Jason's dovetail headstock joint

[Jason Rodgers]

Hi, folks. I've been working on a new headstock attachment method. The idea is sort of a covered V joint, but cut with a dovetail bit and templates.

I've built four scarfed necks, with face plates and backstraps, and they're fine, but I want to try something else. Overall, I want to simplify neck construction and reduce the number of machining steps. I'm going after that with three main goals I hope to achieve:

1- Move the truss adjustment away from the nut. There are surprisingly many steps involved in creating that truss access pocket and cover. I'm going to start using the StewMac Hot Rod with spoke wheel adjustment nut and put it at the heel. One channel for the rod, one little cutout for the spoke at the end of the fretboard.

2- Simplify the neck shaft construction by isolating it as its own component. Necks will be maple laminate construction (perhaps with contrasting stripes), with truss and CF reinforcement bar, 24" long rough stock 1 inch thick. This will also cut down on waste, and tapering and shaping will be easier without an attached headstock in the way.

3- Simplify headstock construction by isolating it as its own component. Thicknessing and shaping a headstock while attached to the neck shaft requires maneuvering and dealing with the neck. If I can thickness, shape, bind, inlay, and drill for tuners without the neck in the way, that's a good thing in my small shop. I really like how Chuck Tweedy does this. In another test of this joint, there was very little carving that was necessary to blend contours into the neck shaft. Plus, I can make the headstock of a contrasting wood, matching the body, for example.

In the pictures, notice how the dovetail is completely contained within the blind mortise in the headstock and the flange around the tenon. I'll need to think about how this "volute" is shaped to be friendly to the hand, as well as aesthetically pleasing. I'm guessing this joint will be just as strong as any other cut headstock joint.

Like any bridal or V joint, the trick will be making that endgrain shoulder cut nice and clean. Any advice is welcome! I have a little tweaking still to do on the jig and templates, but I like the results of this test I did today. Your questions and comments are welcome.

[Bryan Bear]

That's pretty slick! I'd like to see how it looks once it's carved.

[Steven Wheeler]

I like it Jason. Like Bryan, I want to see that carved.

I'm assuming you used a router to cut the dovetail. What if you used the router to cut the shoulders on the neck? With the proper templates to follow, it should make it a one tool joint. Or at the least, make the chisel work to clean up the inside corners off the show part of the joint and give you an easy to follow cut line.

Steve

[Jason Rodgers]

I suppose I should include a shot of the jig for the router to follow. It's a little box. The headstock goes in the left side, and the neck the right. Right now, I secure them with wedges, pressing the work up into the template. That worked fine for the headstock, but in order to hold the neck securely enough, the tenon template deflects upward a little, which changes the geometry of the tenon just enough that there is a slight gap between the flange and the back of the shaft. I'm working on a better hold-down solution.

I could probably figure out a way to cut the shoulder square with a router bit, but it would require another bit and another jig, or at least a router table and fence. There is a 13deg angle cut on the neck shaft, like a typical scarf, which is the plane that the template follows. I do have four routers/trimmers, so if I was doing a batch, that might be worth the setup. If I'm trying to improve consistency and accuracy, getting my dodgy chisel technique out of the mix would make it doubly worth it.

[Chuck Tweedy]

Dude, that is VERY similar to my headstock joint - so, of course - it is awesome!!
I see your challenge in cleaning out the cheeks of the tenon - that is a difficult undercut to get clean.
You don't really need the backstrap over the tenon, if you got rid of that you would have much easier access to cutting that surface. But then it would basically be the same as what I do - so nevermined.
Overall it is really good - The trick is to make it so that getting a good fit is not a 20 hour process. I'm pretty much there, but I don't have that undercut to deal with. In shaping the neck you are going to cut into that pocket, so you need to make it clean. Actually, very Very little of the neck end of the backstrap/volute will be left. You will need to carve one of these back to a real neck profile to see how to proceed. On my joint, the cheek on either side of the mortice that sits on the shoulders of the tenon is very small after the neck is carved. You can kind of see it here: http://handcraftedguitar.guru/dVol_Page.html

[Jason Rodgers]

Thanks, Chuck! This is something I've been pondering and sketching for a bit, and then I saw your headstock and said, "Yep, I can make this work."

