Summer Build Off 2024: Bari-um Meal

[Quills]

As implied by the title, this is an entry for the Summer Build Off 2024. It's entered in the 'not-a-bass' category, but it's more a not-a-bass(partial) than a not-a-bass(scratch) as it's a continuation and hopefully conclusion of a delayed Winter-Build-Off that began over here. Check that thread for a (long-winded) discussion of 'how we got here'.

The specs on this build are the same as in that thread with 2 exceptions:

1.) Owing to shenanigans involving the neck (there were two previous abortive attempts), the neck will now be built from Utile/Sipo again (as originally planned), but it will have a finger board of 'Knobthorn' from Southern Africa (Senegalia Nigrescens, formerly Acacia Nigrescens). One of my suppliers had some and it looked interesting (janka hardnesses comparable to ebony, and its colour-tone almost matches the Utile for a double-roasted look but without the roasting). It apparently machines decently, but can be a bitch to cut, so I'll be prepping the fingerboard separately and then gluing it on, just in case it doesn't work out.

2.) I was originally going to use a 2k spray polyurethane for the body finish, but have decided against that and will probably go with something more domestic like water-based acrylic or oil-based poly (again, way-too-many details in the previous thread).

OK. That's it, I think. Good luck to everyone partaking.

 

[2groggy]

Hey @Quills , I followed the link to the WBO thread. Here's one more way to get a wider nut width - start with a twelve string neck and cut six tuner holes off of the headstock. You won't find an off-the-shelf bridge to match, but I bodge up my own floating bridges anyway.

I'm using this trick on all three of my builds.

 

[Quills]

Hey @Quills , I followed the link to the WBO thread. Here's one more way to get a wider nut width - start with a twelve string neck and cut six tuner holes off of the headstock. You won't find an off-the-shelf bridge to match, but I bodge up my own floating bridges anyway.

I'm using this trick on all three of my builds.

Darn; I didn't even think of that one. Thanks. That's worth a mention. I'll go back and edit the initial post to include that just in case anyone finds it by the vagaries of Google. In my case it wouldn't necessarily solve the length issue, but for anyone who doesn't care about the length it's definitely an option.

 

[Quills]

When it rain, it pour
I came to bring the pain once more (foreshadowing #1, trigger warning: extremely mild violence).

TLDR: I made a fingerboard.

That's enough of a head start, I think. Time to get this show on the road.

I started by taking the stock for the neck and fingerboard and thicknessing it to roughly where I needed it. Fortunately, I had some scraps that were close enough in thickness to act as snipe sentinels. The difference in sound between the narrower SPF sentinel and the knobthorn blank going through the cutters was hilarious.

I was aiming for somewhere around 21mm on the Utile neck blank and about 7mm on the Knobthorn fingerboard. I called quits on the neck at 21.4mm as it gives me some wiggle room. The fingerboard came out pretty close:

I took the fingerboard blank and planed one edge straight and square. I chose the line of the edge to try and make the best straightness of the grain lines, which are slightly wavey. I then used a mitre saw and shooting board to get the two fingerboard ends square relative to that edge. I'm doing the neck a little bit differently this time, order-of-ops-wise, in that I'm going to make most of the fingerboard relative to a fixed reference-edge and then use centre lines and locating pins to glue it to the neck in a mostly finished state.

Somewhere, during the Spring, budget carbide retailers Rennie had router bits on sale over here, so I invested in a 1/4" 2-flute down-cut bit. I used it to create the nut slot by running it along a straight edge on the router table with a straight and true-cornered backer-board acting as a sled:

In theory, this would fit the 1/4" Tusq nut blank I have perfectly, but it didn't; it was a couple of hairs too-tight for comfort, so I added 2 layers of standard white masking tape to the end of the fingerboard adjacent to the nut and did it again. This has the effect of pushing the board out from the straight-edge by 2 tape thicknesses, widening the slot by the same amount.

Warning: If you do this, even though it's a very thin cut, it's effectively a climb cut and the board can jump / kick-forward / catch unless you hold it down fairly-tightly close to the blade (this probability is worse with harder/denser woods). Try not to do it this way; it's one of those things that will work right up until the moment where it doesn't, with potentially disastrous results (for your fingers as well as the work-piece). A better way would either be to (a) move the straight-edge back a tiny amount with a feeler-gauge-spacing trick; or (b) to put the tape on first, make the rearmost cut, remove the tape and make the front cut; that way it's a conventional cut. I didn't do that because I didn't know how many microns / layers of tape would be needed; 'turns out it's two. Now you know.

It came out pretty decently, though:

I placed tape down the centre of the board, then used calipers to locate centre-line locations relative to the reference edge. I clamped a straight-edge down the centre-line and cut the tape exactly to the centre-line. I then removed one half of the tape, so the tape line was effectively the centre (or, at least, a consistent distance from the reference edge). I then clamped a ruler next to the tape and used it to mark out the fret positions. Using one of them-thar fancy protractors, in combination with an engineering square from the reference edge, gave me two ways to mark / double-check the fret positions, which I scored-in with a knife.

I measured all my positions relative to the nut-line, so I went back afterwards and double-checked that they were the right distance from each other individually.

Because my fret-slot-cutting jig isn't designed to use templates (mainly because I'm going to be using off-brand scale lengths and low volume output anyway ...), the fingerboard itself can be the template. So I used a some small saws (including the slotting saw) to cut into the knife scores on the edge of the fingerboard to give the main saw something to locate into once in the jig. This mostly worked well (foreshadowing #2):

I then took the board to my jig and cut away. Twenty-two times. (Why, yes; those are the wing-knobs from an old electric hover-mower; why do you ask? ...)

