Building an open-tubed 12.5" F/12.5 Classical Cassegrain

By Michael E. Lockwood
assembled
Figure 1: Telescope, equatorial platform, and support platform,
waiting for darkness in the fields of Illinois.

Introduction

I've enjoyed watching the sky ever since I was a child, growing up in the dark woods of northern Michigan.  My father has always been a builder of things (houses, hobby projects, etc.), and I am the same way - I am always creating something.  So, building telescopes naturally grew out of my love of the sky.  With some help, I built three telescopes in high school, and used them quite a bit.  After finishing college and grad school and some house rennovations, in early 2002 I decided to finish a mirror making project that I had started in 1989.  With a little advise, it came out quite well.  My first mirror, an 8" F/3.9, is something that I will never sell.

Fast forward three years - after making many Newtonian telescopes and after grinding, polishing and figuring about 20 parabolic primary mirrors (by hand), Mike Conron and I made a 30" F/3.8 mirror with the aid of his grinding machine.  We did the final figuring by hand.  I had figured some very good fast mirrors before, so I ended up taking charge of the figuring of this very big, very fast mirror.  (Mike C. tended the machine through virtually all the grinding and polishing.)  We put in a lot of time, and had a lot of fun (not to mention a few beers!), and when that mirror was done I was thrilled.  (See the end of the article for a link to my web site, which has a section on the 30".)

But what could be the next challenging project after making such a big, fast mirror?

My friend decided to thank me for the figuring help by giving me a full-thickness, Pyrex, perforated 12.5" F/4.2 Cassegrain primary that needed refiguring.  That was the start of the project.  Up to that point, I'd only seen one amateur-made Cass, and one non-commercial cass made by a professional optician, so I figured it must be a challenging project to get both the optics and the mechanics right.

I like a challenge.

Additionally, this type/size of long focus telescope can be used with your feet firmly planted on the ground (no ladders), and often viewing can be done from a VERY comfortable seated position.  More detail can be seen when you're comfortable.  Putting the whole scope on a tracking platform would further increase viewing enjoyment, and would make waiting for that second of steady seeing much more bearable.  The price for the convenience of the viewing position is that you must make a highly precise convex mirror.

Cassegrains do have disadvantages - they can be difficult to baffle, collimation requires a laser collimator and practice, there are more optical surfaces to scatter light, and their mechanical construction is not as easy as a Newtonian.  I chose not to worry about the disadvantages, and focus on the positive aspects of the design.  I wanted to use the scope and judge it for myself at that point, not before.

After a lesson in the equations that define the design of a Cassegrain telescope, I wrote a short, simple computer program to calculate several designs with different focal ratios.  Taking into account the fact that I wanted the focuser to sit on top of the mirror box, rather than behind the primary, I settled on a design with a focal ratio of F/12.5, about as fast as was practical given the primary's focal ratio (which I wasn't about to change!), a secondary of 4" in diameter, and a 32% obstruction ratio.  The fully illuminated field has a diameter of 0.6".  A 1.53" elliptical tertiary mirror, the same type as is used as a secondary mirror in a Newtonian, was located just in front of the primary mirror, and sent the light path out the top of the mirror box to the focuser.  That is, instead of the light passing through the hole in the primary mirror, it is reflected upward into a focuser placed on top of the mirror box.  This places the eyepiece in the comfortable viewing position.

diagram
Figure 2: Telescope diagram - green lines represent light path.
Primary mirror is concave, secondary mirror is convex, tertiary
mirror is flat.  Not to scale.


Another advantage of the Cassegrain design is the long focal length - high magnification is possible with longer focal length eyepieces with more eye relief, and Jupiter's moons are disks, not the "comets" that they become in a fast Newtonian.  (Many do not realize that the performance of a perfectly collimated F/4.5 Newtonian will be limited by coma if an object, such as a planet, is not well-centered in the field of view.)  Magnifications of upwards of 400x are achievable with a 9mm Nagler.  On the other hand, A 41mm Panoptic provides 96x, which is low enough to make finding objects fairly easy, and viewing the entire moon possible.  No vignetting is noticeable with this eyepiece, even with its enormous field stop!

