Making a 32" Dobson
my very first telescope to my current one the progression has been an
increase in aperture. First a 2-inch refractor, then reflectors of various
sizes ranging from a 4", 8", 16", 22" and now
32". For people who have never caught the "aperture fever",
this would make no sense. Why else would anyone sell an instrument which
is fully functional only to make a new one that then will be harder to
make, move and to use.
The answer, of course is that the bigger the scope, the more spectacular the images are. To observe for the first time in a large instrument is a true shock and the only way of renewing this experience is to increase the aperture. This is the cure to the disease know as aperture fever. If it is possible to make a telescope, which will provide better images, then it's hard to resist. If drawing and making telescopes is a pleasure, it's now impossible to resist the temptation. Time spend designing the instrument is the most important part. Smaller 'scopes are less prone to error but when a mirror and cell combined goes beyond 140 to 160 lbs. problems can arise. This project has been great experience.
this project, the choice of the mirror diameter is a simple one; within
limiting factors it should be the largest manageable optic feasible. It's
useless to start manufacturing a monster whose use would be impossible to
realize. What guides any choice is generally the weight of the mirror.
There are 36" and even 40" mirrors available but there isn't any
way possible I can make a scope around that big of a mirror and be able to
move it. A 2" thick 40" diameter mirror alone weighs around 180
lbs. The 22" visible on this site enable me to test some ideas
borrowed from other amateurs. The structure first of all, is a plywood +
foam composite. Steven Overholt largely used this technique for his
30" that he made few years ago. He was really years ahead using this
method. The astatic cell, already used with success on the 22" is now
replaced with a new system, chrysocal blade. The purpose of the astatic
chrysocal mirror cell is to reduce flexure and give a reliable support for
this large thin mirror. Another change in design compared to the 22"
is the use of carbon fiber truss tubes. The carbon fiber is used in order
to lower as much as possible the center of gravity. Other change, the
diagonal cage is a simple single ring instead of a complete classical cage
in order to also have a low wind profile. Transport is also different
because even if the instrument can be transported in a medium size car, a
trailer offers more flexibility such as for example, a temporary storage
place. The plywood used in construction is Baltic Birch for most part,
except for the cage where ordinary very thin plywood is used. This Baltic
Birch plywood is really beautiful, strong and easy to work with|
The mirror cell
The mirror cell is the first part I made. An astatic mirror cell seems to be complex but I've found that not to be true. The 22" cell was a good test to be more familiarized with this system and while making it, I noticed that the tolerances of manufacture were well within the range of an ATM. This time I choose an astatic cell made with chrysocal blades. It's a little different system from the classical astatic mirror cell
It's very powerful and minimizes the problem of mirror flexure. However, pushed by the desire of experimentation, I made a mistake: the levers were too long. I have made new and stronger levers and now it's much better. These new levers allow a more precise adjustment. First I indeed to choose a ratio close to the 1:20 that theory says will function but this gave many problems. This failed in practice because the influence of a very light counterweight (120 gr.) was weak compared to the mass of the axel that supports it. The new levers have a width of 4.7" compared to the former 2.4" in the first version. The new levers resistance to possible deformations is much better. The supports are composed of 9 " T " shapes. These " T's " are made with welded pieces of steel. They replaced perfectly the traditional triangles while offering a resistance quite higher than typical aluminum triangles. Combining an iron "L" shape and "U" shape makes them possible. Their rigidity is excellent. Six of them have an astatic lever and the other 3 are on a fixed point (floating of course). The principal piece of the cell is made using two thicknesses of 3/8" Baltic Birch. The joining was somewhat delicate because of an insufficient number of clamps and no screws to join both plates. This caused one sheet to slip by about 1/16". This was corrected by careful sanding. Six reinforcements were made using 3/4" plywood that were glued together and then screwed. This reinforcement is joined also by two aluminum triangles that connect the reinforcements of the lower part of the cell. In order to further increase rigidity, all around the edge of the cell, on both sides are glued and screwed an "L" shaped aluminum angle. The mirror cell is "connected" to the mirror box thanks to 3 square aluminum shaped tubes of generous section, about 2.5". These sections will receive a stainless reinforcement, 3/16" thick. They are glued and screwed under the cell, at 120° and take part in its rigidity. A structure welded out of steel can offer the same performance while being easier to realize but as I like working with wood, I wanted to try to use it as much as possible. The weights of the cell components are:
Triangles : 10 lbs
Structure : 24 lbs
Levers : 2 lbs
Counterweights : 11lbs
The new levers have larger counterweight but that didn't give any problems because the mirror box was lighter than expected.