The board came out mostly great ...

... except for fret slot number 3 (aft-shadowing #2)...

What had happened was that I must have slightly mis-cut one of the reference notches, failed to check it, and then chose to reference off that one in the jig. Fortunately, I caught it reasonably early; the jig is so square that it immediately showed something was off, so I checked for correctness, referenced off the opposite notch (which was right) and cut it correctly, but not before one side came out over-wide to partial depth.

Normally, I'd be super-Pee'd-off at this point, but I wasn't panicking for three reasons: (1) When I put the chosen fret-wire in there, it still bit in the deeper slot and the wire was wide enough that it still seated over both edges anyway; (2) the fingerboard was yet to be radiused -- that would take quite a lot of the offending section off; and (3) the board is still over-wide at this point; there's about 14mm (~9/16") to come off, which will also remove a chunk of the problem, so I was fairly confident it would come out "OK".

Nevertheless, I decided to "fix it in pre-" just in case it needed a little help, so I readied the troops:

I had some spare knobthorn from the scrap, so I used 80-grit paper to prepare a pile of sawdust and mixed it 2:1 with epoxy. Filled when wet:

Still wet-but-gummifying with tape removed and some careful acetone mop-up:

The next day, dried, with stray epoxy pared-away with a sharp 1" (25mm) chisel:

With the slot re-cut:

With the slots and nut placements pretty much established, it was time to get to the radius. I found a 2"x3" (50mm x 75mm) cut-off from an old gate project and planed two adjacent surfaces straight, flat and square to each other. This was slightly tricky owing to one knot, but the jack plane took care of it:

I had literally just changed the blade in my marking knife and was in the process of carrying the centre-line around to the back of the fingerboard, when the murderous Mr. Bozo paid me the courtesy of a visit (aft-shadowing #1):

While cutting tape to lie adjacent to the centre-line, I caught myself enough to put a not-too-deep but somewhat annoying 3mm (1/8") deep 'notch' in my left index. It's nothing that needs surgery, but I won't be playing the bass for about four days:

I've been playing around with circular saws, table saws, angle grinders and routers for over twenty years and -- literally touch wood -- nothing bad has happened to me in the shop or on worksites (there were two incidents, but on both occasions the guards did their job. I love you, guards). The only times I've ever bled while on the job have been knives and chisels. Every. Single. Time. And, even then, the chisels weren't that serious, just grazes. It's always the knives. Always.

Anyway, because kitchen towels and masking tape are essentially trauma dressings and micropore, we tidied-up and continued.

I taped the fingerboard to the effective-router-sled 3"x2", matched exactly along the centre-line, set up the big router table and fence, and cheerfully joined @Gary_M and @Flying B in the routing-your-radiuses club. After one side (three passes):

I flipped it, matched to centre-line, and did the other side:

Truth be told, I should have been more aggressive here. By leaving an untouched strip down the middle it is effectively a low spot relative to the target radius, and I had to work harder to sand the sides later. I think the best bet would be to overlap the cuts very slightly, which would leave you having to sand a slight high-spot in the middle (relative to the target radius), which would be less work.

Radius-end art shot:

This was the 'glitchy' fret slot number 3 after the router-radiusing:

I probably didn't take the epoxy back far enough but, even at this stage, it's essentially fine. There's still sanding and cutting it down to width to go.

I then put 120-grit on a sanding block and went to work. Knobthorn is, to use the scientific terminology, b@stard hard and, had I known how resilient it was going to be, I'd have started with 80. However, it didn't take too much time or sweat.

I then switched to 240-grit and finished her off:

The 3rd fret slot came out even better (it's basically a non-problem):

Whew. Done. Neck neckst.


TBC.

 

[Quills]

It's like déjà vu all over again.

TLDR: Neck layout and truss rod slots.

Starting on the neck itself ...

The board I've got for the neck seems pretty decent (touch ... err ... wood ...), but the grain was somewhat off-axis from the blank edges so I had to spend a little while lining up the correct angle on which to cut a reference edge.

I had to leave enough room for the width of the headstock, while also putting the headstock over the one face end that had a tiny amount of snipe still in it (as that surface is going to get removed anyway), while picking a good 'average' line as the grain waves very slightly, while leaving enough scrap beyond the heel for locating pins and possibly fingerboard overhang. It took a while playing with a combination square and a template to compute the imaginary centre-lines that would permit this, but eventually I got it cracked and cut and planed the new reference edge along that line.

Once that was done, I could put centre-lines on both sides from the reference edge and begin layout. This one's a little tricky as the order-of-ops requires the template for the final rout be attached from the back (as the fingerboard is already radiused), which means passing reference points either around or through the neck blank if laid-out from the front The catch is that there are several potentially-easy locating-pin positions on the top side (which end up either removed or covered when top-routing) that you can't use on the back side (as they're neither covered nor removed and will remain visible).

Fortunately, I figured out that I can use the machine head locations for transferring front template positions through the headstock and then use the centre-lines and the nut line to align the fingerboard without interferring with the template. I just have to be careful not to cut within 2.5mm (3/32") of the neck outline before doing the final rout as the centre-line and body positions are accurate enough that this should give me enough wiggle room.