So, here's a quick summary of why I did the project:
1)  Challenge of fabricating a different (non-Newtonian) optical system
2)  Lack of amateur-made Cassegrains
3)  Convenient viewing position
4)  Compact telescope
5)  Long focus - high magnification is easy, little coma
6)  I already had a primary mirror blank that only needed refiguring
My goals were
1) High optical quality, minimal spider diffraction
2) Ease of transport and assembly/disassembly, and
3) A unique telescope that might inspire others to make a Cassegrain.
At some point you might also wonder what tools I used.  Most are fairly standard woodworking tools.  Here's a list of the power tools:
    10" tablesaw, 12" chop saw, router, jigsaw, drill, drill press, orbital sander, biscuit jointer.

(I don't have a lathe or mill....... yet!)

Optical Fabrication

I'm an amateur optician, working by hand.  I use the Foucault test to test concave mirrors.

The tester was built in an evening, and costs $20 or so, including a $10 dial indicator.  Nothing fancy here.  It is pictured below.  The "knife edge" is actually a 1/16" x 1" piece of brass stock a few inches long.  The edge is prepared with careful filing and lapping against a ground glass plate (see Texereau's "How to Make a Telescope" for this procedure) to get a very straight, slightly rounded edge.  Such a tester, with careful use, is capable of testing optics to 1/10th wave wavefront accuracy,  or about a millionth of an inch on the glass.  This is my third Foucault tester.

slitless tester
Figure 3: A slitless Foucault tester for testing concave mirrors.

This tester was used to measure the shape, or "figure" of the primary mirror and the test plate for the secondary mirror (more on this below).

The primary mirror refiguring was no different than a normal parabolic mirror, except that it had a hole in the center!  The subdiameter lap that I made was made just large enough that it could be pressed between the hole and the edge of the mirror.  At the beginning of a figuring spell, the lap was used gently after the press until it no longer grabbed when passing over the central hole.  The severely undercorrected outer zones were corrected gradually, and the mirror was soon done with a smooth figure accurate to significantly better than 1/10th wave wavefront error, according to my measurements.  Focault testing was done with a six-zone Couder mask.  Here's a photo of the coating being stripped off the primary mirror before refiguring, and of the subdiameter lap being pressed into contact.

coating off lap press
Figure 4a: Old coating being stripped.  Masking tape dam keeps the
chemicals from running off the sides, or down the hole.

Figure 4b: Subdiameter lap being pressed under water jug and weight.

The secondary was the critical component in the scope.  It had to have the right curvature so that the focal plane was located in the proper place.  This is important - moving the secondary in and out can move the focal plane around and give more or less back focus if necessary, but one should not deviate too far from the ideal spacing.  Moving the secondary too far inward towards the primary results in undercorrection.  Moving it too far out results in overcorrection.  For this adjustment, 1/8" of change in the spacing makes a big difference in star testing results.  The moral of the story is, get the secondary curvature right, or you will have to redesign your scope!

The secondary mirror and grinding tool both began as 4" Pyrex blanks.  With grinding, gradually one becomes concave (the test plate), and the other convex (the secondary mirror).

Because I wanted to make more Cassegrain secondaries in the future and other small mirrors, too, I built a small turntable that I could sit comfortably at while grinding or polishing.  Cost was under $150.  A geared-down 1/3 hp DC motor drives the turntable via a belt.  The motor is powered via a variable transformer (a Variac) and a full-bridge rectifier, which converts the 120VAC from the wall to anywhere from 0 to about 90VDC.  I wired it up myself.  This homemade speed controller spins the turntable anywhere from 0 to about 250 RPM.  (I plan to add a driven arm to this machine very soon, so it can polish mirrors up to about 13" in diameter for me while I watch!)  As pictured, though, it has a manual overarm.  I hold onto the arm and move it back and forth while the turntable turns the work.  The top of the machine is a piece of formica-covered countertop that I had on hand, and the red splash pan is a water tub for farm animals, purchased for a few dollars at Farm and Fleet.

turntable
Figure 5: The turntable, with manual overarm, polishing a 5" mirror.
All equipment on top of the machine is vital!

