Official Luthiers Forum!

15'/28' for us, but those radii are based on what has worked well acoustically, gives the correct fretboard extension fit with the building system, and provides adequate range on handling RH changes. Back radius does have some impact on acoustics, but there's no real coupling between back radius and the neck/body geometry we need to establish...in other words, you might just pick something that has worked for others and go with it.

For top radius, it's really a question of the building system. If you'll be wedging the upper bout to establish the fretboard extension fit, those builders using that approach seem to build with anything between 20 and 30 foot radius. For those using a flat or nearly flat 60' UTB glued on a flat caul, 28' radius provides good extension geometry for 12 and 14 fret bodies from Size 5 to 12 fret dreadnaught.

Having both dishes allows the body to be braced up and closed in the go-bar deck, which eliminates a lot of clamps. With a deck and dishes, 48 bars (24" x 3/16" with vinyl caps from Goodwinds) will run $80 delivered...or about the cost of 4-5 good cam clamps, and will let you get by with far fewer cam and F-style clamps. Also an option to rip up some straight-grained hickory or ash to make up wood spring bars...$15 worth of 5/4 stock gets out a lot of wooden bars.

I notice you don't have a large piece of sandpaper stuck on. Do you remove it after sanding the sides?

OK, don't everybody jump on me.

The old techniques for classical guitar construction included steps that deliberately did not devote effort to get the edge seam perfect. Tops were attached with tentalons (dentalons) and it did not matter if the the sides even touched the top. Then the edge was cut out for the binding, in which case even if you did match the sides to the top, you cut out the entire seam that you so carefully constructed.

The back is a little more difficult. Try to get it close, but do not run a series of clamps to hold the pieces tight. Instead use a backing strip to distribute the clamping force to bridge any local areas that do not exactly close. After you cut the binding channel, the occasional slight gaps are irrelevant. But, if you had run a whole bunch of clamps without the backer, you certainly would deform the back to close the gaps, which would only ripple the back and maybe set you up for future cracks.

This is not just a lazy expediency. The old masters, and many of the current ones, often developed the top with a very complex shape prior to assembly. The tentalon method is pretty much mandatory to get a fit.

I made a radius dish by gluing small wedges around the edge of a circle of 3/4 inch plywood and then placing a circle of 1/4 inch plywood on top, gluing it to the wedges and screwing it in the center. It's probably more of a spline curve but it is close enough for radiusing the top. It's virtue was it was cheap and I didn't have to swallow a lot of sawdust to make it.

That seems like a great idea to me. Do you still use it/like it?