WBO 2024: Bari-metric Pressure

[Quills]

Managed to get a few hours in the shop today for a change. It was a three-steps-forward-one-step-back kind of day. Again, I didn't get anywhere near as many photos as I should, so there's more unnecessary text instead.

I started by cutting 'tabs' in the trued-up side of the neck. These have a few purposes as will be seen shortly, but chiefly they allow me to continue to use the flats of the tabs as an effective reference edge, while also creating a neck edge that can be used as a reference not only for positioning, but also for taking the tabs off when I'm done with them. There are reasons for this, mostly to do with order-of-operations, keeping the centreline parallel with various reference surfaces and how I want the fingerboard to have a slight overlap on the end, but also get carved-away just behind the nut. I thought of other ways of doing it but, although clunky, this seemed like the most straightforward without spending hours carefully configuring router jigs.

I then went to the bandsaw, cut away the waste between the tabs, reattached the neck template and routed the waste areas as if they were the finished line of the neck. I had to create a new locating pin point in the heel, as the old one got swallowed by the truss rod, but it's fine. It's not actually on-centre, but it doesn't need to be; I did the old "align the template carefully and clamp it, then drill the locating hole through the template" trick.

I then did the truss rod adjustment cavity. I wanted to do this before attaching the fingerboard to make sure everything worked and that I could make the connecting hole from both sides.

I used an 8mm (5/16") brad-point bit to make two holes on the centre-line in the appropriate places. I only went in a few mm (maybe 1/8"). I then swapped the drill bit for an 8mm spiral up-cut router bit in the drill press and drilled both holes to full depth (10mm / 3/8"). This was so there would be a flat bottom and the brad point would neither be present in the hole, nor would have gone deeper if I'd drilled the whole thing with the brad points.

I then used a knife and some smaller chisels to excavate the hole to the right depth. To be honest, I didn't overly care about neatness (or I'd have used a router template) as half the hole is going to get cut / abraded away when I do the paddle-headstock transition and the hole is going to be covered, so provided it was 'roughly' neat and functional, it was good enough.

I'm lucky enough to have 3mm (~1/8"), 4mm (5/32") and 5mm (~3/16") aircraft bits. I used an awl to locate a hole about 3mm above the bottom of the cavities on both sides. I then used the 3mm aircraft bit to drill on as-flat-a-trajectory-as-possible from both sides until the hole connected somewhere under the 'nut bridge'.

The clever bit is: you can use a small chisel (in this case 4mm) oriented horizontally bevel-down slightly above (1-1.5mm | 1/32"-1/16") the hole on both sides and split the wood so that the roof of the 'tunnel' collapses into the hole. You then get the 5mm aircraft bit and put it into the split rather than the original hole. This forces the larger bit to drill on the high side of the pilot hole as the lower side has material packed into it (from the tunnel collapse). It worked really well.

I then put the truss rod in and checked that you could actuate at least 60-degrees of turn with a hex key from the hole as-is. You can do it both short-leg-inserted and long-leg-inserted (behave) with a typical allen key. It's pretty tight (more so with the long leg as the angle is more acute owing to the front lip of the hole), but you can do it both ways, which is more than good enough as, again the hole is going to open-up further with the headstock transition.

I then needed to start work on the platform that becomes the paddle headstock behind the neck transition. My idea was to remove most of the waste on the bandsaw before using a router to finish the platform and then the horizontal-belt-sander to finish the curve. Removing the waste on the bandsaw went fine -- in fact, I should have done the whole platform that way and gone straight to the belt sander, as the reference tabs worked perfectly.

The problem came with the router which was where I made the first real mistake of the entire project. I was using an extended base to increase the reference surface but was watching the bit so hard that I lost concentration for a split-second which, as is often the case with routers, is more than enough to ferk everything up. The router ended-up tipping slightly which caused the bit to plunge into the surface. I knew I'd screwed-up just from the sound. I moved the router away, and turned it off to have a look. The news was not good. Mercifully, it was an isolated mark and was confined to a non-critical area, but it's to the point where you can't just sand it out and hope nobody notices.

I see the future: There's going to be either some fairly-serious headstock reshaping or some creative accent-piece inlay work occurring. I haven't decided which yet.

I stopped doing it the way I was doing it and did the rest of it the way I should have done it in the first-place: on the router table, with flat template edges. The rest of it came out fine.

Can you see the mistake (it's very subtle)?

I decided to just accept it for the time being and continue with the plan.

I ran the fingerboard through the thickness planer bring it down from 9mm to about 7mm. There was some slight tearout on the surface as the figure in the fingerboard has some reversing grain, but it's not too bad. It looks worse than it is as most of it you can't even feel; it just registers as an optical difference. I'm reasonably confident that the ones you can feel ought to sand out when the radius goes in.