The truss rod is 6mm (15/64") wide, with the typical larger section for the welded bolt head. I also want a truss-rod access cavity so, after calculating the appropriate positions along the centre-line, I used an 8mm drill bit to set the extents for the access cavity and the bolt-head and a 6mm drill bit to set the final extent for the truss rod itself.

I then set up the router table and joined the larger holes in multiple passes with an 8mm spiral upcut bit. Once the larger channel parts were done, I did the longer section with a straight-cut 6mm bit that's pretty much the exact width of the truss rod (I seem to use this bit only for doing truss rod channels).

I tend to cut the channel very slightly on the short side, as you have to square up the ends anyway and you may as well use that time to finesse the final length of the channel and depth of the nuts (ahem ...). I'm pretty pleased with how it came out; the length and width are snug without being restrictive and everything is just-a-hair below the neck surface.

We fight on. Until zugzwang, anyway ...

 

[Quills]

Just in case it's not obvious and -- as with the last one -- this is nowhere near going to be ready for the deadline. As it is prone to do, Real Life(TM) has been getting in the way again. Ironically, I've just got to the point where I can spend more time on it again, so there will be updates in the coming days, but not with enough time remaining. Another challenge flubbed, alas, but I will be finishing the build in this thread.

Good luck to everyone still rocking or who has finished already. Ta.

 

[Reedt2000]

It happens

Please keep updating though, we're still following along

 

[Quills]

Four. It's the actual magic number.

So: Quick recap.

The first neck died a death because of some dodgy unforeseen grain causing a neck twist.

The second neck died a death because of (i) dubious wood selection (turned out to be spanish cedar instead of bossé) coupled with (ii) a non-terminal but annoying router mishap on the headstock.

The third neck (yes; the one up above) died a death ... well, maybe not died a death exactly but became a zombie, perhaps ... owing to the idiocy of working when your mind's not on the job. Working when your focus is elsewhere can lead to mistakes like ... for complete ... random ... example ... putting a straight-edge template on the wrong side of a neck blank and routing sufficiently aggressively to mean that the resulting neck would be several millimetres too narrow at the heel end (which was unalterable owing to the location of the truss rod slot).

A few millimetres at the nut end would just have resulted in a regular nut-width neck (which isn't what we're going for here but would be useful in future). A few millimetres at the heel is still a useful neck, but it really requires a different bridge, different pickups, a different heel rout and doesn't fit the plan in action, so it has been put to the side temporarily. It will rise again at some point; the nut width is still good, and I'm not against using / making a 49-50mm-spaced bridge and running narrower tramlines; it's the nut that's important here -- buttttt ... it doesn't fit the current layout / blueprint.

So the kid is nothing if not a fighter, and a fella just happened to have more utile laying around as well as some pretty cool-looking pink-ish guibourtia coleosperma (African Rosewood / Large false Mopane [none of that small false mopane muck around here] / Bubinga family).

One more time with even more feeling ...

This picture was taking while laying out the frets in more optimistic times, when the hope of hitting the deadline was still just about viable:

Post fret cutting:

Post radius-ing:

Aaaaaaaand ... it was at this precise point that my bandsaw's capacitor decided that the smokey Djinn had been captive long enough and literally blew-open.
Pop!

It was indicative of the type of week I was having back then. So while the new cap was in the post, I decided to use the downtime to work on the pickups instead, which will be the next post.

 

[Matt Liebenau]

Sympathy like. Definitely had bits like that. Happy to see you’re persevering.

 

[Quills]

Right. I'm going to try and keep the text to a minimum in this one, but there will be a lot of pics. Some are for my own benefit.

Short version: Alnico 5; fender spacing at the bridge; gibson spacing at the neck; nickel baseplates; 42 gauge PU. I wanted them slightly-overwound. Spoiler: kinda nailed-it.

Homemade pickup winder doin' some pickup winding. At home:

Couple of coils, post-wind:

Adding the dog leads:

Using a punch in the drill press to locate the slugs:

Assembling and connecting the pickups and hook-up wire:

All-taped-up and structurally-finished:

Neck DC resistance @ 13 degrees Celsius (55 F):

Bridge DC Resistance @ 13 degrees Celsius (55 F):

The DC resistance of transformer coils fluctuates in interesting ways with temperature (hotter = more resistance). I checked both pickups a couple of days later when it was somewhere around 20-celsius (68F), and they'd both gone up by about 200 Ohms (.2K), which kind-of gives the lie to the type of cork-sniffing that wants to bullseye a pickup DCR to <100ohms. That's only a couple of degrees of temperature variance in both scales.

When I've wound pickups in the past, I've always used a bain marie in a slow-cooker-type-thing to do the potting process. Last year I purchased a dedicated, reasonably-priced wax melter (Mylee) to see how well it would work (and because it was on clearance at Amazon). This was my first time using it. I followed internet advice and purchased a milk thermometer to keep track of the heat: Beeswax and candlewax have a habit of igniting at certain temperatures, and -- obviously -- you want to stay below them. Of some minor importance is that vintage-style, butyrate bobbins have a habit of softening and deforming at a certain temperature and, if using them, you obviously want the heat to be hot enough to melt the wax while low enough not to deform the butyrate. Being the sort of neanderthal who's perfectly happy with PU magnet wire, I'm also happy with ABS bobbins and they don't deform until higher temperatures, so this didn't concern me. Nevertheless, for the sake of consistency and safety, I was trying to bullseye 63-C (145-F) for the mix.

