Official Luthiers Forum!

I've been anxiously awaiting for that comment!

_________________
Jim Watts
http://jameswattsguitars.com

Alright all you nit picking kerfufflizers! A guitar neck joint is a half blind tapered sliding dovetail joint. Who's parts are still a tail and a socket!

One day a very young kid started helping out on a house building crew. The first day he used the nouns generator and compressor, both of which were on the site, interchangeably. I spent the day correcting him every time he confused the two. He was pissed at me all day, accusing me of needlessly nitpicking as some sort of malicious hazing. He didn't last long, (no surprise), and I sometimes wonder if he now understands the difference between the two, or still thinks all machines that end in, "or" and make a lot of noise are the same thing.

I've decided to return my LMI jig, I got an RMA number from them yesterday and as soon as they clear the mudslide that is blocking the road out I'll head into town to get a box to pack it in. It's just not getting the job done for me, very disappointing.

Meanwhile, I've been reading the Gore/Gillet book and I'm going to switch from dovetails to their bolt-on, bolt-off neck design on the next guitar I do. I'll need to make some fixtures for that, and change the top geometry a little -- I'll need to read more of the construction book to fully understand that.

Today I have some more dovetail fine tuning to do, I'd like to get one of these two guitars over the hump and start finishing. I suspect I have some more mistakes to make (and hopefully learn from).

_________________
========
Blog on Making Stuff: http://www.McGlynnOnMaking.com

Its hard to work with junk, a well fit dovetail is hard to beat. Wayne Henderson, his daughter Jane, and John Arnold cut both the neck and the matching neck block on a table saw like the early Martin factory.

Hold on Joe.
Consider trying to trouble shoot further.
Ignore the concurrent discussion going on her.

How do you set the depth of cut?

_________________
Ken

I set the straight router bit to touch the uncut surface, then set the depth stop using a 3/4" spacer. Plunge about 1/8" at a time and clear the waste. Then switch to the dovetail bit, set it to the depth of the cut surface (socket bottom or shoulder of tail, depending on whether I'm doing the final cut on the body or neck.

_________________
========
Blog on Making Stuff: http://www.McGlynnOnMaking.com

JoeM, I apologize for contributing to the thread side track.

I have re-read everything. While the jig may be inaccurate or poorly made, I don't see how it could have created your rocking problem.

Even if you ultimately decide on a different neck joint, it seems you are punting before getting the learning benefit of the struggle you've already made.

I thought Terence Kennedy's post was a clear description of whats most likely going on. I would re-read it and make sure it makes sense to you.

I agree with you about the boxy aesthetic of a flat body where it meets the neck, but if it is curved, and the neck shoulders angle across it, some fitting to eliminate the rocking will be required. Rough cutting the necks taper to the heal and undercutting the shoulders with a chisel should make it so only a small amount of material has to be removed. Trying to fit the shoulders to the body before the basic taper from fingerboard to heal is cut is bound to be a real chore unless you have a very large flat landing on the body as Robbie O'Brien's instruments do.
.

This seems correct but they must match in depth and the templates need to each placed front to back so they are correct.

If you do the mortise first on the body, use the straight bit to clear then change to the dovetail bit and make one single pass at the depth at which the bit touches the floor, your body is done.

Then with the neck held in, and the template suspended above, a new depth should be set for this member. Touch off the straight bit on the body end which is zero (z). Do this at the fingerboard end of the neck part. Because the neck is angled, the end grain part of the neck is not equal depth from the template. Then use the spacer to again set 3/4 inch (-3/4). Plunge in 1/8 passes until the depth is reached. Then change over to the dovetail bit and touch it down to the shoulder of the neck end and lock the depth. Do one single pass to undercut the neck tenon part.
They should slide in perfectly except one problem. The narrow end of the neck tenon (dovetail) is going to be too long by just a slight amount. Because it is angled, there will be more cut off the front fingerboard end than the back end on the cheeks.

