All the coach screws were removed and the three wall frames taken down. I lifted the floor and re-cut it so that it would fit within the inner boundaries if the wall frames rather than being flush with their outer edges. Lifting the floor was more difficult than anticipated as the heads of some of the screws had been damaged when the floor was initially laid and some proved to be particularly stubborn in removal. The floor was eventually coaxed from its moorings and reshaped.

The three wall frames were then screwed directly onto the horizontal fence posts. At this stage, two 3"x3" fence posts were screwed into the two corners formed by the three wall frames. This time 100mm coach screws were employed. The posts were mounted flush with the tops of the wall frames but extended about 9" into the previously dug corner holes where they would later be cemented into position. The addition of the two fence posts anchored floor and walls together into a solid structure.

The next addition was the fourth wall. This was to go at the front of the observatory. Since the door was to go on this side as well, the wall width was shortened with this in mind. A third 3"x3" fence post was screwed into position at the new corner, solidifying the structure even further.

The Observatory Roof Design

Being a roll-off roof design, I had looked around for materials that would facilitate some kind of wheels on/in track arrangement. The roof also had to roll
off somewhere, so the idea was to have the tracks extending from the observatory walls for another 8 feet - the length again of the observatory. Needless
to say, the tracks would need to be supported at their 'free ends.

Unfortunately, I couldn't find the right kind of materials, especially ones that would stand up to repeated use when subjected to a heavy load. What I did find were some 2.5" neoprene castors (the fixed rather than rotating kind) which were mounted in sturdy aluminium frames. The frames contained four holes for bolting the wheels to furniture. The wheels are slightly over 1" wide. They're also quite tall given the additional height imposed by their housings.


Picture 4: The basic frame for the observatory including rails stretching out to the right to their supports and the door.

The Rails to Carry The Observatory Roof

Since I couldn't find any material that could be pressed into service as a running track, I ended up cutting a 1.25" wide (this width was found by trial and error using one of the castors with a spare block of wood), 0.5" deep rectangular groove in two 3-metre long pieces of 2"x3" using a router. Several passes were required for such a deep groove but an attachment lets the tool cut parallel with an edge, so getting a groove central along the full length of the beams was not a problem. What was a problem was that the 2"x3"s weren't long enough to cover the full run of the roof, so two beams had to be joined end-to-end on each side of the building. More work with the router.

With the rails cut to length, the short pieces were bolted (using 150mm carriage bolts) into position first. One went from the top of the leftmost corner fence post to about half way along the top of the rear wall frame. The other was bolted to the top of the nearside corner fence post (where the door would be placed), crossed the gap left for the door and the remainder was bolted to the nearside wall frame (see Picture 4).

With the two short rail lengths in position, the remaining 2"x3"s were clamped in position, parallel with the nearside and farside wall frames. Using a plumb, the positions immediately below the ends that extended into the garden, were marked. The beams were then taken down (to avoid injury).

The Rail Supports

Two 3" square, 24" long, metal fence post holders were hammered into the ground at the marked positions. These post holders can be a problem. Should they hit a large stone or rock on their way down, they have a tendency to rotate off square. One of the posts encountered such a problem as it was hammered in and it was only through brute force that it was deformed slightly and rotated back into position. Levers are wonderful things! In my case, the stone encountered was about 18" down in thick clay and it was impossible to actually lift the post holder out of the soil, so the leverage was brought to bear.

One other problem with such post holders is that they can veer off vertical as they're hammered in. Constant checking with a spirit level is essential. Having gone through these tribulations, I'd recommend that the posts be cemented in position instead!

With the post holders in place, the fence posts were inserted and individually measured and cut to length to take into account any minor differences in the depth of the post holders. The posts were then positioned vertically using a spirit level and the post holder locking screws tightened to fix them in position. A cross-beam was cut and fashioned to fit between the two vertical supports, both to keep them a fixed distance apart and to provide a more solid framework.

Mounting the Rails

With everything now in place, the two long rails were mounted in position butted up against the existing short rail sections, bolted parallel to the long wall frames and screwed onto the top of the two vertical supports. Two "fill-in" 2"x3" pieces were then bolted to the two shorter wall frames, between the two rails. This ensured a level surface around the top of the walls.

The Observatory Door

Since the aperture for the door was a non-standard size, I ended up building the door from scratch. The basic door was made out of one 3"x3" fence post (to house the lock) and three 2"x3"s nailed together into a rectangle. A small margin was left around all four sides to allow the door to open and close and allow for expansion in damp conditions. Two 2"x3" cross beams were added about 15" from the top and bottom of the frame to improve rigidity and provide anchor points for the hinges. I decided against the classic Z type support as I wanted to be able to be able to make fine adjustments to the frame (to make it slightly off square) should that become necessary (which it did).


Picture 5: The basic skeleton including the roof which was temporarily placed in position to evaluate how well the wheels traveled within the rails

Galvanised steel garden-shed type hinges were chosen for connecting the door to the nearside wall frame. These hinges have a standard rectangular end for fixing to a door frame; the other part of the hinge is long and triangular for fixing to the front side of a door. I routed recesses in the two door crossbeams to the depth of the hinges which were then screwed in place. The door was then set in place and the positions of the rectangular hinge ends marked on the wall frame. These sections were then chiseled out. The hinge ends were repositioned and screwed into place.

In practice, there was still some friction between the door and the door frame, so the offending sides were planed and sanded until smooth opening and closing was achieved.

The position for the mortise lock was then selected. Opposing holes were cut in the door and door frame for the lock components and a hole cut for the key. Bear in mind that the door was still just a frame at this point. The lock was tested and adjusted until I was happy with its operation. Building and hanging the door turned out to be the most time consuming part of the entire project.

Roof Construction Because of the height limitation of 6.5 feet and the wall height of 5 feet, the maximum roof height would be just 18". The minimum height was just 2". This was to be the next phase in construction.

Part 1 Part 2 Part 3 Part 4 Part 5 Part 6
SkyShed Backyard Observatory Building Plans
SkyShed Building Plans Lite Version

[phpbay]astronomy observatory, 100, "", ""[/phpbay]

Home Observatory Videos:

[tubepress mode="tag" tagValue="amateur astronomy observatory"]

Products on Overstock:

[phpostock]observatory,1[/phpostock]

Filed under: Astronomy Articles