Infinitum

[Akula]

Time for another rabbit hole of a build.

 

Four years ago, in the grip of a covid lockdown, I built my first all-in guitar after nearly ten years of hiatus. A five string bass guitar modelled after a popular high end model, and with it's flame maple top it was quite a beauty! In the time since, I've built about two dozen guitars, and I'd like to think my skillset has improved just a touch. Back in those days I was building in the garden on a folding trestle table, using a jigsaw as my main apparatus. I foolishly left all my tools in England a decade prior, and I was still months away from owning a bandsaw or even a plunge router. The instrument that resulted has done well, but there are so many things I could've done better... So it's time to strip it for all it's parts and hardware, and do a complete rebuild.

 

Three major differences from my original build. Firstly, I want to experiment with a curved radiused body. This means the body and the top will both have convex and concave surfaces, and fit together for the glue-up. It alleviates the need for a gut carve or an arm carve, and looks damn cool. Secondly, I will be doing a multiscale, with the parallel at the bridge line. It's a smaller difference in scale length to the usual, which should result in less of an angle on the nut, and the parallel at the bridge means I can use standard hardware and pickups. Really, it just allows me to have a slightly longer scale length for the bass strings, for a little less flabbiness. And thirdly, I changed the body shape and plans for this build compared to the last one. It was just a little too similar to the popular brand-name guitar for my liking - those on this forum who know me know I like to take influence but not carbon copy. My instrument is going to be longer, thinner, wider at the hips and smaller at the waist, and just all around sexier.

 

In short: a 5-string, chambered, radiused body, multiscale bass guitar. Not decided on timbers yet, but I'd like to see Walnut and Tasmanian Blackwood. I'd opt for a maple neck laminated with some other accentuating species, possibly purpleheart if I can source it on this side of the world.

 

Lacking a CNC machine, I figured the best way to impart a radius onto the body would be using a router sled. Here's what a bunch of pine sticks and plywood got me.

 

I made two sleds, one for routing a concave shape and the other for a convex shape. I had to crack out the math textbook to work out radii for drops in thickness over a chord length of the circle... Then the added depth of cut for the router bit, plus a bit of wiggle room. The good thing about working with timber as opposed to something like solid metal is that wood has a little bit of bend to it. If I can get both surfaces within a millimetre or two of each other, good glue and clamp technique will take care of the rest. I would prefer precision, however, to alleviate any future tensions on the workpiece. 

Here's the other sled, with a pine-and-ply demonstration body being worked.

 

Nearly a good fit. The jig needs a little fine tuning. 

 

I've been monitoring the mass of this demo-body quite closely. I can use it to work out the surface area of the guitar body, and extrapolate to the density of the hardwood timbers I'll be using for the actual build. Throwing in a good amount of CAD means I can plan against the neck dive which plagued my previous attempt. 

 

In all the fun and games, I broke a bandsaw blade. Damn! Get back to you guys in a coupla days.

 

 - Jam

[Asdrael]

I really like the radius body and I'm looking forward to see where it is going!

[Bizman62]

You're doing a helluva job in radiusing the top as well! I've built two neck-thru guitars with radiused bodies and I just bent the top veneer. A moist towel laid over the top will expand the outside and if clamped on the radiused body it will follow the shape pretty soon. I had to do that for the wings separately as the neck was visible all the way through.

[LFlab]

That's a very cool idea to route the radius on the body. For the other build I am doing, a fretless medium scale 5-string bass with a chambered, and radiused, body, I was thinking about how to tackle the shaping of the body.

Was thinking about using a handplane, but your jig seems pretty practical and not too hard to implement. It'll also be easier to do the chambering first, with a handplane, one runs the risk of breakout.

Do you have a drawing of what you are planning to do?

 

[Akula]

Thanks @Asdrael! If I can get this build anywhere near the precision of your builds, I'll be a happy man.

