How the light travels

A Newtonian is about as simple as a telescope gets, which is exactly why I chose it. Light comes in the open end, hits the parabolic primary at the bottom, and converges back up the tube. A flat secondary sitting at 45° kicks that cone out the side and into the eyepiece. That's it.

Choosing the parts

The math that set the tube length

Focal ratio is focal length divided by aperture. So an f/5 mirror at 203 mm gives a focal length of 1015 mm. Everything else follows from that one number.

The eyepiece focus sits roughly 100 mm outside the tube. Add 12.85 mm of tube wall and 114.3 mm from the inner wall to the centre of the secondary, and 227.15 mm of the light path gets spent before the primary to secondary run even starts. Subtract that from 1015 mm and the two reflective surfaces need to sit 787.85 mm apart.

Magnification

Magnification is just the telescope's focal length divided by the eyepiece's. I cannot change the 1015 mm, that is fixed by the mirror I bought. But I can change eyepieces and add a Barlow. Smaller eyepiece, bigger magnification.

Here is the thing I'd tell anyone building their own scope. Run these numbers before you buy your mirrors. My magnification came out way lower than I expected with the kit I had. Try it:

The disc stands in for how large a planet appears. Apparent size scales with magnification. With the 40 mm and no Barlow I get 25.38×, which is superb for framing something enormous like the Andromeda Galaxy and disappointing for Jupiter. It's small! With a 4 mm and a 2× Barlow, Jupiter is massive.

The glue debate

Search for the right way to secure a primary mirror to its cell and you find a surprisingly heated argument. Most people claim gluing causes astigmatism, where the mirror surface ends up with different curvatures in perpendicular planes and the image goes blurry. Others insist it is not an issue at all.

That left me nowhere, so I worked it from first principles instead. Three RTV silicone pads, 10 mm across and 1 to 2 mm thick, spaced 120° apart:

• The adhesive's shear capacity vastly exceeds the mirror's weight. Roughly a 14× factor of safety on napkin math. It cannot slip or detach.
• The pads are thin, so curing and thermal stress cannot distort the mirror. It essentially floats on flexible supports.
• Silicone RTV has an elastic modulus around 1.6 MPa against roughly 92,000 MPa for the glass. Being that much softer than the mirror, it yields first and acts as a natural adjuster instead of warping the optical surface.
• Three evenly spaced pads make a statically determinate system, which minimises optical deformation.

So, good to glue. I primed the wooden mirror cell with automotive primer to give the silicone an even surface to grab that would not soak it up, and gave it a full 48 hours to cure at room temperature. RTV has held aquarium walls together for fifty years under constant load. This should be good for twenty.

Assembly and first light

Once every tube section was printed, I applied a thin layer of JB Weld plastic bonder, snapped the connectors together, and let the whole tube cure. Black for the tube because black absorbs stray light and keeps noise down. White for the secondary vanes. I also cut a 2 mm slit in the top of the tube as an iron sights style viewfinder, for lining the scope up with things I can already see by eye.

Then I couldn't resist. I collimated it, carried the bare tube outside that night, laid it on the ground, and pointed it at the moon sitting low on the horizon. It focused. First light was a success. And I immediately learned something important about magnification.

The base

The Dobsonian mount comes from John Dobson, who worked out a simple, cheap altazimuth mount in the 1960s for holding big homemade Newtonians. I chose it for exactly those reasons. Affordable, simple, easy to use, and able to carry a large aperture for very little money.

The mount printed across eight plates. About 1.77 kg of filament, a little over a day of print time, roughly $44 in material. A lazy Susan bearing sandwiched between the mount and a circular base gives full 360° rotation in azimuth.

A good telescope also needs to be balanced, so it pivots about its centre of mass and doesn't tip forward or back when you're making fine adjustments. I did not predict the centre of mass correctly. I ended up balancing it with adhesive tire weights instead. Cheap, they stick straight to the curve of the tube, and they are easy to move around.

What I'd do differently

• STEP files instead of STL for the tube. It gives a much smoother edge.
• Thinner tube walls. 0.25″ walls at two perimeters and 10% infill turned out plenty strong, and would have saved a lot of filament.
• Metal vanes for the secondary holder. The printed ones were tricky to print and aren't strong laterally. Laser cut metal would be stiffer and easier to collimate.
• Not the cheapest lazy Susan. Mine lets the two plates separate and shift vertically, which makes the whole base wobble. Solution: don't buy the cheapest lazy Susan bearing online.

Next up is an equatorial tracking mount. Without tracking you can't take the long exposures that galaxies require, because your eye simply can't collect enough photons. You also can't stack frames the way every serious telescope does, Hubble and JWST included. That is its own project.