Monday, March 20, 2006: The secondary cage forms the top end of large Dobsonian telescopes. It is an evolution of earlier solid tube designs, shaped by the requirement that large portable scopes break down into smaller, easily transportable pieces. One design goal for the secondary cage is that it be as light as possible to ensure the scope can be easily balanced without adding lots of weight at the bottom end. Since the scope is essentially a long tangent arm, a pound of weight added at the top requires approximately five pounds at the bottom. At least, it does in the Kriege design. This formula is subject to variation depending on altitude bearing diameter, but more on that later.
I decided to generally follow the Kriege/Berry design, which calls for the secondary cage to be constructed from two rings made from plywood and four struts to separate them. Initially, I decided to size my cage rings following the Kriege formula, which suggests an inside diameter of about 22.8 inches for a 22-inch f4 mirror. I rounded the inside diameter up to 23 inches. After cutting out the two ring disks, however, conversations with ATMers on CloudyNights as well as with Steve Kennedy and a few others, caused me to conclude the Kriege secondary cage ID is too narrow. As one person put it, Obsession scopes are “notorious” for having undersized secondary cages.
Two problems result from under sizing. The lesser problem is that the smaller inside diameter contributes to light path vignetting. The secondary cage itself actually blocks a percentage of light the primary mirror can direct to the eyepiece, decreasing the level of off-axis illumination in the field of view. The most important factor on this issue is the secondary mirror itself. The secondary mirror size represents a compromise for various reasons — make it too large and the secondary obstruction reduces contrast in the field of view (and increases cost!) unnecessarily, make it too small and the off-axis illumination is lower than it should be. The secondary cage ID plays a role in determining off-axis illumination, but it is much less important in this regard than secondary mirror size.
The other problem caused by an undersized secondary cage is that it forces rising warm air into the light path, decreasing resolution. This is a non-issue for people who do not use a light shroud, but every large scope owner I know does use a shroud. Here in Missouri, dew is a significant problem. That reason alone is enough to convince me that a light shroud is essential. However, an unfortunate (not to say highly improbable) incident last fall involving my 31mm Nagler Type 5 eyepiece (the pride of my eyepiece collection) and a high flying night bird, convinced me I would never expose a large mirror to such stealthy aggressors.
So how big to make the cage rings? I called Randy and Alan at AstroSystems — an invaluable resource for telescope makers, not to mention excellent large dob manufacturers for those less-inclined to spend months in the workshop building their own. Having purchased my secondary mirror, spider, and holder from AstroSystems, I felt justified soliciting advice (of which Randy and Alan have a plentiful supply they generously share with customers). Randy said they size their rings with a 24-inch ID for 22-inch mirrors. Some ATMers resort to arcane mathematical formulas, spreadsheet calculations, and scope modeling software to determine the optimal secondary cage ID. I decided to go with Randy’s recommendation, set the original disks aside, and cut two new disks with an outside diameter of 28-inches. (I decided to make the rings two inches wide).
I decided to use Baltic birch plywood for my project. Apple-ply, featured prominently and strongly recommended in the Kriege/Berry Dobsonian telescope, is not commonly stocked by lumber sources in our area (or in nearly any one else’s from what I have heard). Baltic birch is available from many commercial lumber yards, though not from HomeDepot or similar outlets. It features an all hardwood core (birch) with many layers, making it much stronger than plywoods stocked at home centers. These have softwood cores with fewer layers, often with voids in the center which have a tendency to show up in the worst places for your project. Baltic birch is sold in 60×60 inch sheets — it is sized in metric dimensions. The 3/4-inch width is really 17mm, a little under 3/4 inches.
To make the 28-inch rings, I cut two 30-inch squares on my table saw and then screwed each square (in turn) onto sacrificial spacers placed on my work table. The screws are long enough to pass through the spacers into the work table, holding the stock rigidly in place for cutting. An MDF table table top is cheap and easy to replace, a definite advantage for this approach! The spacers prevent the router from cutting grooves in the table. I put several screws in the center area and several at the corners, being careful not to place any screws where they would be in the ring cutout. The router I am shown using is a Porter Cable 2Hp plunge router, one of the top-rated commercial routers. I use a 3/8-inch down cut solid carbide router (with a half-inch shank). The disks are cut out in multiple shallow passes. This is a messy business, requiring frequent cleanup, as well as an effective sawdust mask and ear protection. As you can see in the above picture, flying sawdust even found its way to my camera lens.
I am shown here unscrewing the ring disk and cutouts from the supports.
Here are the work supports after the first ring disk was removed. The supports are just scrap pieces of particle board — nothing fancy.
I have clamped one of the ring disks onto the corner of my work table and am using a smaller fixed-base router and a 1/4-inch round over bit (1/2-inch shank) to finish one edge of the disk. This has to be done on both sides. A plunge router could be used, but unlike the initial cutting operation (where a plunge router is a necessity), the fixed base router is easier to handle for this operation. This is a 3/4 HP Porter Cable router. While I usually work in a shop apron, wearing an apron while routing is a good way to get apron pockets loaded with sawdust.
After rounding over with a router, the edges are smoothed with a palm sander. I prefer Norton stereated paper. Stereate is a soap compound. It keeps the paper from jamming with sawdust. The paper costs more, but lasts longer and sands better than normal garnet type papers.
After sanding, the disks are reattached to the work table for the second router cut. The second cut completes the ring. It is important to keep the same side which was up when the first ring was cut facing up for the second cut. Otherwise, the slightest misalignment in the pivot guide hole results in an eccentricity between the inner and outer cuts — that is, the ring would not be perfectly symmetrical. Since the guide hole has to be made with a hand-drill, it will not be exactly perpendicular with the face of the disk. Instructions that came with the circle cutting jig (from WoodCraft) stated that a pilot hole is not necessary. A quick experiment convinced me that it is. The pilot hole makes it easier to ensure the pivot is dead on the center mark and also makes it easier to insert and remove the circle cutting jig.
The second cut, which forms the ring, is made in a series of shallow passes. Take your time and do not hurry here. Trying to cut too much material in a single pass will make the inside surface of the ring (where the light baffle will be attached) unnecessarily rough, and could result in splintering of the surface plies or even an inaccurate cut. It is also a good idea to make sure the router is set for the correct inside dimension. You will have to reduce the circle by twice the width of the ring, AND make sure the outside of the router cut defines the inside of the ring — the reverse of the first cut where the inside edge of the bit defined the outside of the ring. To ensure the correct setup, I made a set of reference marks on the edge of my work table and set the circle cutting guide from that. I did NOT use one of the disks to set up my router. Doing so is a great way to have the opportunity to cut a third disk…
Here is a finished ring. Only the center waste area is screwed to the table. The ring itself is not attached screwed down — you don’t want unnecessary holes through your cage rings. An advantage of the work supports is they keep the ring from dropping when the cut is finished. Holding it in position reduces the chance of nicking the ring with the router when the cut is finished.
Here are the finished rings. Cutting out, rounding over, and sanding the rings took three hours.
Not including clean up time. (Hint: Do not try this in your living room…)