Lateral support are piano wire with a diameter of 2 mm, they are placed at 120° . A small pieces parts have been specially made for that purpose. I must keep testing but currently it seems to be good. Those supports are on mirror box in order to minimise flexion. At that place wall thickness is about huge because collimation threaded rods go throw them.
During transport, the telescope is
tighten on the trailer. So, in case of a crash, the base/fork/mirror box
will not move a lot. The mirror, free of movement might go up quite
violently. In order to "limit" as much as possible the possible
damage, I try to avoid as much as possible a instant stop when the mirror
hit the retention clips (5 mm thick stainless steel with a thick foam
under them). I’m sure a strong crash will destroy everything but in case
of light to average crash, it may be safer for the glass. In a classical
system, the only choc absorber is the thin layer of foam under the clip. For this cell I have made a second
protection. In case of a strong crash, the retention clip rotate slightly
along a axe a long rod go to crash in a second layer of thick foam.
Pictures are better to explain the system. The rotation axe is tighten but
there is only one screw, it can rotate. I hope this will avoid some
possible damage. Also, in order to try to limit damage, the piano wire can
be remove and replace by thick foam. It’s safer to avoid as much as
possible a contact between metal part and the glass. The collimation is achieve with 3
threaded rods of ¾ inc. The 3 screw have a small motor that is connected
to a hand pad, so it’s possible to tune the collimation at the eyepiece
anytime you want. Special thanks to Francis Tisserant for is help ! The
box he made for me can also "drive" 3 fans with variable speed (two
small fans with a variable speed adjustment and a big fan with a separate
adjustment). I haven’t test them yet.
During transport, the telescope is tighten on the trailer. So, in case of a crash, the base/fork/mirror box will not move a lot. The mirror, free of movement might go up quite violently. In order to "limit" as much as possible the possible damage, I try to avoid as much as possible a instant stop when the mirror hit the retention clips (5 mm thick stainless steel with a thick foam under them). I’m sure a strong crash will destroy everything but in case of light to average crash, it may be safer for the glass. In a classical system, the only choc absorber is the thin layer of foam under the clip.
For this cell I have made a second protection. In case of a strong crash, the retention clip rotate slightly along a axe a long rod go to crash in a second layer of thick foam. Pictures are better to explain the system. The rotation axe is tighten but there is only one screw, it can rotate. I hope this will avoid some possible damage. Also, in order to try to limit damage, the piano wire can be remove and replace by thick foam. It’s safer to avoid as much as possible a contact between metal part and the glass.
The collimation is achieve with 3 threaded rods of ¾ inc. The 3 screw have a small motor that is connected to a hand pad, so it’s possible to tune the collimation at the eyepiece anytime you want. Special thanks to Francis Tisserant for is help ! The box he made for me can also "drive" 3 fans with variable speed (two small fans with a variable speed adjustment and a big fan with a separate adjustment). I haven’t test them yet.
The mirror box
It is of octagonal shape for reasons of size, rigidity and personal taste. It is a laminated composite of Birch and extruded polystyrene foam. The plywood thickness is 3/8 " and the foam is 2" thick for the two big lateral panels of the mirror box. Some of the panels are reinforced with 1/2" ordinary plywood. The connections between these plates are made with a composite of 3/4" wood with foam. Their width is of 3" thick. These pieces are very thick because they must support the collimation screws, which are 3/4 " thick. This octagonal shape was no easy thing to make to make because of all 45-degree angle cuts. Compared to the 22", this meant much more work. Because of its shape using the 45° cuts these joints were also more delicate because to have a good pressure to clamp all in place it was necessary to have parallel guides to secure everything while the glue dried.