 

@Bizman62, I am preparing myself to bend timber for the top if I can't improve the radius jig. I've found my issue - the concave shape of the convex jig, if that made any sense, is deflecting with pressure from the router. I'll maybe fabricate another one with thicker timbers and more bracing. The deflection is resulting in all of my mathematics being thrown out the window, and for the top and body to not quite fit together. Even worse than that, the top is now not a uniform thickness across the curve, as the inside and outside curves now have different radii! Check this:

 

I'd love to aim for a total body thickness, perpendicular to the arc, of 40mm, and therefore I've left enough meat on my pine demo to try and give it another go. Here's a quick CAD view of what I'm really going for:

 

 

As you can see, machining the top to fit the body perfectly will be a good solution, but only if my jig can accurately reproduce an exact radius. Otherwise, it would be simpler to create a curved surface on the bottom of the body, match that radius on the top of the body with planes and sanding, then bend the top to match. 

 

17 hours ago, LFlab said:

Do you have a drawing of what you are planning to do?

I do indeed! They're not finalised yet, but I can send the unfinished plans in a PM if you'd like?

 

I've got a new bandsaw blade in the post, and a rather large sheet of clear acrylic for templates-making. This build will succeed or fail upon the quality of it's templates and the preparation of it's operations. 

 

 - Jam

[Asdrael]

Are you going to leave the very top flat to install a bridge or keep it radiused with single saddles?

[Akula]

33 minutes ago, Asdrael said:

Are you going to leave the very top flat to install a bridge or keep it radiused with single saddles?

The entire top will be radiused. The bridge is recessed, and that routing procedure will be done the same way as the pickup bays - with a flat template riding on shims either side to produce a flat-bottomed route on a curved surface. It's gonna look sick - probably!

 

 - Jam

[LFlab]

That's not needed. If you want, just the post the plans here when they are ready. Just interested.

22 hours ago, Akula said:

I do indeed! They're not finalised yet, but I can send the unfinished plans in a PM if you'd like?

 

[Akula]

Here's a rough outline of the thing. I'm still not sold on the headstock - I've been sitting down every night and moving bezier handles by a bee's.

 

The scale lengths are 863-876mm. I briefly went over this in the original post, but the idea is to extend the lowest string a bit while allowing for standard hardware and pickups. I'm going to have issues with the nut, though. I can't use the adjustable nut I currently have, and will need to make a new nut, yet I want the adjustable feature. So, I'll try my hand at machining a multiscale adjustable nut from brass. The idea is mechanically simple: a normal nut, slotted and shaped, riding upon two pins and sitting hard against grub screws on a brass plate, which is either screwed or glued to the shelf at the end of the fretboard. 

 

I've been working on a spreadsheet to calculate the weight distribution of the instrument. I figured, knowing the density of the building materials by weighing them in a known DAR format, I could use CAD to measure surface area times thickness, subtract any cavities and roundovers or carves, then do the same calculation for the neck profiles as a cross section, averaged, and times the length. The headstock was actually the easiest to calculate: surface area x 0.014m, minus 5 x Pi R^2 of 0.005m. Throw these into a centre of mass for both sections, and I've got a fairly good predetermination that the instrument will lean back into the strap quite nicely.

 

Next item on my list of mathematical endeavours is the neck dimensions. I've drawn up plans at a 1'and a 2' neck angle, and I've decided a smaller angle will allow for less grain run-out on the neck blank, and also letting me sit underneath the standard timber sizes here. Another degree could mean buying a 75mm thick board instead of a 50mm one, and wasting a lot of it. However, this means my neck blank won't allow for a straight run into the headstock, and shall require a scarf joint. This is interesting to me because of the nut angle - I'll have to attempt a compound scarf joint. The concept is really simple, but the execution could be rather hard with a multi-laminate neck blank. If the laminations won't line up, and I see no logical reason why the would, then I'll probably throw a veneer slice into the mix to distract the eye from the aesthetic issue. You'd see a shoddy line on a scarf joint, but throw in a dark veneer line or two between the joint and I bet you'd skip over.

 

I've ordered the veneer for accent lines on the laminated neck and scarf joint. It's a species called Dillenia. Looks quite dark, takes glue and stain well, fairly stiff. I'll try my best to soak veneers with black stain before gluing up, to add to the effect, but if it affects glue ability or won't soak uniformly through the fibres then I'll deal with having brown rather than black. Also got some carbon fibre rod in the post, plus a truss rod and some jumbo SS fretwire. 

