Stud Wall Pre-Assembly

Back to work! After a couple of weeks doing other work around the place, I decided to buy a decent table saw to cut and shape the stud wall end studs... I wish I had bought this thing earlier, as it has made life soooo much easier in general!

Anyway, I reduced my small mountain of raw lumber into lots of smaller pieces and set about cutting out the rebates in the lintle and foot-plates of the stud walls. The end studs on each of the eight walls must be cut to a trapezoidal shape to make everything fit snugly:


So, after a bit of fun with the glue pot and a large hammer, I knocked out 8 stud wall sections, which I was delighted to find all fit together perfectly. With each pair of walls aligned and G-clamped together, I drilled through the studs in three places and bolted them securely together. And here's the result!


It really is quite roomy inside - and I haven't even put on the roof yet!!

Next job is to drown every squate inch in builder's grade Protim to preserve and protect the wood - a smelly, messy job!

Dome & Wall Ring Sub-assembly

Having cut out all 32 x 45degree sectors for the dome ring and wall ring out of 8 x 3/4" sheets of hardwood marine ply, the sectors were glued and screwed together in groups of three, offset to lend maximum strength and support to each other, with each sub-assembly representing 90degrees of the full ring.


After gluing and screwing the three sectors together into a double thickness of 3/4" ply, a small forest of G-clamps was used to squeeze out the excess glue and assist in a uniform laminate.

Laying this lot out on the ground, in a full circle, gave me a real sense of just how large a 16' dome was going to be! Man, this thing is big!! It was gratifying to see that my trigonometry skills haven't completely deserted me - all the sectors were exactly the same size - to the millimetre!

Cutting the Dome Rings

To support the lower rim of the dome, lend it more rigidity, and provide a hard flat surface for the dome transport wheels, I cut out out 16 x 45 degree annular sectors of 3/4" marine ply that will be glued and screwed together to form a double thickness (1.5") ring. This Dome Ring has an outside diameter of 2500mm, exactly matching the outside radius of the dome itself, and is 150mm wide.

Another, double thickness ring was similarly cut and will be attached to the top edge of the octagonal observatory walls. This Wall Ring is 50mm smaller in diameter than the upper dome ring, allowing the dome ring to overhang the wall ring by one inch all round. The wall ring is also wider than the dome ring, at 330mm, so that the eight outside corners of the walls are aligned with the outer circumference of the ring and the inner corners with the inner circumference.

At a later stage a thin, 6" wide metal strip will be attached to the outer edge of the dome ring. This strip will hang down below the wall ring and prevent rain from being blown into the gap between the rings. A draft-excluder brush strip, as used on the bottom of doors, will also be attached to the outer edge of the wall ring with the brush pointing up and gently rubbing against the dome ring to keep out insects and dust.

Using graph paper and compass I experimented with how many sectors I could cut out of each sheet of ply. The best fit was achieved with two dome ring sectors and two wall ring sectors nested tightly together. My 10:1 scale drawing suggested that I should have 30mm to spare... however, my pencil lines and slightly springy compass meant that my error factor was a little over 30mm! Cutting the first sheet was going to be an act of faith!

10:1 scale drawing of how two dome rings and two wall rings

fit onto a single sheet of 4' x 8'

Eight sheets of 3/4" x 4' x 8' Hardwood Marine Ply are required to cut out the 32 sectors required for the two rings. For those who haven't had the joy of trying to handle this stuff, a single sheet is humongously heavy even for two men. Dragging this stuff around my yard and into the garage single-handedly for cutting nearly cost me my life on numerous occasions! My arms hurt. My back hurts. Everything hurts!!

Before cutting each sector, the two ends of the sectors were carefully defined so that they would be accurate 45 degree sectors. This was done using simple geometry... on a large scale!

I knocked up a long radius arm for the router out of a length of skirting board, using a screw as the pivot point. The plywood sheets are raised off the ground by some lengths of scrap chipboard to prevent damaging the cutting bit. Three passes with the router were necessary to cut all the way through the very hard 3/4" ply.

Slow progress... as you can see, there is precious little space between the cuts

every millimetre is precious!

And the proof of the pudding is in the cutting...

Full house - they all fit!!!

Dome Control Hardware

There are a number of commercial systems on the market to control the dome (opening and closing the two-door shutters, and slaving the dome to the telescope so that they are both pointing in the same direction.) However they are rather expensive for the piddling amount of electronics and driver software involved. I was just about to reach out and 'touch' a friend to see if he could weave his magic (and soldering iron) to come up with his own control system, when I stumbled across the LesveDome group, run by Pierre de Ponthiere, who had developed dome control hardware based on a generic off-the-shelf USB controller board. The USB board is used to control a number of power relays for activating the dome shutter and azimuth motors, and to take inputs from a number of sensors and limit-switches. The ASCOM compliant driver software is still under active development by Pierre but is fully stable in its present incarnation. It's also dirt cheap!