Right, I want the fitting of this joint to not take a full afternoon. If I can get everything fine-tuned, I should be able to do everything inside an hour. Though this method may have some fiddly steps with the jig, plus a few cuts on the table saw, it will be far less time than cutting a scarf, truing the scarf, gluing the scarf, gluing the ears, thicknessing the headstock, gluing the faceplate, gluing the backstrap, and then finally arriving at a point that the headstock can be profiled and drilled.

You're also right about things looking different once it's carved. I did another mockup a while back, all with chisels (solidifying that that is NOT the route I want to take), and the shoulder does get quite a bit smaller. The flange on this test is wider, but an initial squaring cut on the table saw, with minor material removal of the hidden inside corner, might do the trick. The corners on the leading edge of the mortise could also be trimmed back a little. That is one benefit of this joint that wouldn't be found on a through dovetail: the inner surfaces can be a little ugly, as long as that shoulder is clean.

I'll carve this test piece and report back.

[Jason Rodgers]

Ok, I got out to the shop this morning before my chauffeur duties began. I carved the profile with a straight spoke shave and followed up with 100 grit on a stick. The blending on the headstock was done with a 1/2" chisel followed by the sanding stick.

Remember, I still have some adjusting to do on the jig and templates to ensure properly mating dovetail surfaces, but I like how this is looking. I found that I could make about a 1/4" cut on the shoulders with the table saw and then back- or under-cut the small amount of material on the inside corner fairly easily with a sharp chisel. The flange around the tenon could be faired down even further, and the whole "volute" rises from the back surface of the headstock only about 1/8"-3/16" at its highest.

Notice, the ONLY carving done on the headstock is just under the ears. In the full process, the neck shaft is rough carved, headstock fitted, neck and headstock transitions blended, then the headstock is removed and worked separately until it's time to glue.

[Brian Evans]

So here is my question. Maybe I am not seeing this correctly, but I see p potential weak spot in the neck created in line with the cheeks of the headstock/neck join. If you discount the strength of an end-grain to end-grain glue joint from the cheeks of the cut, that part of the neck only has the cross section of the male dovetail and the overlapping projection that gets glued to the back of the headstock. It seems like a truss rod rout might reduce that even further. Aren't you worried about creating a weak joint at that point? I've probably missed something... It seemed to me that the Chuck Tweedy joint had virtually the full cross-section of the neck incorporated at the base of his triangular tenon.

Brian

[Jason Rodgers]

No, your observation is correct. The continuous grain from the neck shaft only passes into the headstock via the tenon, which is an inch at its widest. So, your concern is, for example, that in an impact the neck shaft might split along the edge of the tenon?

Remember, there won't be a truss access slot in this headstock, and since I'm using multiscale 24 fret fretboards, my fretboard is a little longer than usual, so the truss slot will probably be terminating around the 1st fret.

My glue surfaces will be three sides of the tenon, plus the flange-to-headstock area, plus the shoulders. Hmm, I suppose I could widen the tenon to shrink the resulting width of the shoulders.

[Brian Evans]

I'm just commenting as a guy who had two 1960's Gibson's lose their heads (in that splintered manner of speaking) about exactly at the point that I'm contemplating in your joint design. It took a cat and a guitar stand... and a small dog to chase the cat! If I had to guess, I would say that the amount of wood in the joint proper might be about the same as the amount of wood in a Gibson neck at the truss rod adjuster. I was thinking you might be able to maximize the dimension of the tenon there in some way. I do think the joint should be extremely stable mind you.

Brian

[Chuck Tweedy]

Yep, that will work. It looks better than my first mockup

EDIT: this sat a while before I posted so I didn't see Brian's comments.
The back-strap/ volute adds a lot of wood in the joint. It is in tension, the wide part of the tenon is in compression - it is actually shaped like an I-beam, so it should be stiff and strong.