 

Thinking of timbers for the build. Maple neck is a given, but the 5-piece laminated neck gives me some options to throw some other wood in there as well. I was thinking Wenge for lams number 2 and 4, but I may just go for maple all round with the accent veneer. Wings will be walnut, as I've been in love with that particular grain pattern and feel for a few builds now. Top? I'm thinking flamed Tasmanian Blackwood like my DC build a few years ago. I'd like to have at least one or two proper Australian woods in the mix, and I simply adore the look of blackwood. Tiger Myrtle is another option, as is fiddleback Redgum. Fretboard will be either Sheoak or Wenge, depending on the particular pieces on offer.

 

Quite looking forward to having all the bits and bobs in front of me. First cut is going to feel great.

 

 - Jam

[LFlab]

Looks great! Like yourself not sold on the Warwick headstock, and also, I think it is preferable to have the tuning machines not parallel but have them taper so the run from tuning machine to nu is straight. At least, that's how I designed in my parametric model, Especially that middle string on a 3+2 headstock is often at a severe angle.

Wouldn't mess with a compound scarf joint, just make it straight and solve it in the volute, as long as your glue joint is neat, the extra line is not too noticeable (and you could always put veneer on the back of the headstock)

[Bizman62]

1 hour ago, Akula said:

I'll try my best to soak veneers with black stain

A Crimson video about the subject looong ago revealed that plain soaking won't dye the veneer thoroughly. A vacuum chamber might help. Another option might be using a pressure kettle but boiling the veneers may cause them to disintegrate in the manner they make cellulose for paper.

[Akula]

11 minutes ago, LFlab said:

Looks great! Like yourself not sold on the Warwick headstock, and also, I think it is preferable to have the tuning machines not parallel but have them taper so the run from tuning machine to nu is straight. At least, that's how I designed in my parametric model, Especially that middle string on a 3+2 headstock is often at a severe angle.

Wouldn't mess with a compound scarf joint, just make it straight and solve it in the volute, as long as your glue joint is neat, the extra line is not too noticeable (and you could always put veneer on the back of the headstock)

Thanks! Yep I definitely like a good straight pull on the strings past the nut - my 6 string headstock, like most modern designers, is built around a straight pull. I designed around the tuner holes, not the shape. A re-design is in order. I shall, however, be messing around with a compound scarf joint. It's in my nature to try something I've never done before, however I shall be messing around with spare stock before I do it on the actual neck blank. Nothing like a badly scarf-jointed knotty pine neck to put myself off the idea and follow your advice!] There'll be time to experiment before the real thing.

 

11 minutes ago, Bizman62 said:

A Crimson video about the subject looong ago revealed that plain soaking won't dye the veneer thoroughly. A vacuum chamber might help. Another option might be using a pressure kettle but boiling the veneers may cause them to disintegrate in the manner they make cellulose for paper.

Righto! I picked a darker coloured timber just in case. There is a luthier supplier this side of the world who does "black veneer sheet" for about five times the price of raw timber veneer specifically for this reason. I'll deal with not having jet black accent stripes, they look just fine to me.

 

 - Jam

[Asdrael]

2 hours ago, Akula said:

Here's a rough outline of the thing. I'm still not sold on the headstock - I've been sitting down every night and moving bezier handles by a bee's.

If there is one thing I learned from overplanning my builds, is that moving by a hair on a CAD won't do much as soon as sandpaper gets involved. imo don't sweat it so much and just go with the flow once the rough shape is in.

 

2 hours ago, Akula said:

Next item on my list of mathematical endeavours is the neck dimensions. I've drawn up plans at a 1'and a 2' neck angle, and I've decided a smaller angle will allow for less grain run-out on the neck blank, and also letting me sit underneath the standard timber sizes here. Another degree could mean buying a 75mm thick board instead of a 50mm one, and wasting a lot of it. However, this means my neck blank won't allow for a straight run into the headstock, and shall require a scarf joint. This is interesting to me because of the nut angle - I'll have to attempt a compound scarf joint. The concept is really simple, but the execution could be rather hard with a multi-laminate neck blank. If the laminations won't line up, and I see no logical reason why the would, then I'll probably throw a veneer slice into the mix to distract the eye from the aesthetic issue. You'd see a shoddy line on a scarf joint, but throw in a dark veneer line or two between the joint and I bet you'd skip over.

