radius dish made with thickness sander

[tippie53]

the easiest way to make an accurate dish is the long compass method. this video shows Chris Paulik doing it.

[bunny]

So the rails here just have an angle instead of the radius? I can't find the forum I saw with detailed info on calculating the angle according to this Long Compass method.

[Stray Feathers]

I am a wannabe instrument maker, and I am certainly not a mathematician, so I was also a little confused by the Chris Paulik video (which I watched many times, along with other videos.) It appears that there is a sloping bed between the router and the operator, but the router itself appears to sit level through the process. I could not find other references and could not contact Chris Paulik, but ultimately came to the assumption that the bottom of the rail must have the arc cut into it, and, as this slides over the edge of the box, the router cuts the arc into the dish. It is also cut out to clear the dish as it moves along, which accounts for its asymmetrical shape I think. I don't see how this type of sled is better than having the router travel along a pair of rails that have the arc cut into the top of them. I was further confused by the description of this as the "long compass method" until I finally figured out that refers to the tool used to scribe the arc described by a long radius, and not to the particular method of creating a dish using a router sled with the arc cut into the underside of the rails. I find this method a little complicated and I think I will try a simpler one with a router moving along the arc cut out of the top of the two rails (the correct arc having been drawn using a long compass.)

[bunny]

The bottom of the rails is straight, there's no arc. It is the long compass already, not for drawing an arc on the rails but for routing it right into the dish. I found the discussion and the formulas on the OLF, Chris gives an explanation on the method to another math-challenged guy like us.:))

[MaineGeezer]

One idea intrigues me, though in practice it might not work well at all. If one had a high ceiling (in a garage or barn), one could attach, say, a 16' rod to a swivel joint at the ceiling, attach a router to the lower end, and use the suspended router to cut a 16' radius dish. One would have to either build a bench of the proper height to put the blank dish on, or figure out some way of adjusting the height of the upper mounting point.

One could also do it by cutting steps. For example, using a CAD program I've figured out depth-of-cut for a 2' diameter 12' radius dish every 0.1" of the radius. Set up the router over the blank, with a way to accurately move it along a radius in steps of 0.1", and rotate the disk under the router at each step. (I'll do it on my milling machine, with a rotary table, but I think it would be do-able with a router as well.) You'll be left with a slightly stepped surface, which could then be sanded to a uniform curve.

(The drawing in the attachment will be more legible if rotated 90 degrees right.)