I eyeballed the amount of wax and got it pretty much spot-on. For anyone trying these Mylee wax melters, a mostly-full-ish bowl is about 320 grams. I used a 6.25% mix by weight (20g of Beeswax in 300g of paraffin wax). There is enough room to pot two pickups at once, but it makes moving them around to exhaust all of the air awkward, so it probably makes more sense to do it one at a time.

Starting-up (I didn't add all the mix at first as I didn't know how it would compress as it melted):

After 15 minutes:

After 30 minutes:

After 45 minutes (at which point I added the rest of the mix):

After 60 minutes:

As you can see, it takes pretty much an hour.

Neck pickup being potted with air escaping:

Bridge pickup being potted:

Because I was curious about the heat effect on resistance, I tested the DCR while the pickup ('normally' 8.4K) was in the hot wax @63-C ...

On my machine, this is roughly the point on the dial where you get a consistent 63-C (145-F). You can get away with going somewhat higher as the wax is melting (checking constantly with the thermometer), but I'd probably just stick it here and leave it:

Done and done:

I've got a few more updates before we're up to speed, but I'll get them up tomorrow.
Peas. And peace.

 

[Quills]

A few days afterwards, the new capacitor arrived and got installed, so I could carry on with the build.

I've shown a lot of this earlier in the thread(s; we've had a few run-ups ...), so I'll try and keep the commentary down.

I did the thing with the thing and the thing:

Then I did the thing with the thing (6mm for the channel, 8mm for the head and the access slot) and chiseled the ends to fit:

Checking the fit:

Added some space-age damping technology:

Henge:

Post-henge:

Did another thing with another thing:

Fender used something like a 52.4mm diameter drum here (2 1/16"); I used 38mm (1 1/2") as the headstock is shorter being a 3x3. It gives it a rather drastic-looking curve, but it's kinda cool. For the record: I use a router (you could also use a bandsaw) to set the inital headstock over-thickness step. I then use 80-grit until I'm within 0.8-1mm (~1/32") of the desired thickness, then 150-grit to smooth it out for the final approach. If you're doing this for the first time, those margins are too tight and you should be much more conservative until you have a good sense of the degree of change that occurs with each fence adjustment.

Having done this several times I've learned that, unless you exactly match the centre-line of the guitar to the centre-point of the full travel of the spindle oscillation, you're always going to end up with one side of the headstock thicker than the other as it receives differential wear owing to the bulge of the the compressed spindle drum. In practice, it's really difficult to do this as a chunk of the spindle travel is beneath the platform surface which means the travel centre-point is lower than you think -- and, even if you do it perfectly, you tend to end up with a very slight cross-sectional dishing along the centre-line of the headstock.

The solution is either:

(a) not to care (we're talking 0.4mm, 1/64"), or

(b) not to bother about the centre-point-matching and, having taken the bulk off going through the correct way (against the spindle rotation), then do a potentially perilous climb-cut/grind from the opposite direction to cancel out the imbalance. As long as you keep good control of the workpiece, the climb-grind is not too dangerous as you're only taking off the small asymmetry margin, not the full material bulk. Warning: Your fence needs to be properly flat and square to the surface for this to be true; the less perfectly-correct the fence is, the more this technique is considerably more dangerous as the asymmetry is even greater. On the climb-grind, be fully prepared for the workpiece to suddenly yank! occasionally as it grips; you have to keep it flat to the fence when it does this, or you'll have ripples in the surface, which is why conservative margins are helpful.

Anyway, I was aiming for 15mm thickness and called it when it was close enough:

I then trimmed off the sides ...

... and routed to the template (the fret slots needed some recutting, too, after the radiusing ops):

It's really interesting seeing the final proportions with a ~46mm nut. It does look kinda-like an acoustic, dimensionally. I'm digging the Guibourtia-Utile colour match, too. Should look great under some finish.

OK. Finally starting to get somewhere ...

 

[Quills]

Warning: dweeb-ish number-crunching and nothing else.

People who've read the other thread (and have functional memories as it was so long ago ...) may remember that I was occasionally weighing the thing obsessively as, at one point, it looked like it could well be a lunker (10-11lb). In post #40 of the other thread I said:

It's going to be a push to get it under 10lbs (4.54kg). At the moment, I'm targeting mid 9s, but that's optimistic; the hardware load on the body is probably a bit higher than 700g now that I think about it. It's all good, though; as long as I can get it under 11lbs, I'll be happy.

The current state is that ...

... the body is now ~2.5 kilos:


... and the neck is about 0.7Kg:


... which means the 'wood bulk' is 3.193Kg, which is just a third of an ounce over 7lbs. That's looking more optimistic for getting it towards 9lbs.

I then realized that I now had enough information to get an accurate-ish estimate for the upper bound of the weight. To save confusion (for me, at least), I did the whole thing in grams and then converted later (It's kind-of more intrinsically accurate, anyway: a gram is about 1/28th of an ounce).

Here we go:










Skrrrt. Quick maths:
2505+688+122+123+179+10+68+7+68+267+90+48+37 =
4.212Kg
which is 9 lbs 4.5 ounces.

(i) There are a few things unaccounted-for which will add to the total (grain fill, finish, control cavity cover, truss rod access cover, strings, solder, component wiring, shielding tape, neck bolts and washers, threaded inserts (if used), small screws for the covers and the string retaining bar), but they don't weigh too much; I'd be surpised if they were over 200g (~7oz) combined, possibly less.

(ii) There's a fair chunk to come off the above sum (neck carve, neck mortise, pup cavities, sanding, jack socket hole, string retaining bar recess, excess material (the fret wire and string retaining bar particularly), packaging weight).