I ground the secondary mirror and test plate from 120 grit Carbo down to 5 micron Aluminum Oxide, measuring their curvatures frequently with a crude spherometer made from a pulley, some ball bearings, and an inexpensive dial indicator.   The concave tool later became a test plate, a polished optical surface that is figured to a concave hyperboloid that perfectly matches the convex hyperboloid that the secondary will become.  Below is a photo of the spherometer, sitting on a fairly flat surplus optical window, zeroed and ready for measurement of the curvature of mirrors.  The large bolt is tightened by hand, and holds the dial indicator stationary in the bore of the pulley.  To the right are both 4" and 5" test plates and secondary mirrors.  The 4" became the secondary for this telescope.  (The 5" secondary later went into our club's 16.25" F/12.5 Cassegrain - it was another large project in it's own right.  See my web site link at the end of the article for more info on that scope.)

spherometer
Figure 6: Spherometer being zeroed, future 4" and 5" secondary
mirrors and test plates.  (The 5" was for another telescope.)


The concave test plate can be tested using the Foucault test, and figured just like a small primary mirror, however it has several times the correction of a parabola of comparable size and focal ratio.   (For mirror makers, this one had about four times the correction of a 4" F/6 parabola - not too bad.)  I finished figuring the test plate in a few evenings of work, and I noted that I had a small, slightly depressed ring just inside the edge of the mirror.  This would be important later, when I was trying to make the secondary mirror match the test plate to within a millionth of an inch.

Once the test plate was done, I had to polish the secondary mirror out without having the curve drift too far from the curvature of the already-finished test plate.  Using the machine, the secondary polished out less convex than it should have been, so I had to use a special lap to wear down the outer part of the mirror until the curvatures matched approximately.  The lap had some of the pitch carved out from the center, so that the edge would wear more than the center, thereby making it more convex.

After polishing sessions, the curvature of the secondary was compared to that of the test plate (both polished) using a simple Newton interferometer, which consists of a few compact fluorescent lights, a diffuser screen, a red filter (for lighting effects), and a two way mirror.  This is homemade gear, costing about $40 to build from plywood.  It is NOT an expensive commercial interferometer.  To perform a test to see how the mirror and test plate fit, the mirror and test plate are cleaned, and then placed in contact with each other under the interferometer.  They settle into intimate contact if they are clean, separated by an air gap only a few wavelengths of light thick, or about 100 millionths of an inch.  (That's why they must be clean!)

Newton interferometer
Figure 7: Diagram of simple Newton interferometer.  Convex mirror
sits on top of concave test plate, separated by a thin air gap.

The simple setup produces a fairly good source of diffused monochromatic light, which allows the differences between the secondary and test plate to be observed in terms of interference "fringes", the dark bands seen between bands of red below.  The spacing between each dark fringe represents 1/2 wave of surface error (about 14 millionths of an inch for the red light used here), which is one wave of wavefront error.  The fringes are viewed from several feet away to avoid parallax errors, which distort the fringes if you view them from too close to the tester.

In a nutshell, curved fringes mean there is work still to be done - they mean that the mirror and test plate don't match exactly.  The appearances of the fringes themselves means that the top piece is "tilted" with respect to the bottom piece.  It is easy to observe the "squiggles" and curves in the fringes.  In the photos below, where the lines curving to the right represent a high spot.  Thus, in the first photo, a high ring exists at approximately the 70% zone, that is, 70% of the way from the center to the edge.  (This is normal in figuring a Cassegrian secondary.)  In the second photo, the outer areas are high.  Photos are taken at a distance with a digital camera - this is exactly how it looks to your eye when looking into the tester.

fringes1 fringes2
Figures 8a, 8b: Photos of interference fringes during testing.
Fringes curving to the right indicate high spots.


fringes3
Figure 8c: Interference test for finished mirror.