The total mass of the mirror box (only parts out of wood) is 35 lbs. Rigidity is excellent In order to minimize its height during storage and transport, I have thought make the bearing split into two parts. It's possible to reassemble parts to observe but the problem is a possible lack of rigidity. I preferred to join all parts. In order to reinforce the assembly, some small pieces of wood were added on the two faces of the "bearing" and act as straps. During the assembly, it appears that the "bearing" had a different thickness than the lateral panels of the scope. This was because the foam didn't always have the same thickness. A sanding machine is then an invaluable tool. The assembly of the panels was facilitated by the use of screws. The clamps cannot always do everything. The precision of the unit is, I hope, correct because as it turned out, it is approximately 1/32" to 3/32 " off compared to the ideal dimensions. For 1000 mm, that gives 1003 mm length and on the others faces of 999mm a 1001 mm. That should not be too much of a problem. In order to smooth the bearing, I used the same method that one use to make equatorial forms, rotation of the parts on a grinder, which is fixed. With a design like that of the 22" or this 32", the problem it is that there is no actual of the center of the bearing. Thus I had to make an assembly in order to grind those bearings. The axis used was an aluminum bar 1" in diameter. It will latter be used for the axel of the telescope wheels. The operation is a little delicate and it is necessary to take care not to remove too much all at once. The grinder is placed on an increasingly thicker plywood plate in order to remove the material gradually. It is a little delicate and should be taken slowly. If the grinder hangs too much, it digs in a begins to turn the mirror box. The result of this method is astonishing. The surfaces that were rather irregular are now perfect. It should be hoped that the axis did not move (but it does not seem to) during all the operation but if not, the form obtained is still a beautiful disk. Having a smooth regular surface, a 1 mm thin stainless section, has been glued in order to reinforce this surface and give a good surface for ball bearing. Four layer of polyurethane coating have been apply on the wood to protect it.
La batterie de serre joints utilisés pour le collage.
The truss tubes
I have tried to maximize the base of the triangles formed by the truss tubes. As the shape of the box with mirror is octagonal, the serrurier truss is inverted with regard to the classic model, which implies a cage with a square shape. It is certainly possible to realize a classic shape, even to use only 6 tubes instead of 8, but if we wish to maximize the strength of the structure, it is necessary to maximize the width of the "base" of the triangle. The square cage allows me to obtain this most important rule.
The choice of truss tubes was an especially tough one. Using a spreadsheet article that appeared in an issue of Sky and Telescope by Roy Diffrient, it was easier to test various combinations. The aluminum tubes are usually used. Considering the size of this scope, they are unfortunately too heavy to use, difficult to find, and expensive. Due to the short focal length of the mirror, F/3.75, it was necessary to look for something stiffer and of a lighter material. There is not a great deal of material available that can answer these two criteria. Using carbon fiber tubes seems to be the ideal combination of weight and rigidity. The price is certainly higher compared to using aluminum tubes. Their "exotic" size that is offered is still quite close to standard aluminum tubes. There are various types carbon fibers. Not being expert on the subject, I simply noticed that the carbon fiber named "pultrusion" was the best adapted to a serrurier truss by being almost twice as stiff as the classical fiber. The classical carbon fiber certainly is interesting, but its strength is much lower.
This difference between both types of carbon is very important because the module of Young, an indicator of rigidity is almost twice superior. ...Unfortunately, there are only a few companies that make such a product. The Structil Company, based in the South of Paris, makes a good product. I bought 8 ft lengths in a diameter of 1.2-inches and .078" wall thickness. The weight of the 8 tubes is about 12.6 lbs. The cutting of the tubes has been done with a disk. but it seems that it is necessary to take a few precautions: Wear gloves, a dust mask, and glasses and clean up as much as possible the very irritating dust. I will try to cut that tube using a Dremel system
The cage is square. The foam is 2.4 inch thick and the plywood is 1/8 inch by 3.1 inch wide. The foam that is not covered with plywood is protected with a wooden veneer 1/32" thick. A coat of polyurethane varnish over a layer of epoxy help to protect it from moisture. As mentioned previously, the square cage allows, in this design, to obtain the maximum space at the base of the truss tubes, about 40-inches and thus will obtain a maximum strength. The weight of the cage is around 6 lbs. The spider is finished. The aluminum blades are 2 mm in thickness. The central body raises a problem because it is also made with aluminum and its weight was high. This issue was corrected following a lot of machine work to lighten it by milling away a lot of its mass. The secondary holder is a special one, the mirror can be extracted without any tools or loss of collimation. It’s safer to remover it during transport or set-up. The focuser is attached on a separate panel with a foam "arch" cover by carbon fabric and epoxy.