Between (i) and (ii), is there a net weight-loss of 4.5 ounces (which would take it under 9lbs)? I don't know (depends on the exact hardware used for the neck joint, I think, plus maybe the finish thickness). The chamfering of the body took off just under 1 ounce and the body contours took off just under 2.4 ounces, so I'm conscious that selective wood removal doesn't take off quite as much as one might think. Having said that, the pup and neck mortices are cubic whereas the contours were kind-of oblique and not that deep.

Even so, I don't think there ought to be much of a net gain even if there is one so, that being the case, I think putting the upper bound at somewhere around 9lbs 7oz (4.28Kg) is conservatively fair; it could be considerably lower but I'm not confident either way. We'll see. If that's what it is I'm significantly happier with that than 11lbs.

 

[Quills]

The Hand Job (part I)

The time had come to do the neck carve. I was both looking-forward to this and slightly dreading it, but so far it's been fun. I was going to set up a router trench to do the taper but, quite frankly, it seemed like a lot of micro-faffing for something that had the potential to go wrong (either mathematically or in execution). This is partly because my good router trench is a hair too small for this neck, and the bigger one is too 'rustic'; I really need to make a good larger one.

So I decided to do it by hand. First the taper, then the carve. We'll only get partway there in this post. I use a similar method to Tornelli Guitars, which itself is just one minor variation of the way people have been doing this for hundreds of years; I think he goes too close to the nut, though, but that's just me.

Marking the taper with tape on both sides (the taper here is relatively small: 25-22.5mm (63/64"-7/8"), so you have to do this carefully) :

Create initial start and stop carves with something like a Iwasaki file. You want these to be mostly flat, but the important thing is that the bottom of the trench at the corners just kisses the top of the tape, acting like the 'maximum extent':

I do this at the front position (halfway to fret 1), at the 12th fret and at the final position (30mm in front of the 16th fret, which is the body joint); this gives some wiggle room for the fade carves later.

Mark the back of the neck and carve to the tape line, leaving a tent-like ridge down the centre of the neck. The angles aren't 100% consistent, but it doesn't really matter at this point; what matters is bullseyeing the top of the tape line, or just a micro-hair above it if you want wiggle room, as this sets the edge position of the taper. As pictured, I mostly used a shinto rasp for this, with assistance from files and sanding boards if needed.

Mark up the back of the neck. You can see the rough topology from the shape of the stop start grooves:

Start flattening the back of the neck with whatever you like; I mostly used the shinto as the interlocking grain of Utile can cause problems with hand planes. The marks will tell you when you're off to one side or another; this gives you time to get it relatively flat.

This is an iterative process; you may need to re-mark and re-carve several times to get an accurate picture of what remains to be done. As you approach the final line, though, you should have it pretty close. Then you can use a square, rasps, files and a scraper to true it up. It doesn't need to be perfect at this stage: the edge lines and the centre line need to be pretty close to perfect, but if the rest of the back flatness is only, say B+, that's OK; it'll throw out the facets a small amount, and the better it is, the smaller the cosine error, but micro-perfection is going to make no difference.

Pretty much done:

On to the facets.

I re-mark the centre line on the back using calipers at the 1st fret and 12th fret positions. This line stays undisturbed until the end of this process as it's the master reference for "where is middle" and should only disappear at the point where you have no need to know where middle is any more (i.e. last). I'm going for something approaching a "classic C", although it'll end up slightly flatter and D-ish in the lower frets as the neck is wider than normal there.

Marked-up:

Sneaking up on the tape line:

First facet pretty-much done:

Starting to kiss the tape on the second facet:

Not too shabby:

I weighed it again, just to check:

The taper and primary facets had taken off 70 grams, which is exactly 2.5 ounces. Encouraging.

To be continued in part II ...

 

[Quills]

Neck Carve part the second

Taped up the secondary tapers:

I mark the carves in pen (sharpie) from this point on as the angles between the facets are close enough that it can be hard to tell where the carve ends and the workpiece starts. Approaching-done carvery:

The carves were light enough at this point that I started using the large 1st-cut half round file to do some of the sneak-up-to-the-line work.
Secondary facet:

Both secondary carves done:

When it comes to these second-order facets, the measuring, marking and taping-up takes as much if not more time than the carve itself. It's worth it, though.
Tertiary facets taped-up:

The gaps in the centre of the tape towards the headstock in the image above are from me 'abusing' the vinyl tape's stretchiness to mark tightly to the line, not because the lines themselves are curved.

Tertiary / all the facets finished:

After taking off the high ridges with gently-used card scrapers (I love card scrapers ...), I then wrap-sanded in order to show the high points and smooth out the curve. Rinse and repeat. A neck starts to emerge:

Almost done:

At this point, I have some light finishing (mostly light sanding) on the main neck to do, and I still need to feather the ridges into the fingerboard edge, but it's almost done (and feels awesome). After that, which shouldn't take too long, we're then on to the headstock and heel blend carves.

To be contin-wah-d ...

 

[Quills]

Necking III

I didn't get too many pics of this stage as I was too busy doing it to remember to snap.

First thing I had to do was take off the last two facet lines right next to the fingerboard and then consolidate the whole neck carve, which was straightforward enough. Then it was on to the headstock and heel blending carves.