I carefully aligned the fingerboard with a line about 10mm (3/8")beyond the nut line and did it in such a way that (a) the ribbon pattern in the grain would hopefully float relatively-consistently down the fingerboard and (b) the fingerboard itself substantially overlapped the reference tabs, but did not overflow them (so they can still be used for referencing). When it was in the right position, I clamped it in place.

I then took it to the drill press, put a 4mm (5/32") brad-point bit in it, and drilled through the fingerboard into the tabs.

I then cut 4mm dowels / florists rods to length to use as chunkier locating pins:

I took some 120-grit sandpaper and scuffed-up both the surface of the neck and the back of the fingerboard to give them some tooth for the glue. I then took some acetone and thoroughly de-greased the back of the fingerboard. I almost didn't do this but was glad I did as a ton of residue and oil came out of the board; in fact, I needed two paper towels to get it all.

I then (thankfully!) remembered to put the truss rod back in the slot and intermittently added some very small beads of rtv silicone to the top of the rod, focused towards the centre; this is hopefully to stop it rattling.

And then it was just a case of (a) gather clamps; (b) pre-rip some paper towels; (c) perform bathroom check; (d) take a swig of juice; (e) apply glue and ...

(f) Clampus Hengeus.

We'll see how the recipe turns out tomorrow.

 

[Quills]

Good against remotes is one thing. Good against the order-of-operations; that's something else.

The more I thought about it, the more I realized what a pain-in-the-hole the router damage from yesterday was. It's in a fairly innocuous place, but the sequence of how the fingerboard gets shaped to the neck requires that the paddle of the headstock be done early (not least because the camber behind the nut is dependent on the tabs and I can't remove the tabs until I've done the camber. In addition there are other logistical blockages, such as that you can't reliably use a router table on the front of the headstock once the fingerboard's been radiused. So I realized that it needed to be sorted out sooner than later.

I took a piece of paper and drew around the actual headstock, making a note of the location of the damage. I then scanned the outline and put it in Gimp, rescaled it to the right resolution and aligned it with the original plan image. I then made several attempts at redrawing the outline of the headstock so that I could just reshape it on the horizontal belt sander ...

Some were OK, but no matter how many times I re-jigged it, it didn't look as good as the original shape (partly because 3x3s need a bit of space above the D and G tuners otherwise they just end up looking truncated).

So I decided to take what's frequently the best course of action with a damaged work-piece: damage it some more. I wanted something that should look OK but shouldn't take too many hours to implement, so what I came up with was this:

I'm going to inlay a piece of walnut (the teal section) within the headstock and then inlay some of the dyed figured maple (the red-brick section) within that. It should hopefully echo what's going on with the body top.

As far as logistics are concerned, I only need to get the walnut and the cavity for the maple done now; the maple itself can be done later (which allows me to more safely match the body dye for extra bullsh coherence). The pattern itself is based on a 74mm-diameter circle (roughly 2 and 7/8") and the walnut/teal is supposed to intersect with the edge halfway-though the column of the G-string tuner.

So I needed a 74mm template but, ideally, I also needed something that would hopefully be on the same plane as the un-radiused fingerboard, so I could use it straightforwardly with the neck upside-down on the router table. The drop from the top of the fingerboard to the surface of the paddle is 11.5mm (about 1/32" less than fender's) so, ideally, we needed something about that thick for the template. I thought I might have to mill something but, bizarrely, one batch of "12mm" plywood that I had in the scrap pile turned out to be within 4 thou (0.1mm) of it. Thanks, universe!

I stuck it on the drill press and cut a 74mm hole using a hole cutter:

I bandsaw-cut the template so it would fit on the headstock, then marked the points where the circle intersects with the edge of the headstock. I aligned the template with the marks; clamped the template to the headstock; turned it over and marked the position of the headstock on the back of the template. This gave me accurate enough positioning to position the masking-tape for the tape-and-CA-double-sided trick.

I then set-up the router table and carefully gouged more of the end away. It came out pretty decently.

The rout itself is about 4.5mm (just under 3/16") deep. I have some thicker walnut stock I could have milled down, but I decided to laminate two pieces of 2.5mm sheet/veneer together, (a) because I have them to hand, (b) because they're the same sheet that was used in the body, so will be a pretty good match, and (c) they'll only be slightly proud of the surface of the headstock, so they won't need any extra processing: I can sand them down during the final headstock paddle profiling on the belt sander. I want the walnut to surround / 'encase' the maple both from the sides and below, and this will allow me to use ~3.5mm (1/8") maple while still having room to get beneath the glue line of the walnut lamination.

Checking the fit (it's good ...):

The only downside is that I need to use quite a large amount of walnut for what's actually going to be a tiny accent piece, as I'm going to use a reversible hole cutter to cut the disc that fits in the recess, so the workpiece requires clamping, which means it has to be substantially oversize to clear the hole cutter.

I cut two pieces of 130mm-long (5 3/16") walnut sheet; pseudo book-matched them, so the grain reflects and the slight cup is in opposition; quickly knocked up a packing-tape caul; used the extracted disc from the template as the central pressure caul and ... yup ... it's that time again ...