The base and the rocker
Its structure depends mainly on the choice of the motorization and the form with which I will move the instrument when is on the ground. The motorization is a delicate element. There are several possibilities: Use a DobDriver II, Mel Bartel's system, a ServoCat system from www.Rxdesignonline.com or an equatorial table, or. This last solution is very pleasant to use but for an instrument of this size, it starts to become a very big platform. The eyepiece is already at about 9 feet at the zenith. Gary Myers’s solution (ServoCat) has been choosen, they will be a page devoted to that subject soon.
The goal is to integrate as much as possible the motorization so that it is the least cumbersome possible. The base must have an accurate shape in order to accommodate the azimuth motor, like Charles Stark did to install his DobDriver on his obsession. The base's edge is covered with a polyamide tape in order to offer a good "grip " to the toothed wheel of the driving pinion. As Gary’s solution requires a low friction movement, good quality ball bearing, like those used for roller blades, has been chosen instead of the traditional pads of Teflon on FRP or Formica. For a telescope used on an equatorial platform, a DobDriver II without roboscope or a tradition ground board, the Teflon/FRP combination is preferable because the bearings are sometimes more delicate to tune and can require the use of brakes in order to control the quality of the movements. I also to use bearings for the pivot but I do not know yet how. In order to minimize the height of the instrument, the base is bored in its center, leaving the free passage in front of the mirror when the 'scope is aimed at 45°.
The structure of the base is also in composite of 3/8" plywood (visible panel) and 2.4-inch foam and 1/2 inch plywood for the low part of the base. Four inflatable wheels of 10-inch diameter with roller bearings are placed on a 1-inch aluminum bar. They are removable by the use of pins. The axels of these wheels pass through the sides of the base. At the place where these axels pass in the fork, there is a plywood reinforcement, 3/4-inch thick. The edge of this is 3/8" thick. Removable bars, intended to exert the push during transportation of the instrument, will be fixed on the side of the telescope. That also has the advantage of avoiding a possible swing of the mirror box out of the pivots during Transport
Since I have seen the beautiful trailer used by Patric Lequevre to transport his 18", I want to adopt this solution. The instrument being of rather large size, approximately 40" x 40" x 120", it is difficult to transport in one piece, so it will be dismantled for transport. Dismantling also allows helps to reduce the size of the trailer
A 40" mirror about 2" thick weighs around 180 lbs. This Russian 32" blank I'm using is a thin meniscus disk, 1.6" thick. The meniscus shape allows compensating partially for the weakness in a 1.6" thick blank of this size. The weight saved by this shape is interesting because this thin primary weighs only around 100 lbs. The problem of it's "convex" back is that the contact points of the cell are more delicate to adjust. It taked some time to tune it. The secondary mirror is 6-inch minor axis. The FD is 3,75 and a Paracorr is used for low power view. The secondary is a 6" model, so obstruction is around 18,7 %.
It’s a one person setup scope. First the truss is install and a thin alu blade with Velcro glue on it keep the space between truss element at the good distance. Then, a piece of wood keep the scope close to an horizontal position, the cage is installed and the blockage system is removed.
Some tuning soon
The mirror is not centered on the bearing line (85 mm back) so the center of gravity is not at the good place. A Tom Krakjii virtual counterweight will be added soon to avoid this problem. Also, baffling of the secondary will be added in order to have the best possible contrast. I don’t want use a full baffle system to avoid wind sensitivity as much as possible. Different solution will be tested till I find a good one. Collimation motor currently stall but not because of the load, it’s because of too much friction on treated rods, those friction have been removed a other test will say if the motor are now strong enough.
First light has been tough because of bad weather…Murphy’s law again…First, the mirror has been put in the mirror box. It’s a 100 lbs piece of glass but thanks to the very low profile of the box, it’s really easier to do with two people than to put the 20" in a classical dob. After some tune of the astatic cell, the scope has been aimed at M13, like for the 22". …Amazing !!! pin point star cover the field of view, and even more impressive, the mag 12 galaxy just next to M13 is now a beautiful galaxy, no longer a faint smudge of light. M57 is also great. A thin line of different "color" is now visible all around the classical annulus. This one is visible on some pics as a thin red line. The central star is visible 75 % of the time. Needless to say I’m happy !!!