I started by drawing a circular section to define the end of the carve. I tried various things (including the can of Raid bug spray that's in the background), but eventually settled on a small can of WD40. The last time I made a neck, I started the blend carves with a pretty aggressive rasp, which worked, but I wasn't happy with the way it overly-defined the curve. This time, I wasn't quite sure of the best way to start, but I'm pretty good with a chisel, so I just grabbed one and set at it. It has to be properly sharp for this to work well. Wu Tang said that you have to protect your neck, so I stuck three thicknesses of low-tack tape on the adjacent curve, which turned out to be for the best as I (mildly) tagged it a couple of times.

Truth be told, most of the bulk was removed with the chisel, much of it freehand paring. It takes a lot off quickly, but it's also very controllable with different techniques and the cut is clean. It's then something of a game of chicken as to when you should stop. Errr ... about ... here:

I then reached for a small Iwasaki half-round rasp to complete the shaping. That thing is excellent for this job. I then rotated between the rasp, a 6-inch bastard-cut half-round file (which, on wood, behaves like a mild Iwasaki), a card scraper and two grades of sandpaper (p120/p220) wrapped around a dowel. That was it.

In all honesty, I possibly went about a dozen or two microns too far (partly because I was having fun), but it came out well-enough and most-importantly, it feels incredible.

Then it was on to the heel. I'd gotten my eye (and wrists) in now, so I just repeated the technique: roll of electrical tape for the template, chisel for the broadphase, Iwasaki for the 2nd pass. File, scraper and paper to bring it home. After the chisel and initial Iwasaki pass:

Near the end:

Neck done (mostly):

That's a major step (and one of the more intimidating ones) in the can. Pretty happy. Not quite sure what to do next. I'm torn between completing the neck and doing the neck mortise / bridge alignment on the body. It needs some thought.
TBC.

 

[Quills]

And we're back ... again ... again.

Quick life update: 'We' (as in 'fortunately not me', thank-cosmos-touch-wood ...) suffered not one but two broken humerus(es) in the immediate family in November. This is debilitating enough, especially over Christmas, but it had the effect of reducing me to "the only available and uninjured driver" for two households. As you can imagine, this put something of a crimp in my fabri-ma-cating. I only managed to get the odd hour here and there, so I made use of the time by making jigs as I didn't want to get stuck into things I couldn't take a run at. On the up side, I now have a fancy, new router trench and a working home-brew vacuum bag set-up, so ... wahey! ... I guess.

Anyway. People are driving again, and I'm getting more time in the shop, so things are looking up (thank-cosmos-touch-wood ...), so time to get back on the mission.

Following on from the above post, I decided to mostly complete the neck before doing anything more on the body. This requires doing a couple of fiddly little jobs (nut, truss rod cover etc.) before I was planning on it, but it has the upside that, once it's done, the body work should be straightforward, as you're essentially just fitting a pre-fab neck.

Necking IV: Grain fill

As utile is a relatively open-pored wood, and as I'm going to use some kind of lacquer-esque finish rather than just straight oil, I decided that the neck needed grain filling. I don't generally like pore-filling operations, but this one went so well I'm slightly revising my opinion.

I assembled various sanding implements ...



... and went at the neck, smoothing everything to about p240 ...



... feathering in the headstock and heel transitions and adding a micro-chamfer to the headstock:



I then taped off the neck.



I used Nitorlack water-based grain-filler in the mahogany colour (it's a really good match). I like it. The only thing to be aware of is that, if you're used to more general grain fillers, this one dries fast. Like: super-fast. Mine was drying in about ten to fifteen minutes. This means you have to move relatively rapidly if you're covering a large area. It hardens up over the first few hours and doesn't seem to have much, if any at all, shrink-back, which is good for a water-based product.

The general algorithm is: mix (well) with water to a consistency like melted ice cream; use a pad to apply the filler to the wood in a circular motion (i.e. both cross-grain and with-grain); use gentle scrapers (like putty knives / spatulas / palette knives) to take the excess off after about 10 minutes; let it cure for a while (2+ hours; I left it overnight); sand it back; repeat if needed.

Because of the rapid dry time, I had no time to snap pics while I was doing it, so here's the neck after it was mostly completed:



The next day, I sanded it back. Here's the back of the headstock with filler still on the neck and sides:



And here's the neck, pretty much sanded back:



TBC.

 

[Quills]

Much ado about not doing very much ...

Necking V: Truss rod access cover.
Warning: Pic heavy.

I kind of have a bugbear about crappy truss-rod access, which is one of the reasons I'd left a somewhat gaping orifice at the north end of the neck. Generally, I prefer truss rod covers, as they're discreet, and allow whatever nefariousness is occuring with the rod end and its access to remain concealed. There's just one problem: for simplicity -- and because it suited my tool availability -- I've built a fender style headstock, with the access hatch in a curved transition (and mine's steeper than Fender's). This is where you have to pay the karmic debt for not having taken the time to make a scarf jointing jig; you have to spend the time elsewhere instead.

My original plan for dealing with the problem involved 2mm black pinseal ABS. I was just going to heat it up above its transition temperature, use foil to shield the wood and literally bend it in place to act as the access cover. The issue with this is that ABS truss rod covers can look tacky if not done exceptionally carefully and well. It's by far the cheapest-looking thing about the early-run Ibanez BTBs, for example.

Fortunately, as I was fishing out the ABS in my stash of random small sheet goods, I found a number of small 0.6mm (3/128", 0.024") cherry veneers:



Instead of doing something relatively straightforward, we can do something unnecessarily convoluted and error-prone instead. Yay. So I cut four equally-spaced pieces. This should give us a starting thickness of 2.4mm (3/32"), which hopefully isn't too thick.