The fringes gradually get straighter, and the final result is shown in Figure 7c.  The larger "wiggle" in the fringes about 3/8" inside the edge is due to a depression on the test plate, not the secondary, so the figure of the mirror turned out quite well.  A perfect mirror/test plate match would show perfectly straight fringes.  The test plate is safely packed away in a closet, ready for use if I ever need a secondary mirror with the same specifications.

With both mirrors finished, the optics were off to L&L Optical Services for coating with enhanced aluminum.  For this type of Cassegrain, which has three mirrors, having 95% reflectivity (enhanced alumimum) on each optical surface provides about 86% light throughput, versus 70% throughput for three mirrors with 89% reflectivity (standard aluminum).  That's a big (almost 20%) difference, and worth the extra cost.

Telescope Construction

While my primary goals for this telescope were high optical quality and a convenient viewing position, I also wanted a telescope that I could easily disassemble, and which would not take up too much space.  The solution for me was an open tube with oversized tubes at the four corners of the mirror box.

The mirror box was made of 1/2" baltic birch plywood.  It features two baffles, one each near the top (front) and bottom of the box, which fit in dadoes (slots) that are cut in the inside of the box sides.  The box sides are joined at the corners with biscuit joints - slots are cut in the wood which accept small wooden "biscuits".  These biscuits align the joint and strengthen it.  When glued with good quality wood glue, the biscuits and dadoes form a very strong box with no other fasteners required.  This means no screw or nail holes to patch.

Here's a photo of the mirror box, viewed from the rear, assembled without glue to check the fit of the parts.  Note the holes in the front baffle to accomodate the 1.5" diameter aluminum poles.  The edges of the holes don't line up exactly with the insides of the box - this bends the poles slightly when the scope is assembled, and thereby creating a stiffer tube structure.  The poles are stuffed with styrofoam peanuts to damp vibration, and have threaded inserts in each end.  The threaded inserts screw on to four bolts inserted in the the back baffle in the mirror box.  The heads of the bolts are visible at the four corners of the back baffle (nearest in this photo).

box1
Figure 9a: Mirror box partially assembled.  Top of box has been removed.

Here is another photo of the box, again without the top, with the poles inserted and threaded onto the bolts.  The metal bracket at the top of the back baffle serves to reinforce the attachment point for the mirror cell.

box2
Figure 9b: Mirror box partially assembled.  Poles installed.

The mirror cell was made mostly of 3/4" strips of baltic birch plywood, laminated together in places for strength.  Its triangular frame is attached to the back baffle in the mirror box (above) with a bolt and a type of rod end that allows a bolt passing through it to pivot with very little mechanical slop.  The bottom two attachment points also employ the same part of rod end, but these points are movable to allow for collimation.  Threaded knobs allow tweaking of the primary collimation at the eyepiece, using a star.

The primary mirror is supported with a four-point mirror cell.  A photo of the nearly finished, painted mirror cell is below.  The top two contact points (white plastic squares) are mounted on a piece of aluminum tubing which is supported at the center - this is effectively a teeter-totter.  The bottom two points are fixed.  This supports the full-thickness mirror enough to prevent detectable flexure of the primary mirror.  The two L-brackets that support the bottom of the mirror were later replaced posts made of 1.5" diameter Nylon.  A teflon contact point was added to the top bracket, and PVC mirror clips were fitted to all three brackets to keep the mirror in the cell.  In this type of cell, the mirror clips move with the mirror, so the mirror need not have lots of room to rattle around.

cell
Figure 10: Mirror cell.  4-pt. support is enough for the full-thickness primary.

My goals for the secondary cage were light weight, low wind profile, and minimal diffraction caused by the secondary support.  This meant using a spider that was made of thin wire, something I had never seen done with a Cassegrain telescope.  Additionally, the spider attachment points on the central hub are offset to resist rotational vibration of the secondary.  (The offset spider concept was introduced decades ago in Texereau's book, "How to Make a Telescope"; still, few people realize its advantage.)

