simple support structure to build. For an 8 or 10 inch reflector the base plate would be made out of brass strip 1" wide by 64 thousandths thick. The supporting vanes would be made out of brass strip 3/4" wide by 32 thousandths thick. The whole affair is simply soldered together with standard soft solder. A wooden jig can be built so as to properly and accurately align the parts until soldering is complete. Variations of this design can include the elimination of the base strip (see picture below) so as to allow for mounting in circular tubes. Or, perhaps better yet, the base strip can simply be bent to the curvature of the tube. Many variations are possible. The nice part about this design is that it can be removed easily and replaced without a lot of fuss and bother. As far as centering adjustment is concerned, the tube should be accurately measured and the spider support designed to be a trifle short so it can be built up with one or two washers as needed. Elongated mounting holes can allow for lateral adjustment. 

Rigidity is maintained by the triangular shape and thickness of the vanes in one dimension and the breadth of the vanes in the opposing dimension. While the use of a third opposing vane could allow for thinner vanes this would itself add to the diffraction. The observable net result appears to be the same with the added advantage of the two vane design being simplicity of construction, ease of adjustment, and, in some cases, expanding the potential for mounting in new and different tube designs.

I have made several telescopes using variations of this design and all have worked out extremely well. I would say that this design can be made to work with telescopes is larger as 12". In no case has vibration been a problem.


As mounted in my 10" f/8 rotating top

Oh yes, about those diagonal cells (like the one pictured at the top of the page) that use three set screws to adjust the tilt of the mirror; if you relax one screw the whole affair becomes loose and you have to tighten one of the other screws (take your choice), making the diagonal invariably point in a direction you don't want. And so you go around and around, loosening and tightening. A trick that solves this problem is to cut a little circle of 1/8" thick inner tube rubber and place it between  the central socket and the ball, thus creating a compressible and yielding structure. Now you can tighten or loosen one screw without having to adjust any of the others. Works like a charm.