I grabbed a piece of random SPF, and did a sloppier version of what I did to the headstock to create the original contour.



In case anyone was wondering why I'd used the 38mm (1 1/2") drum originally, it was partly because I was using a shorter 3x3 headstock and needed a more rapid transition than Fender's, but also partly because I had a trick up my sleeve. This is what remains of some 38mm doweling (think it's oak) that we used for some curtain poles about ten years ago.



In fact, it's slightly smaller than 38mm, but in a way that works with us rather than against:



So I cut a piece off and planed a flat spot on it:



I glued and screwed another block of scrap oak to the back of the contour jig we made earlier so we can use it as a clamping surface.



I put some packing tape on the contact surface of the dowel, as I didn't want the oak tannins leeching. I was a little concerned about springback as the veneers were so dry, so I cut a couple of 4-degree wedges on the bandsaw; not enough to matter if there wasn't any, but might help if there was:



As mentioned, the veneers were dry as a bone, and they wouldn't have made it around the curve without snapping -- even though they were very thin. I checked that they were originally air-dried; they were (kiln-dried wood doesn't bend so well...). I decided to boil them, the same way you might do straight-grain pieces if they were thin enough not to need a steam-box.

I boiled them until they no longer floated. This took about ten minutes, but I left it about 5 minutes longer as I could see that the boiling process was leeching out some of the tannins from the cherry, and the more they matched the maple of the guitar top, the better as far as I was concerned. When they looked like they'd had about enough, I clamped them up with a wedge for some slight extra bend:



I left them in the jig overnight but they were still damp the next day (albeit bent), so I re-clamped them and left the jig in the airing cupboard for 4-5 days. When they came out, it had worked well (tightest curve on right):



I put packing tape on the curve jig to repel glue and glued them all together in the same jig, using one of the wedges as a caul for the top surface:



They came out like this:



I used the guitar neck to figure out where to cut the curve off, and used the bandsaw to take it to a more manageable size:



It was time to use CAD (cardboard-assisted design), so I broke out the cheap french curves and came up with something that fit nicely around the underlying cavity, but with room for screws.



It then went back to the bandsaw, where I cut it as close as I dare:



After some shaping and sanding:



After even more shaping and sanding:



Pretty much done:



My OCD (not the pedal...) wishes that I'd paid a little bit more attention to the grain orientation of the top veneer -- it was the least co-linear of the bunch and I somehow contrived to get it on top of the stack in the initial clamp-up and get it a bit off axis as well -- but I don't mind that slight disorder.

I may dye it to match the top just to repeat the motif on the headstock (as I'm not doing an overlay), but I may not. It depends what I do with the top. If I go more natural, it'll stay more natural etc.

Anyway; pretty happy with the way it came out.
Yup. Two posts in one day, but I had a bit of a backlog to clear.

 

[Matt Liebenau]

Much ado about not doing very much ...

Necking V: Truss rod access cover.
Warning: Pic heavy.

I kind of have a bugbear about crappy truss-rod access, which is one of the reasons I'd left a somewhat gaping orifice at the north end of the neck. Generally, I prefer truss rod covers, as they're discreet, and allow whatever nefariousness is occuring with the rod end and its access to remain concealed. There's just one problem: for simplicity -- and because it suited my tool availability -- I've built a fender style headstock, with the access hatch in a curved transition (and mine's steeper than Fender's). This is where you have to pay the karmic debt for not having taken the time to make a scarf jointing jig; you have to spend the time elsewhere instead.

My original plan for dealing with the problem involved 2mm black pinseal ABS. I was just going to heat it up above its transition temperature, use foil to shield the wood and literally bend it in place to act as the access cover. The issue with this is that ABS truss rod covers can look tacky if not done exceptionally carefully and well. It's by far the cheapest-looking thing about the early-run Ibanez BTBs, for example.

Fortunately, as I was fishing out the ABS in my stash of random small sheet goods, I found a number of small 0.6mm (3/128", 0.024") cherry veneers:

View attachment 7170905

Instead of doing something relatively straightforward, we can do something unnecessarily convoluted and error-prone instead. Yay. So I cut four equally-spaced pieces. This should give us a starting thickness of 2.4mm (3/32"), which hopefully isn't too thick.

View attachment 7170907

I grabbed a piece of random SPF, and did a sloppier version of what I did to the headstock to create the original contour.

View attachment 7170910

In case anyone was wondering why I'd used the 38mm (1 1/2") drum originally, it was partly because I was using a shorter 3x3 headstock and needed a more rapid transition than Fender's, but also partly because I had a trick up my sleeve. This is what remains of some 38mm doweling (think it's oak) that we used for some curtain poles about ten years ago.

View attachment 7170911

In fact, it's slightly smaller than 38mm, but in a way that works with us rather than against:

View attachment 7170913

So I cut a piece off and planed a flat spot on it:

View attachment 7170917

I glued and screwed another block of scrap oak to the back of the contour jig we made earlier so we can use it as a clamping surface.

View attachment 7170919

I put some packing tape on the contact surface of the dowel, as I didn't want the oak tannins leeching. I was a little concerned about springback as the veneers were so dry, so I cut a couple of 4-degree wedges on the bandsaw; not enough to matter if there wasn't any, but might help if there was:

View attachment 7170920

As mentioned, the veneers were dry as a bone, and they wouldn't have made it around the curve without snapping -- even though they were very thin. I checked that they were originally air-dried; they were (kiln-dried wood doesn't bend so well...). I decided to boil them, the same way you might do straight-grain pieces if they were thin enough not to need a steam-box.