If you take the neck out of the jig and try to slide it all the way in to the body mortise it will stop short because it is too long at the narrow end. Uust by a little bit, but significant enough to prevent neck from going all the way in. It will also rock because the parts only match when all the way in.

You can do a surfacing pass to route the end grain or take the neck off and cut it off or sand it off so this part is shorter than the socket. The end of the neck at the dovetail does not need to touch and actually shouldn't.

I think I can describe a way to line up the two templates so the members can be placed front to back correctly. Let me know if you are interested or of you are going another direction.

_________________
Ken

As I tried to explain above, this has been my experience as well. Finessing a dovetail can certainly be tricky, but I think this rocking leads many people (myself included) to assume something is amiss when all may actually be just fine. I also think most of the popular neck jigs were created primarily to help set the neck angle and don't necessarily improve the process of cutting and fitting the actual mated parts. This, I believe, is an important distinction that can save a fair amount of frustration.

Full disclosure: I've only done a handful of dovetails, so I am NOT an expert.

_________________
George :-)

That is the best explanation I've heard, thanks for that.

On the second guitar I sanded a dead flat landing pad on the end of the body, flat from the soundboard to the back and at least 2.5" side. I checked it with a straightedge in both directions and on both diagonals. After cutting the joint it didn't fit any better than the curved body on the first.

After a lot of test fitting and paring the sides of the dovetail it's nearly all the way home (again, with the length of the dovetail cut to about the correct length for the socket I'm starting with 5/8"-3/4" of misfit...that is the top of the neck where the fretboard will glue to is 5/8"+ above the soundboard.

I'm using blue dental paper to find the high spots in the joint. I only get contact points in the middle of the tail -- the middle along the protrusion of the tail, not along the length. Or said differently, the middle between the shoulder that abuts the body and the end of the neck. I'll have to take a picture as I'm not describing that well.

And the length of the contact point, especially when I first start fitting it is very short along the length of the tail. It does get longer as it drops into the joint further.

So it behaves as if the angle of the male and female parts is off (when viewed from the end). Or, said differently, like the angle between the two templates is slightly different, although I don't think that's the case. I'll check that today.

But in the machined joint, the fact that the contact point is only about 1/8" wide in the middle of the joint seems to indicate that the walls of the joint aren't dead flat. That could be lack of rigidity in the fixture or it could be a manufacturing error in the dovetail bit I guess.

So, criticism of the jig:

1. It's certainly not rigid enough. Even in Eric's video of his shop-built/improved version he talks about being careful about downward pressure and how the weight of the neck and router on the fixture throws the accuracy off.

2. It doesn't have the neck accurately positioned under the template, and there is no way of adjusting that. I drilled new locating holes for the pins, but it is still off a little bit. The misalignment is obviously a manufacturing error on my jig, the lack of adjustability is a design limit that is only a problem because of the manufacturing error -- if it was dead on I would care about wanting to adjust it. But the design should allow for manufacturing errors and tolerance stack up problems.

3. The LMI templates have the joint centerline engraved, but it's a wide line (maybe 3/32+), it's not really conducive to accurately aligning things. A laser engraved centerline would be helpful here. Or maybe an actual laser centerline beam.

Those are all things that could be improved on either in building a new jig or modding this one. I like the traditional aspect of the dovetail, and I like the simplicity, once it's accurately cut and fit you should be able to apply glue and a single clamp to seat it and be good to go. The bolt-on approach with a mortise/tenon has a lot more build complexity, but my guess is that it's more predictable than what I'm currently experiencing.

I may get inspired to fire up my CAD software and see if I can doodle an improved jig.

_________________
========
Blog on Making Stuff: http://www.McGlynnOnMaking.com

As mentioned the end of the tenon may need to be shortened to allow a little gap at the end so it does not bottom out and with a dovetail create a space for a steam needle for neck removal. Since the neck is angled the end of the tenon has to be shortened a little more at the tip.

Here is a little belt sander jig that works for that. Keys off the truss rod slot and has a small shim on the bottom for the angle.

_________________
It's not what you don't know that hurts you, it's what you do know that's wrong.