For the secondary cage, I chose to make a square front board with a circular hole to define the telescope aperture, reinforced with an octagonal frame that would serve to anchor the spider.  The four poles attach at the corner of the squares, sitting in recessed holes.  The secondary cage assembly was glued together, and this proved fairly complicated.  Here's a picture while the glue dried.

cage glue
Figure 11: Assembly of secondary cage.  You can never have too many clamps....

The forest of clamps was removed, and holes were later drilled in the octagonal frame to lighten it and allow wind to pass through.  Before this, though, the holes for the four stainless eye bolts that anchor the spider were drilled.  As the spider is of the offset variety, these holes are not equally spaced around the octagon.

The central hub consisted of two 3.5" x 1" x 0.25" aluminum bar, connected by two stainless bolts which were threaded all the way to the countersunk head.  The bars were locked onto the bolts with nuts.  Four small screws serve as anchors for the ends of the two pieces of wire that form the spider.  A piece of baltic birch is attached to the aluminum bar.  Three adjustment screws pass through the wood and secure the aluminum secondary holder. Here is a closeup of the central hub.  The front is at the bottom.

hub
Figure 12: Central hub of the wire spider.  Secondary mirror
attaches to wood block at top.


To string the spider, a jig was made consisting of one strip of baltic birch, with holes drilled so that the central hub could be securely fastened to it.  Stainless steel wire, 0.015" thick, was secured under one screw and washer on the front of the central hub assembly.  The screw was tightened securely, and the wire was threaded through one of the eye bolts on the front of the central hub.  It then passed through an eye bolt on the secondary cage (wrapping completely around it once), back to the central hub and through the eyebolt on the back side of the central hub, through another eye bolt on the secondary cage, and finally through the eyebolt on the back of the central hub.  The wire was pulled fairly tight, and secured under another screw and washer.  This formed the bottom two legs of the offset wire spider.  The sequence was repeated for the top two legs, with another piece of wire.  The wire was then de-tentioned by loosening the eye bolts on the secondary cage, the jig removed, and the wire re-tightened.  Thus, the spider was completed.  Here's a photo of the secondary cage, central hub, and spider jig before stringing.

spider jig
Figure 13: Secondary cage and spider stringing jig.  Ready for wire.

The next photo shows a closeup view of how the wire was wound and secured on the central hub.  Where the wire passed through the screw eyes on the cage (seen in the photo above), it was wrapped once around the eye before going back to the hub.  The two black knobs are used to adjust the tilt of the secondary holder.  Compressible rubber washers are placed between the aluminum disk seen in the photo, which is part of the secondary holder, and the piece of wood.  Later, a strip of aluminum was bent around the aluminum disk, a lip was bent over on one side, and it was secured to the disk with screws.  The secondary sits inside this assembly, with polyester batting behind it to press it gently against the lip, much as is done with many secondary holders in Newtonian telescopes.

wire closeup
Figure 14: Closeup of central hub, suspended by wire spider.

Finally, here's a photo of the finished secondary cage, complete with painted secondary holder, and a painted and stained cage.  The blue knobs attach the poles (by threading into the threaded inserts in the tubes), to which the cage is attached in the photo.  Note holes in the sides for weight reduction.

cage
Figure 15: Finished secondary cage.  Blue knobs thread into inserts in
the poles.  Unpainted aluminum in hub is not visible to the optics.

For the tertiary mirror holder, I wrapped a thin piece of aluminum sheet around a 1.5" dowel, and used a chop saw to cut it off at a 45-degree angle.  The aluminum was slid down the dowel a little bit, and a lip was created around the 45-degree hole.  This would hold the tertiary mirror in.  A metal plug was inserted into the other end of the aluminum, and a bolt passed through and angle bracked and into the plug, securing it to the tertiary assembly.  The tertiary assembly tilt is adjusted via sections of threaded rod that can be accessed from the rear of the telescope.  These act as push bolts, and there is a central pull bolt attached to the tertiary holder assembly.  (Tyipcally, this mirror is quite stable and needs no adjustment.)  A photo of the tertiary holder is shown below, installed in the scope, in front of the slightly dusty primary.  The round disk attached to the two standoffs is a ring baffle, which blocks stray light from reaching the tertiary.  This is much simpler than a traditional round baffle, and is much easier to implement.  A classic "pipe" type baffle would probably work slightly better, but its inside must be carefully baffled to prevent reflections at grazing incidence.  For my use, this baffle works well.

tertiary
Figure 16: Tertiary mirror in 45-degree holder.  Ring baffle is to
the left.  Light from the secondary passes through the hole in the
ring.  Disk behind tertiary is plexiglass, painted black.

