Some of the literature on violins suggests that the shape of the cross-section of the top and back is based on cycloids. Whether this is actually true is a matter for debate. Certainly much of the shape of a violin can be generated from circles so it is certainly possible that the masters based the shape on a cycloid which is generated from circular geometry. However, some modern makers say that they shape the top until it looks right rather than until it matches a template, so requiring the shape to fit precise geometry is probably requiring too much precision from the process.
Regardless, here is a script to generate the shape of the top of an arch-top string instrument using curate cycloids. I’ve written it in Python. I’m not experienced in Python so please excuse any errors and let me know how it could be better in the comments below.
The script takes in a PNG file that holds the shape of the long-arch down the centre of the instrument, plus the shape of the edges of the curved section. Using a PNG to hold this information means you can generate these curves however you like.
I destroyed the 3018 CNC router by trying to fit a more concentric collet chuck to the spindle. This moved the motor bearings. Given how problematic this thing has been I’m reluctant to spend more money on it.
So I’ve been trying to do the plates by hand. On the whole this has been more successful than using the CNC, given the CNC’s habbit of self destruction taking the wood with it. By hand is almost as fast and less liable to destroy the workpiece. I’ve been using a large gouge bought second hand on EBay plus a plane I made myself.
I’ve bent the sides round the mold – this was easy enough. Hair curlers work very well! I did it in two sections, leaving it overnight clamped to the mold to dry out.
I need a couple of small planes for finishing soundboards. These are expensive so I figured that the best thing to do was make my own.
The blade is a strip of gauge plate or ground flat stock – O1 carbon steel in unhardened form, readily available from EBay. I’ve had this bit lying around for years left over from something else. Being unhardened it is easy to shape with hacksaw and file.
I didn’t measure anything on this other than the angle of the bed of the blade (45º). The pictures should show how it all goes together.
The wood is oak from an old chair leg, cut on the bandsaw and sanded to a good finish.
This blog has been quiet for the last few months as I’ve been working on a big, fairly dull, project.
We’ve got a concrete sectional shed at the bottom of the garden that was put up by the previous owners of the house. Nice building – suspended ceiling, chipboard floor, plasterboard walls, good lights and lots of electrical sockets. Only problem was damp.
There was a damp proof membrane in the concrete floor slab. One problem was the membrane was 150mm underground. One end of the slab was embedded under a retaining wall; the uphill side of the building was paved level with the slab and had no drainage other than across the top of the slab.
I had to make clamps for the soundboard from threaded rod and bits of wood.
Making the clamps took longer than anticipated due to the number of them but they seem to work ok. I drilled the wooden bits at a very slight angle so they apply the force at the tip rather than near the threaded rod. Padding was via neoprene sheet glued on with Titebond – I’ve got a big sheet of the stuff left over from boat window gaskets and it works well.
I’m reusing a reclaimed parquet floor tile for the fretboard and other fittings. I think it might be walnut – not sure. What I do know is that the grain is very wavy and horrible to work – tear out is hard to avoid. I also found out what happens if you use hand tools to plane down the top surface of a used parquet floor tile – it took me a long time to work out why the tools became blunt almost instantly. I assume that bits of grit had become embedded in the top surface. Memo to self – saw off the top of the next tile!