I boiled them until they no longer floated. This took about ten minutes, but I left it about 5 minutes longer as I could see that the boiling process was leeching out some of the tannins from the cherry, and the more they matched the maple of the guitar top, the better as far as I was concerned. When they looked like they'd had about enough, I clamped them up with a wedge for some slight extra bend:

View attachment 7170922

I left them in the jig overnight but they were still damp the next day (albeit bent), so I re-clamped them and left the jig in the airing cupboard for 4-5 days. When they came out, it had worked well (tightest curve on right):

View attachment 7170923

I put packing tape on the curve jig to repel glue and glued them all together in the same jig, using one of the wedges as a caul for the top surface:

View attachment 7170924

They came out like this:

View attachment 7170925

I used the guitar neck to figure out where to cut the curve off, and used the bandsaw to take it to a more manageable size:

View attachment 7170927

It was time to use CAD (cardboard-assisted design), so I broke out the cheap french curves and came up with something that fit nicely around the underlying cavity, but with room for screws.

View attachment 7170928

It then went back to the bandsaw, where I cut it as close as I dare:

View attachment 7170930

After some shaping and sanding:

View attachment 7170931

After even more shaping and sanding:

View attachment 7170932

Pretty much done:

View attachment 7170934

My OCD (not the pedal...) wishes that I'd paid a little bit more attention to the grain orientation of the top veneer -- it was the least co-linear of the bunch and I somehow contrived to get it on top of the stack in the initial clamp-up and get it a bit off axis as well -- but I don't mind that slight disorder.

I may dye it to match the top just to repeat the motif on the headstock (as I'm not doing an overlay), but I may not. It depends what I do with the top. If I go more natural, it'll stay more natural etc.

Anyway; pretty happy with the way it came out.
Yup. Two posts in one day, but I had a bit of a backlog to clear.

Very nice. I never would have thought to do that but now I may, assuming I remember…

 

[Quills]

Very nice. I never would have thought to do that but now I may, assuming I remember…

Heh. Thanks. The profusion of pics and excessive detail in most of my posts is chiefly so that I can remember, several years hence.
("How on earth did I do that? ..." etc.)

 

[Quills]

Necking VI: Heavy, shiny nuts.
TLDR: I made a nut.

My original plan had been to use TUSQ for the nut and, as pictured above, I have a white one just for the purpose. However, I also have some 1/4" brass bar lying around (foreshadowing; we may be seeing it later), and it's been playing around the edge of my consciousness that I could use some of that for the nut instead.

For the record: I'm not in the camp that makes any magical claim for the quantum-harmonic properties of brass nuts -- most of my other instruments are Tusq and bone -- but in this case:
(a) most of the hardware is gold(-ish);
(b) I seem to have the world's last remaining repository of Jescar 55090 in Evo-gold and, given (a), this seems like a good time to use it -- and it may help balance the tone of the fretted note with the tone of the open note;
(c) I kinda feel like taking the scenic route.

I'm aware of the claims for brass jamming up with certain strings. In my experience, provided the nut is cut correctly and lubed decently, this tends not to happen. Also, if it doesn't work, I can always make another nut.

The brass itself is a hair under the 1/4" of the nut slot width, and a hair or two over 6mm, but I've figured out that two layers of scotch masking tape shims the back side of the brass bar stock perfectly. I'll just have to trim it to relative invisibility once installed.

I used some Poor Man's Dykem (PMD; aka Blue Sharpie) to mark up the bar stock to the width of the slot and approx 7mm high:



I thought it wise to break out the little 3" (75mm) Irwin vice en route to hacksaw ridge:



I carefully filed then ends to width and then did a test fit:



The width is perfect but the height is too ... conservatively excessive. More sawing required. So I lightly scribed a line along the top of the fingerboard ...



... and used a radius gauge to put an equivalent line just under 4mm (~5/32") above it, then trimmed off the excess:



That saved a lot of filing, but there was plenty remaining. At this point I got more caught up in the doing than the photo-ing, but I started on the back edge and attempted to match the curve all the way to the front, while maintaining a downward direction backwards. When the curve intersected with the scribed high-line curve on the front, I stopped.



Test fit in situ:



I then PMD'd the front face again and, along it's flat bottom edge, scribed out the calculated string positions using the calipers (where any remaining marks will be hidden by the fingerboard). I then used an engineers' square to transfer these locations carefully to the top edge; this was to stop any cosine errors from measuring around a curve.

The next step was to take a thin (0.25mm / 0.010") modelling saw and cut the marks. This was followed by enlarging those cuts with a 0.4mm (0.016") modelling saw. The goal isn't to go anywhere near depth; just to locate the positions. I then took two triangular needle files of different sizes and enlarged the grooves.



Obviously, these slots aren't correct -- they need to be lower at the rear than the front for starters -- but it's to establish the initial working positions. By my calculations, there's at least 1mm of leeway left to cut -- possibly 2mm.

I'm reasonably happy with how they came out. My only annoyance is the position of the G string; it's not terminal but it's nearer the D than I'd like it. I think I must have filed more on one side than the other at that point as everything looked good after the saw cuts, but it's close enough for now, and I have enough leeway (I think ...) that I can finesse it over a bit using the safe edge on the triangle files before cutting the final slots. That'll have to wait until final set-up, though. I'm not committing to anything before that.



Anyway; I think we're finally there. Time to start fretting in both senses. TBC.