The various wooden parts of the telescope were stained and polyurethaned.  Internal parts were painted Krylon Ultra-Flat Black.  Teflon and formica were installed on the rocker box, and the side bearings were bolted on the mirror box.  The side bearings are 1" thick, composed of two thicknesses of 1/2" baltic birch.  1/8" x 1" aluminum bar stock was bent around the outside edge with waterproof glue underneath.  Screws secure the bearing at various points around the circle.  The glue fills in any gaps under the aluminum, filling in any slight void and eliminating any tendency of the scope to "bounce" when move small amounts in elevation.  Below the masking tape layout is shown for the side bearings, with the balance point of the scope marked.  Four bolts secured each side bearing.

side bearing
Figure 17: Marking the balance point and installing the side bearings

The focuser is a blue-anodized 2" Crayford model from Moonlite.  It sits on top of a wooden platform on top of the mirror box.  It is smooth in operation, and the blue color matches the blue of the tubes and the collimation knobs.  This blue color goes well with the golden color of the woodwork.  The focuser can be seen in the photos of the assembled scope below.

First Light

First light happenned in my driveway near the beginning of April of 2005.  I set the rocker box on a piece of plywood, which was placed directly on the concrete of my driveway.  I knelt to look in the eyepiece, and located a bright, poorly collimated image of the waxing moon.  Two friends looked too.  I realized it was time for some serious collimation work, installation of light baffles, and a folding platform to privide that convenient viewing position that I was so fond of.  The equatorial platform arrived soon afterward, so it was time to build a support platform for it.

Folding Platform

The folding platform was my idea and creation.  It folds up (to about 24" x 20" x 5") for easy storage.  When unfolded, two threaded knobs are installed, locking the sides together.  It is triangular to match the shape of the equatorial platform that I bought from Tom Osypowski; with the equatorial platform sitting on top of it, it is extremely rigid, and does not vibrate much at all.  Here are photos of the platform being unfolded.  Note the selective removal of wood with bracing remaining intact.

platform folded unfolding
Figures 18a, 18b: Support platform folded, and partially unfolded.

set up
Figure 18c: Support platform with locking knobs installed.

The space inside the platform is a very convenient place to place the 17 amp-hour battery (a "jump start" power pack) that I use to power my telescopes.  It sits completely out of the way, and no one can trip over the cords!

Pole Case

While I was in a building mood, I made a case for the four poles.  It was made of 1/4" and 1/2" Baltic Birch.  I made the box, and glued the sides to both the top and bottom.  Then I cut the lid off with my tablesaw, ensuring a perfect fit of the lid.  Here's a photo of the tubes in the box, complete with some strips of weatherstripping to keep the poles from rattling.  (The end of the poles that goes inside the mirror box is painted flat black, the rest is Rust-Oleum Royal Blue.)

pole case
Figure 19: Pole storage box, poles inside.

Ventilation

The full thickness primary mirror needs cooling.  There is no alternative.  So, I added six 60mm, double-ball-bearing 12V fans to the mirror box.  Four push air up across the primary from the bottom, and out through doors in the top of the mirror box.  Two fans move air across the back of the mirror from the top corners of the mirror box.  Here's a photo of the fans blowing across the primary:

bottom fans
Figure 20: Ventilation fans for front of primary mirror.