There is a registration fee of €30 for the driver software (after 60 days free trial) and the USB controller board can be picked up either ready assembled or in kit form for around €56 including shipping. Being a consummate fiddler, I plumbed for the kit, which looks like this:


The kit is a doddle to assemble for anyone who can wield a soldering iron safely without putting out one of their own eyes or accidently setting fire to the cat, and requires no knowledge of electronics. After sorting out all the components and checking that nothing was missing, thirty minutes saw the unpopulated circuit board transformed from this:

...into this:

The kit comes with a CD of interface software for general use and for testing the completed board. I bought the USB Interface Board Kit from Ramsey Electronics in the USA using PayPal and had it in my hands less than a week later. Excellent service!

The next step is to knock up a small circuit board with the power relays and other ancillary gubbins - more on that later.

Full Size FSP Layout

To double check the final dimensions of the Folded Stevick-Paul optical tube assembly I laid out three 4'x8' sheets of plywood on the ground and drew out the optical path at full scale. The white rectangles in the photos are full scale drawings of the mirror cells that will be bolted onto the tube assembly. Doing this allowed me to double check all the dimensions and angles that I would need for cutting up the 1" box-section steel for the OTA and fine tune the whole support structure.


The photo, below, shows the light path of the convirging cone of light to its focus point, and how it passes close to the tertiary 10" flat folding mirror (in fact, through it's cell!) This mirror cell will have to be built in such a way so as not to impede the light path. Cooling fans on the tertiary mirror will heve to be ducted away from the light path to avoid turbulence.

It is interesting to note that in faster (f/12 and faster) versions of the Folded Stevick-Paul design the light cone actually grazes the edge of the tertiary mirror, which needs to be very carefully bevelled to minimise the obstruction. This is another reason I decided to design my FSP at f/15. I had hoped to make an f/20 system, but that would have meant building an even bigger observatory! Also, I can *just* test a 14" f/15 mirror inside my house. The mirror must be tested at its radius of curvature (twice its focal length) which is 10.5m!!!

Sub-Assembly

1st May 2008: Wonders will never cease! After something of a hiatus, I finally managed to pull the finger out and start work on this project again!!! I finished off nailing and gluing the 2" edge support ribs to all of the hex-pents and drilled them to accept two M6 gutter bolts to assist in joining the panels together for final assembly. Stone me - it works too!! Using bolts to hold the panels together while the glue sets removes the need for lots and lots of G-clamps (been there... and gave up!)

To make final assembly of the complete dome a little easier, I have glued and bolted together five sub-assemblies consisting of a central pentagon surrounded by five hexagons. These sub-assemblies are about as big as can be easily lifted and carried by a single person (i.e. me!) They will eventually be joined together with five sets of intermediary sub-assemblies (middle picture, below.)




My next move was to cut out the 4mm plywood for the inner skin of the dome, to hide all those nasty looking support ribs and provide a smooth inner surface to the dome to facilitate heat to escape through the open shutters. Without a smooth inner surface, the ribs would effectively provide lots of pockets for warm rising air to get trapped in... which would result in a slower cool-down of the observatory and ultimately, crappy seeing.

Internal Support Ribs

Although geodesic structures are inherently strong I wanted to strengthen the Hex/Pent panels around their periphery with a 2" wide strip of 1/2" ply. This would also make final assembly much easier (i.e. possible!) when it came to joining the panels together. The ribs also allow an inner skin to be fixed to the inside of the dome giving a nice contrast in geometry - outside you have 360 triangles - inside you have 55-odd hexagons and pentagons! More importantly the ribs offer something for the shutter-opening brackets/hinges/widgetry and motors to be bolted to, without going through the outer skin.

Each rib consists of a 2" wide strip of 1/2" ply. Both long edges must be cut at an angle unique to the triangle edge that it is supporting (I cheated here and fudged the angle at 4 degrees as the actual difference in this angle for all the different triangle edges only varies by about 0.6 of a degree. Life's waaaaay too short, so sod it!) I used a borrowed table saw for this job, cutting up around four sheets of 4'x8'x1/2" ply into 8' strips. The saw blade was set to 4 degrees off verticle.

The ends of each rib must be cut at TWO specific angles simultaneously if they are to fit together accurately and offer maximum support to the panel as a whole. A mitre saw, a drill-bit and a little trigenometry can overcome this minor headache reasonably well. Doing this by hand, however, was a major pain in the bum, so after struggling through about half of the ribs with the hand mitre saw I invested in a power mitre saw on which I could set the angle and tilt of the blade. The blade on this saw did not cut as neatly as the hand saw, but it made short work of the remaining ribs.

The ribs were then glued and nailed to the inside edges of the HexPent panels and clamped tight around their periphery with a ratchet strap while the glue cured. The ribs remove a lot of the flexability in the panels and add a lot of strength to each panel and the whole dome assembly.