[Jason Rodgers]

Since this mockup has served its purpose, perhaps I could glue it up and test to failure. Of course, being poplar, it wouldn't be a realistic test, but I might learn something about where the weak points lie. It might split into the shaft, but I'm guessing the more likely failure will be through the face of the headstock. I, too, have a 60s Gibson with a typical headstock break (before I got it, and repaired poorly), but short grain mahogany has nothing on long grain maple.

By the way, poplar is great for R&D. Easy on tool edges, quick to carve and sand, and inexpensive.

[Chuck Tweedy]

you'll need to glue a FB on there to make a suitable test dummy
oh, and cut a truss rod slot - and put a rod in it - and cap that.
OyE!

[Jason Rodgers]

Definitely something to serve as a fretboard. I think I have some oak that would work.

[David King]

WHile you're at it Jason I'd suggest (if you have time and wood), that you make a traditional scarf-jointed neck of the same woods and dimensions and see which one holds up better.

If I had to design this I'd see if there was a way to make the tenon come to a point and maybe even assemble the headstock from two halves so that the same point could easily be created in the headstock mortice/socket. I wouldn't want to bury a dovetail router bit into endgrain and then try to chisel out the remainder.

[Barry Daniels]

Your stated goal was simplicity, but it that was true a simple scarf joint would be impossible to improve upon.

[Jason Rodgers]

I'm not denying the simplicity or strength of the scarf joint. That is not a problem that needs fixing. Through the design of this joint, I'm trying to change the time invested in building an accurate and consistent neck unit. Quite necessarily so, a typical neck requires the most steps and time to assemble. While I'm far from finding my stride, it probably takes half or more of the total build time for me to construct the neck. No, there's nothing simple about this dovetailed joint. I see the compression of time and build steps around it (while maintaining or improving upon accuracy and consistency), though, to be simpler. Fender built the one-piece neck - with the step-down headstock and truss rod through the back - to achieve this. It might turn out to be more trouble than it's worth, especially if I can't fine-tune the templates and jig, but I'm going to give it a go.

David, what you describe is basically Chuck's solution. It's a good, repeatable process, and the results look great, too. You're right about the dovetail bit into end-grain: it wasn't too much trouble with poplar, but maple/walnut/cherry will burn like hell. Hmm...

I may not be able to get back to this until after Spring Break. For the impact test on this mockup, I was going to clamp it to the end of a 3 foot length of 2x4, stand it on end, and let it fall back to the floor. I think I have some cutoffs of the same boards that I could throw together for a scarf test. When looking at the results, no breaks would be great, but it might be more about seeing where it's likely to break and how easy/difficult it would be to repair. I use fish glue on everything, though, so that makes the reversibility of glue lines a little easier.

I'll report back! Thanks, everyone!

[Jason Rodgers]

I keep sneaking out to the shop, when I really should be packing. Here is my headstock impact test rig. Poplar 1x2 clamped to a maple plank. About 7lbs total. I tipped it over from standing, once striking the back and another striking the face on a block on the floor. No visible damage to wood or the glue lines, aside from dents where it hit.

Of course, the mass distribution in this test doesn't match a real guitar, nor is there string tension, but I think this joint will be plenty strong. When we get back, I'll clamp it up and apply pressure to the end of the back of the headstock until it fails. That will be more force than a guitar would experience in an accidental fall, but it will show the likely break points.

[Pete Halliday]

Chuck, do you mind sharing how you cut and fit yours? The Spanish vee I've been using is a time sink (at least for me).

[Brian Evans]

Re; breakage jig - stand on it (or the equivalent). It should ideally be stronger than plain wood in the same dimension. If/when it breaks, if it doesn't break in the joint, you win. It doesn't actually have to be stronger, it just has to be strong enough, but that is kind of a first approximation test. I still think the point where the dovetail tenon leaves the neck is the weak spot. The unknown is the thickness of the "back strap" that wraps up the neck.

I would like this a lot if I could build necks from commonly available wood stock, ie 3/4" thick boards finished 4S from the home store. Can you get that joint out of 3/4" or do you need 1"? BTW, kudos for coming up with this, it's really a nice idea.

Brian