Go for the "second" scarf joint (with the join in the headstock instead). Less issues really.

Looking good tho, keep at it!

[Bizman62]

2 hours ago, Akula said:

I'll deal with not having jet black accent stripes, they look just fine to me.

Sounds like a plan. Plus many dark woods look almost black when you apply some BLO, especially if you have some pale wood like maple around it. The "almost" part actually makes the contrast more subtle, from brownish pale to brownish black instead of yellowish pale to blueish black.

As for the headstock, I've also been sold on the straight pull idea. It just makes sense more than the fancy looks that require extra bobbins for tuning stability. Following your sketch, something along these lines:

[LFlab]

Your "Test it on scrap first" approach is commendable! Takes a bit more effort, but you do prevent major f-ups on something you've spent a lot of time (and potentially, money) on already.

[Akula]

Redesigned the headstock. Longer, thinner, a bit less bulbous, and straight string pull.

 

Dropped by the timber yard today and picked up a stash of wood. For anybody based in this hemisphere, Anagote Timbers is highly recommended. I sent through a list last night, and they replied at 08:01am this morning. I went down there after work, and we spent well over an hour picking out the nicest slabs and milling them into boards. My blackwood top came from a pretty thick slab, and they managed to get two sets of bookmatched caps out of it. And there's enough wenge for four fretboards, too, although two will have much straighter grain. 

 

Maple neck lams with dillenia veneer accents. Got this one glued up this arvo - the build has started.

 

 - Jam

[Bizman62]

36 minutes ago, Akula said:

Redesigned the headstock. Longer, thinner, a bit less bulbous, and straight string pull.

Looks nice, there's a Danelectro Coke bottle vibe to it.

36 minutes ago, Akula said:

Maple neck lams with dillenia veneer accents.

Those proportions are tasty!

[Akula]

On 6/21/2024 at 4:07 AM, Bizman62 said:

Looks nice, there's a Danelectro Coke bottle vibe to it.

Those proportions are tasty!

This headstock shall henceforth be known as "The Coke Bottle"! And cheers, the proportions for the neck laminations were considered thoughtfully. I did think about tapering them to follow the string path, but I figured I was doing enough complicated stuff for one build.

 

I cut my scarf joint and headstock angle at a compound of 13' away from string path and 22' across the nut. This required cutting the headstock face from the stock using a tilted bandsaw table, and then flipping the offcut for one more cut before gluing onto the underside of the headstock. I've never done an under-the-chin scarf joint before, only the ones where it meets the fretboard below roughly the third fret. I'm trying something new on this build, and with a backstrap and headstock veneer I think it'll be quite strong.

 

Here's the effect of the tilted headstock plane from down the neck.

 

I set about the task of creating the geometry of the wings, body radius, and neck plane. To do this I relied on my CAD drawings - I've never had my laptop in the workshop before, but it helped massively. 

There's a lot of angles going on here. A neck angle of 2' respective to the body, a taper of the neck-through blank which extends all the way to the end of the guitar, and an anhedral angle of the wings glued to the neck-through blank, to allow a slightly greater radius to be imparted upon the top. Ever try gluing a single-angle scarf joint and have the clamps and pieces slipping all over the place? Nightmare with three angles in the mix, so I simply used some hardwood dowels in tactical places. There's also a layer of veneer in between the wings and neck, because why not.

 

So here's the glue-up shot. I did the headstock ears at the same time, after having planed the funny-angled scarf joint sides to 90' true to the headstock face.

 

And the top, with the tried method of clamping cauls to the table and "popping" the top boards down into place and more cauls on top. 

 

I waited eagerly overnight for the glue to cure, before using my radius jig on the real thing. I've been planning this cut for nearly a month, so watching it work out for real was incredibly liberating. 

