New Twist on Tilted-Mirror Telescopes.
Because the laws of reflection are colorblind, telescopes with reflective optics are always free of chromatic aberration; they bring light to one focus regardless of its color. Yet the images in most reflectors are compromised in other ways, since the entering light is scattered by the secondary mirror and any support vanes that may be present. However, if the mirrors are tilted, reflecting telescopes can be built that lack central obstructions, and such instruments offer the potential for excellent high-contrast performance.
These TCT (tilted-component telescope) arrangements generally require two or more mirrors, and they often suffer from trade-offs involving the mirror tilts and spacings. In some instances as many as five reflections are required to make the final image plane accessible to an eyepiece or camera. Furthermore, that plane is often tilted with respect to the optical axis, and unless the eyepiece includes this tilt there will be an out-of-focus condition at the top and bottom of the image. In any event, the tilt is tolerable only if less than 10[degree sign]. Nevertheless, TCTs can be configured to yield "diffraction limited" images - the sharpest that can be theoretically achieved by a given telescope aperture - across wide fields of view.
I have recently designed and built a relatively low-cost, compact, and fairly easy-to-make TCT. Since its design is based on mirrors that have roughly the same curvature, the task of mirror making is somewhat simplified. Yet NEWTRYAD (one of my "new try" designs) is a good performer: it minimizes focal-plane tilt and yields very clean star images, images that lie within the diffraction limit of its 6-inch aperture. It's also an enjoyable instrument to use: the eyepiece is aimed along the declination axis, making for convenient viewing, and most observing can be accomplished while seated. I believe my design offers an option to those wanting high performance without the higher cost and longer tube length associated with refractors. The only catch is that you can't go out and buy one: this is a do-it-yourself project.
Deriving the Design
In 1969 Richard Buchroeder presented a TCT design that used three spherical mirrors, two concave, one convex (S&T: December 1969, page 418). The mirrors had the same focal length, easing construction. Access to the image plane was easily gained in Buchroeder's design but at the expense of a relatively large focal-plane tilt. He eventually offered an improved scheme that reduced the tilt to 9[degree sign], but it required a very shallow curve on the third mirror. Builders of this f/20 instrument offer glowing reports on its performance.
Later, Jose Sasian produced a two-mirror unobstructed Newtonian that he introduced to the readers of Sky & Telescope in the March 1991 issue (page 320). This instrument had a focal-plane tilt of less than 8[degree sign] but required a toroidal ("potato chip") figure on the secondary. Although readily produced by advanced amateurs, the toroidal surface was a bit intimidating for me.
Next, enter David Stevick. While fine-tuning his personal TCT ray-trace program he developed an outstanding design now known as the Stevick- Paul telescope. It gave the most remarkable ray-trace* results I had ever seen (ATM Journal 3, spring/summer 1993). Stevick's SPOTPLOT program is available for free at http://bhs.broo.k12.wv.us/homepage/alumni/dstevick/weird.htm, where you will also find Sasian's outstanding TCT design software.
*Ray tracing is the process of mathematically following light as it passes through a hypothetical optical device.
I was so impressed with the Stevick-Paul design that I decided to build a 6-inch unit. However, I chose to gain access to the image plane in a different way. The result was an arrangement of mirrors that became the genesis of NEWTRYAD. I placed a flat mirror between the secondary and tertiary mirrors and wound up with a tube that projected from the bottom of the optics box and contained the tertiary. I liked this alternate arrangement because the tube containing the tertiary could also be used as the declination shaft for an equatorial mount. A side benefit of the flat is that images are oriented "correctly," as they are in a Newtonian, and not mirror-reversed.
Next, I began to play around with other changes to see if I could reduce the focal-plane tilt and maybe further compress the package. To do this I wandered from the Stevick-Paul parameters and changed spacings, tilts, and mirror curvatures while holding the general shape of this "new try" approach. Using Stevick's program, I found my new design after much grunt work and many false starts. Largely by trial and error, I deduced empirical equations relating the mirrors' focal lengths, tilts, and separations. Although this may sound tedious, the process went very quickly owing to the friendly nature of SPOTPLOT. (Those interested in developing their own designs can contact me for details.)