Here's a photo of the back of the mirror box, showing the tertiary support framework (T-shaped square tubing), tertiary adjustments near the center, and the two fans in the upper corners:

back view
Figure 21: Back of mirror box.  Ventilation fans at upper right and
left.  Tertiary adjustments at center.  T-shaped structure is the
tertiary support, made of steel tubing.  Blue knobs are for primary
collimation.  Back of box is open for better ventilation.

The six fans result in much faster equilibration, and improved images during cooling.  No vibration is noted from the fans at high power.

I have experimented with a fan blowing on the secondary from below, but I haven't concluded whether or not it is beneficial.

Collimation

After adding flocking paper to the focuser (which resulted in the single largest increase in contrast of any step that I took) prevented me from inserting my 2" laser collimator all the way into the focuser.  So, I now use a 1.25" collimator.  First, the tertiary is aligned to place the laser in the center of the secondary (this usually does not require adjustment).  Next, the secondary mirror is adjusted so that the reflection of the laser returns to exactly where it originated.  Finally, the primary mirror is adjusted while viewing a star in the eyepiece, to eliminate coma.  The two blue knobs in the picture above are the primary collimation adjustments.

That's all there is to it.  I can have the scope collimated in a few minutes, provided I can find a star to align the primary with.  (A barlowed laser would probably allow daylight collimation of all the components, but some tweaking of the primary might be needed once it got dark.  I hope to try this soon)

Performance

The performance of this telescope has exceeded my expectations.  With a 32% obstruction ratio, I expected planetary detail to be washed out, and reduced from what I was used to seeing in my Newtonians.  I was wrong, especially after a little collimation practice!

Planetary detail was excellent.  Saturn was sharp at 450x when the seeing permitted.  Jupiter showed numerous bands and subtle detail.  I could discern color in one of Jupiter's moons as it transited, and could follow the disk across.  The central star in M57, the Ring Nebula, flickered in and out with the seeing conditions.  The double-double (Epsilon Lyrae) was easily split, and diffraction rings were visible when the seeing permitted.  Even a wheelchair-bound fellow club member was able to catch views of some objects (a shorter platform would have allowed him to see nearly all of them), and my friends enjoyed sitting comfortably and observing.  For public events, the intimidation factor is simply not there, compared to large Newts with ladders.  One interesting phenomena is watching people approach the scope; they walk over, staring at the secondary, apparently suspended in midair!  Once they get close enough they can see the wires, and then marvel at how thin they are (0.015").

Here's a photo of the scope set up for observing in the flatlands of Illinois, waiting for Jupiter to emerge in the twilight.  Not shown in this photo are the trap doors that allow air to escape after being blown across the primary by the fans.  Note the Rigel Quickfinder on the mirror box - this is all I need to find objects with a low-power eyepiece.

Set up
Ready for observing.  Wish those clouds would dissolve....

Setup and teardown require 15 and 10 minutes, respectively.  All components of the telescope (including the equatorial platform) will fit behind the back seat of my Subaru Outback, along with eyepiece cases and a tool box.  Because of this, it gets used a lot.

My conclusion is that this instrument is a resounding success.  Of course, my mind has already wandered to future scopes!  An 18" or 20" Cassegrain is within reach, but it will require a faster primary (F/3.5 or less).  Such a scope can still be extremely portable, and will have the same viewing position - feet planted on the ground.  A thinner primary is a necessity for cooling and weight reasons.

In closing, I'd say that I am very happy with what I have learned in the construction of this telescope and  its performance.  I hope that in the future it inspires others to try the design, and shows that Cassegrains don't have to be inferior in terms of optical quality.  If the optics are top notch, then a scope with a well designed mechanical structure will yield excellent images.

final view

final view 2

I'd like to thank all my friends for their support and encouragement, my girlfriend for her understanding, John Pratte for his help welding the tertiary support frame, Dick Wessling for his advise regarding the optics and design, and Ed Jones for double checking my calculations of the optical system.

Thanks for reading.  Here's one last photo of the telescope and me as daylight fades....

sunset
Photo by John Stone

Build a Cassegrain!  It's fun!

        Mike Lockwood


Back to the other 12.5" Cassegrain Page
Back to telescopes