 

I realised very quickly that passes needed to be about 6mm apart along the long axis of the guitar - let's call that the x axis. If you go in the y axis (perpendicular to the centreline), you end up with imperfections in the jig manifesting as ridges and valleys across the grain. Along the grain is much easier to sand, plane, or scrape out at a later time, and much easier on the eye if you miss it. Note: we do this at work all the time - if you have a drum riser at a concert built out of 1x2m decks, for a total of 2x2m square riser, would you put the line down the middle across the stage, or up-to-down stage? Across stage, obviously!

 

I'm very glad I tried this out weeks ago on scrap - since then I've rebuilt the bed of the jig with higher rails (hence the spacers seen here), and rebuilt the sleds with stronger and thicker timber to avoid sagging from the weight of the router. I used G-clamps on the edges of the sled as stops to keep it square on the rails.

 

I'm quite happy with how it turned out. With a uniform thickness of 34mm, the body blank has a "drop" of 6mm in the centre over a width of 320mm, which is more or less the radius of 2100mm I was aiming for. 

Of course, removing that much material from the board can reveal imperfections that weren't on the surface. I knew there was a knot in the walnut which could appear, but I can probably hide that with a battery access cover. The dark marks in the maple are a bit more concerning, but there's not much I can do about it besides stabilise with epoxy and move on.

 

There was a lot of mathematics that got me to this point. Perhaps if I were using a CAD program slightly more powerful than Inkscape SVG I'd be better off, but deep down inside I just love whipping out the old scientific calculator. Let's hope I don't fark it all up with something dumb like a slipping router template.

 

 - Jam

[Bizman62]

27 minutes ago, Akula said:

This headstock shall henceforth be known as "The Coke Bottle"!

Just so you know, the "Danelectro Coke Bottle" is a thing, although they can't have registered it by that name for obvious reasons. "Jam's Coke Bottle", maybe?

30 minutes ago, Akula said:

Here's the effect of the tilted headstock plane from down the neck

Yikes! Looks like the neck were heavily twisted!

32 minutes ago, Akula said:

And the top, with the tried method of clamping cauls to the table and "popping" the top boards down into place and more cauls on top. 

That's a method that may not be clear to all readers. A small tutorial under 'Articles/Tutorials/Tools and Workshop Tips' wouldn't hurt. I've seen it done on a separate board but if you have time and space to occupy your table for a few hours, clamped rails works just as well.

[Akula]

31 minutes ago, Bizman62 said:

Just so you know, the "Danelectro Coke Bottle" is a thing, although they can't have registered it by that name for obvious reasons. "Jam's Coke Bottle", maybe?

Damn! Every good idea...  

 

31 minutes ago, Bizman62 said:

Yikes! Looks like the neck were heavily twisted!

Indeed! My camera angle is wonky, too, so the visual aspect of that photo is really out to get you. Trust, my neck is flat, square, and true, but with a headstock that angles downwards towards the treble side. It's a mindfart!

33 minutes ago, Bizman62 said:

That's a method that may not be clear to all readers. A small tutorial under 'Articles/Tutorials/Tools and Workshop Tips' wouldn't hurt. I've seen it done on a separate board but if you have time and space to occupy your table for a few hours, clamped rails works just as well.

I figured it was a fairly well-known method for clamping thin boards. I'll elaborate - say you have two boards at 160mm wide, you clamp two rails to the bench at a distance of 320mm. Then knock one of them towards the other with a hammer, ever so slightly but squarely. Then take your two boards to be joined, and pop them into the rails and press flat against the bench. The join will be under pressure. Simples! I'll gladly write up a proper tutorial if needed. It's a great method for joining thin boards.

 

Here's another shot of the walnut wings after scraping. Lovely.

 

 - Jam

[Bizman62]

54 minutes ago, Akula said:

take your two boards to be joined, and pop them into the rails and press flat against the bench. The join will be under pressure. Simples!

What if the two boards aren't rectangular? Figured top veneers often are of irregular shapes although symmetric. That's when you use wedges and potentially extra side rails for uniform tightening. Another trick is to lay a slat (appr. 40x10 mm) along the centerline under the boards. That will keep them loose enough for fine adjusting the figuration and also give just the right amount of clearance for good clamping pressure when removed.