The mirrors in my TCT can have spherical surfaces with no appreciable loss of performance. But an ellipsoidal figure on the primary will reduce the already modest spherical aberration to nearly zero.
Once the mirror configuration was known, I had a packaging problem to solve.
Using computer-aided-design (CAD) software, I came up with a layout that provides unobstructed clearance across a 11/42[degree sign] field of view. This same layout served as the "blueprint" for assembling the optical components.
Mirror Making and More
The materials required to construct NEWTRYAD include three 6-inch Pyrex mirror blanks, a piece of 6-inch-diameter aluminum irrigation pipe, 11/44-inch tempered Masonite for the optics box, and lumber to make the beams and the tube clamp.
If an f/10 to f/13 primary mirror is already available, a matching pair of additional mirrors (one concave, one convex) can yield a design similar to NEWTRYAD. Even slightly varying focal lengths can produce good results. This is exactly what I did, since I had a number of mirrors on hand when I started this project. The primary came from a planetary Newtonian I was using, while the secondary and tertiary came from a Stevick-Paul project that was never completed. (I found the folding flat mirror by patiently shopping on the Internet.)
I assembled a machine to grind and polish the mirrors. (This could just as well have been accomplished by hand, but I like to tinker with machinery.) A pen laser was used to check the surface polish on each mirror; only after its beam passed through the front surface undetected was the polishing deemed complete. I wanted to obtain a very good polish because any residual surface scatter would be contributed by each of the four reflections.
Once a mirror is completely polished, the testing and figuring process begins. These techniques are amply described in numerous books on the subject. Since the telescope will perform very well with spherical mirrors, the simplest test of all can be applied: the null test. As seen with the help of a Foucault knife-edge and slit, a spherical mirror will appear as a uniformly gray disk when examined with this method. The convex mirror can be tested by mating it with the concave primary and looking at the resulting interference patterns.
For the adventuresome in search of optical perfection, the primary's ellipsoidal figure can be very accurately measured with the caustic test. I have a modified test procedure available that is particularly useful for checking the figure of long-focus mirrors, and I invite those interested to contact me for details.
Don't be intimidated by the precision of the specifications in the table below. Those numbers represent the "as built" mirror curves and are not to be construed as fabrication targets. Likewise, the separations shown are not as touchy as one might expect. A little time spent at the eyepiece, adjusting tilts during star testing, goes a long way toward balancing any separation shortcomings. Ray tracing also suggests that the radii of curvature of the mirrors can vary by as much as 20 percent as long as the mechanical layout is adjusted to compensate for any such departures.
NEWTRYAD's design yields a focal plane tilted by slightly less than 6[degree sign]. The eyepiece must be perpendicular to this focal plane to provide sharp focus over the entire field, and for this reason it needs to be a little cocked with respect to the telescope housing. I solved this problem with a pivot arm that places the eyepiece's field lens parallel to the focal plane. Focusing is accomplished by turning a knob located between the eyepiece and the pivot point of the spring-loaded arm. For small focusing excursions the pivot arm moves the eyepiece in a nearly perfect straight line. Parfocal eyepieces are needed to keep the arm positioned in its nominally ideal spot. (I suspect that NEWTRYAD would produce good results even if the eyepieces weren't tilted, but I have yet to try this.)
I had on hand an equatorial mount from a previous project. This mount couples to the tube containing the tertiary mirror. Saddling this tube (a 6-inch-diameter aluminum irrigation pipe) to make it the declination axle allows me to observe, in most cases, while sitting in a chair.
Under the Stars
The optical elements in any telescope must be collimated, or aligned, before the instrument can be put to productive use. This is especially true if the high performance of a complex design is to be realized. On paper, collimating a TCT like mine would appear to be a logistical nightmare, since each of four mirrors can tilt in any direction! Yet experience has shown that collimating NEWTRYAD is surprisingly easy if done in stages.
The preliminary adjustment is done once. An inexpensive pointer-style laser is plugged into a temporary holder at the sky end of the system and centered on the optical axis. The side is then removed from the optical assembly so you can observe the laser beam and make tilt adjustments in accordance with the optical-path layout. First the beam is aimed at the center of the primary mirror. Next, the tilt of the primary is adjusted to send the beam to a dot temporarily applied to the center of the secondary. The process is continued from secondary to flat, from flat to tertiary, and from tertiary to eyepiece. The mirror separations need not be precisely adjusted since the final mirror-tilt adjustments during star testing will compensate for any small errors in mirror separation.
Having accomplished this preliminary collimation, you can do the final tweaking while viewing a reasonably bright star - preferably on a night of good seeing. Using the simple star test described by Harold R. Suiter (S&T: March 1995, page 42), a final adjustment of the secondary screws removes any residual astigmatism not cleared up by the laser adjustment. While stroking the eyepiece in and out of focus, pay close attention to the appearance of the star. Initially the out-of-focus shape will appear elliptical, most likely oriented at an angle to the telescope's line of sight. Moving one of the side-tilt screws will rotate the ellipse until its long dimension points toward the test star. At this point you can turn the top adjusting screw until the ellipse becomes a perfect circle. That's all there is to collimation! These final adjustments are made while sitting at the eyepiece, where the secondary mirror's adjusting screws are within easy reach.
On nights of good seeing the details revealed by this telescope attest to the benefits of unobstructed reflecting optics. It is as though a final veil has been lifted from views of the dynamic surface of Jupiter. As the orbiting moons cross the face of the planet, they remain visible as luminous miniature globes floating above the bands and festoons of the gaseous giant. Colors are particularly crisp; even the surface of our Moon reveals subtle hues I had not noticed previously. Saturn's crepe ring is seen so clearly that I wonder why its discovery took so long.
Most of the objects I observe are fairly bright, but I'm also pleased with the telescope's performance on the Orion Nebula. I can see a surprising amount of structure in the form of wisps surrounding the central Trapezium stars.
Pick Your Size
Unlike some unobstructed reflectors, NEWTRYAD's design scales up or down very nicely. One of my current projects is a compact binocular version, inspired through conversation and correspondence with Richard Buchroeder. It is based on 411/44-inch mirrors and will operate at f/10. On the subject of upscaling, the largest unobstructed telescope in the world is the 60-inch (1.5-meter) McMath solar telescope on Kitt Peak. A 60-inch version of NEWTRYAD would produce a diffraction-limited field of view more than 3 inches in diameter! An altazimuth mount would make it possible to reach the eyepiece without a skyscraper ladder.
However, I doubt the atmosphere would cooperate often enough to take advantage of the diffraction limit at this aperture, particularly at the locations where most amateurs observe. On the other hand, 8-inch, 10-inch, 1211/42-inch, or even 16-inch versions would be quite practical, and, depending on the climate, they may actually produce images that would justify the effort required to build them. Here in Kansas City my 6-inch unit is a practical size, considering the number of available good-seeing nights.
Although my design resulted in a 6-inch f/15 system (which would be bulky by refractor standards), I can easily move NEWTRYAD from my house and set it up in my backyard within five minutes, thanks to its modular construction. The equatorial mount weighs 31 pounds (14 kilograms) and travels separately. The optical assembly, at 43 pounds, attaches to the mount with wing nuts. Finally, the 23-pound counterweight bolts to the mount to complete the setup.
I was able to complete NEWTRYAD in just three months (discounting the design time) and am very pleased with the result. This project owes a lot to the published work by such TCT designers as Anton Kutter, Richard Buchroeder, Art Leonard, Jose Sasian, and David Stevick, as well as numerous others, such as Michael Brunn, whose designs I studied to gain a better understanding of the options.
John Francis is an engineering manager for a material-handling company in the Kansas City area. Readers with further questions on his telescope-design and mirror-testing procedures are invited to write him at 13104 Larsen Street, Overland Park, KS 66213, or at firstname.lastname@example.org.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||constructing a tilted-component telescope|
|Publication:||Sky & Telescope|
|Date:||Jul 1, 1999|
|Previous Article:||Less Magnification for Photography.|
|Next Article:||Subaru Sees First Light.|