[Akula]

Got some progress done, slowly working towards the moment where I can glue the top. I chambered the body, using an acrylic template to follow the radius. With small tolerances and a fair few angles involved, it's important to think about where the bottom of the router bit will be cutting as opposed to the top. I used a cove bit around the edges to follow the eventual external round-over, and also just to keep the inside of the instrument as pretty looking as the outside.

 

Electronics cavity and battery compartment.

 

I rough-cut the top and cut the f-holes as well, but neglected to take a photo. I actually template routed the f-holes this time, having always opted for a coping saw and patience. Worked pretty well, right up until I got tear-out. Chunk the size of a coin cracked away, but I found it on the shop floor and glued it back on with a few rubber bands holding it in place. Finished it off the next day, with a coping saw and patience.

 

Made a headstock veneer and backstrap with offcuts from the Blackwood top. Resawing timber to 1.5mm thick requires push-sticks. The headstock veneer is already glued on, but I'm currently in the process of bending the backstrap before gluing - it's been bent and is currently clamped to the back of the headstock and volute overnight. Hoping it doesn't spring back in the morning. Not sure if it would make a difference, but I had the forethought to use waterproof wood glue for joining the bookmatched halves, lest the joint open up under heat and moisture while bending.

 

Tomorrow I'm hoping to get the potentiometer holes drilled, and their recess dishes routed. I want to do that while the top is still flat, so I can bend it to radius and glue it to the rest of the guitar. Then the fun and games really start - final geometry, body shape routing, and the gigantic round-over.

 

 - Jam

[Bizman62]

Can't help myself, "f-holes" sound like a dirty word despite knowing exactly what they are and what's the origin of the word. It's just that an eff-hole sounds like something one could eff. Blame the effing prudes that decided that the f-word was vulgar after having printed it freely for 400 years in dicktionaries and other books.

[Asdrael]

Looking good, loving the confidence with the chambering

 

Question however: have you given any thoughts to potential neck dive? The body is obviously going to be lighter than a regular guitar body (lol Les Paul) and the neck heavier... ?

[Akula]

22 hours ago, Bizman62 said:

Can't help myself, "f-holes" sound like a dirty word despite knowing exactly what they are and what's the origin of the word. It's just that an eff-hole sounds like something one could eff. Blame the effing prudes that decided that the f-word was vulgar after having printed it freely for 400 years in dicktionaries and other books.

f-holes, nuts, and horns all around! What a provocative business this is!

 

16 hours ago, Asdrael said:

Looking good, loving the confidence with the chambering

 

Question however: have you given any thoughts to potential neck dive? The body is obviously going to be lighter than a regular guitar body (lol Les Paul) and the neck heavier... ?

The confidence with chambering stems from a good amount of planning and, for the first time on my part, CAD. I made two acrylic templates, one for the body shape itself and the other for the chambers, and comparing the two put me at ease while routing the chambers in a square body blank. Datum points provide a reference for positioning, both in the world of CAD and on the physical workpiece.

Potential neck dive... Yep, that's been on my mind a lot! So I made a spreadsheet:

 

Densities were taken from the timber boards themselves, straight off the mill I measured their dimensions and weighed them. Surface areas were provided by CAD, and the rest provided by good old mathematics. Now, I don't imagine I've got everything perfectly correct, but I reckon the instrument will be about 3.4kg when done, with well over half of that being in the body, not the neck. There are a few more considerations besides volumes and densities - my calculations are for the body (on the left) and the neck (on the right) - but how much of the neck is "behind" the location of the strap button? It's actually only about 80g, but hey. And of course measuring the densities and comparing them to the projections of the finished volumes can't account for the differences in densities through the timber boards themselves - what if there's a dense knot which gets removed during machining? I stopped short of pedanticism such as weighing the pickup mounting screws, versus the timber removed in a 2mm diameter hole. 

One thing that makes me confident about this instrument having no neck dive is this: I can draw a centre of mass point on the blueprints. My software is all 2d, it has no idea that the body is 40mm thick and chambered, or that the neck and headstock are 14mm at their thinnest, or that the weight of hardware in the body is greater than the weight of hardware in the neck, or that the timber densities are different. But the point lays around about here:

 

Anyways, I'll weigh the thing when it's done and we can all see how wrong I was 

 

 - Jam