Clomp hinge:

Until the morrow ...

 

[Quills]

A good day's work. Managed to grab 50 mins this morning and 80 this evening. Got quite a lot done.

I've seen people online (not here) talk in very scathing terms about the General Tools / Mibro fly/hole/disc cutters and their ilk (notsponsored). In my experience they work pretty well provided you actually read the small print (i.e. the instructions). Short version: (a) never use them in a hand-drill, only ever in a drill press; (b) make sure you clamp the work-piece down tight and check that the tool has the necessary clearance over the clamps; and (c) keep your rotational speed under 500rpm -- and, the wider the hole, the further under 500rpm it needs to be (surface speed increases with radius etc.).

I tend to find that most of the problems usually emanate from violating one of the above and it's frequently (c). The problem with (c) is that most of the cheap bench-top drill presses can't go under about 550rpm, let alone get down to 200, which might be needed for larger holes. It's for this reason that I always recommend to people that, even if you're going to buy a cheap chinese drill press, spend 200 currency and above rather than 150 currency and below (it's one of the biggest 100-currency differences in tool purchasing). Big forstner bits need to run under 300; hole cutters need to run under 300; some glass cutting bits need to run at under 200. For the sake of 100 quid/bucks, not only will you get more room, a better table and a longer throw, you'll also get 9/12/16 gear ratios instead of 5. You're only buying half a press if you can't get the speed down.

Anyway, if anyone has one of these cutters and doesn't have one of these jigs ...

... you should remedy that immediately. I knocked mine up quickly for an expeditious project about 8 years ago on the assumption that I'd make a better one at some point. I haven't. It's functioned fine for all that time. The trick is ensuring that the 1/4" (6.35mm) hole is centred on the same plane as the ruler scale. This minimizes cosine errors. If you have more than one of these General/Mibro cutters (they make several), your jig will work with all of them as they all use a 1/4" pilot bit. They're simple to make and use:

The keen-eyed pedants among you may want to point out that the scale isn't completely aligned with the edge of the board, so I may be getting a scaling cosine error in that plane. Rest assured, however, that the important plane is the other edge, as that's the edge that's on the drill press when you drill the hole and is therefore at right angles to the drill bit.

Not that my jig is perfect by any means. I've got the scale perhaps 0.2mm (1/128") too far from the hole, which in practice just means setting the cutter to your desired distance and then removing just-the-right-degree-of-smidge. I've got it down now.

I set the radius to JTRDO-smidge under 37mm (being half of 74mm, if you're up to speed on yesterday's post), and cut a test disc. One of the things about using these cutters to do anything serious is that you always have to cut a test hole/disc, sometimes several; don't assume it'll bullseye first-time. It's like an offertory to the Gods-of-making-stuff; cutting the test hole ensures you'll bullseye it first-go 95% of the time; not cutting the test hole means you'll always be off first-go. Always.

Oh yeah. They're messy boogers, too. Especially if you're going slightly too fast for the radius as I was here.

Anyway, because I'd made the right sacrifice -- here, a scrap piece of baltic birch -- of course I bullseyed it first-time. I tested it by placing it into the original template ...

... then tested it against the actual work-piece:

More than good enough for Gibson.

I then took the laminated walnut board from yesterday and used the same cutter in the same configuration. As I have more than one of these cutters, I'm going to leave this one set-up as-is, so I can use it on the maple inlay with the same template too.

I realised that I could get a couple of incomplete shapes-worth if I drilled it a certain way, which would give me more choice of grain layout and cut quality, so I did. I picked the best one, applied high-viscosity CA glue to both the walls and the base of the cavity in hopefully the right amount, oriented the walnut piece in the hole, held it by hand until it tacked-up, then got a locking clamp on it.

I should probably have cut it nearer the finished size first, but I was rushing to get it done in the time window I had.

I left it like that until this evening. It came out reasonably well:

I cut it down on the band-saw (with a support platform underneath) and then flush-cut it on the router table. At this point it's still at its original height and has its original surface:

The interaction between the paddle transition and the nut location is important, and I wanted to locate the nut and start work on the transition, not least to see what effect it had on the inlay.

So I took the neck to the bandsaw and, while leaving the tabs, cut off all of the overlap of the fingerboard to the neck. I then started work on the paddle transition on the horizontal oscillating belt and spindle sander (I believe the not-entirely-cool kids are calling it a HOBS or HOBSS, so I'm going to join them, as otherwise it's a lot of typing for what's essentially one noun). I got it to a point where I didn't want to go any further without knowing exactly where the nut was going to be, so I stopped here:

Then I used both top-cut and bottom-cut template-profilers in the router table to flush the fingerboard to the neck blank.

Almost starting to look like something.

Same thing we do every day, Pinkie. Try to make headway on the build.

 

[Matt Liebenau]

 

[Quills]

Nut-cracker.

More order-of-operation stuff. As mentioned, I need the top side of the paddle pretty much finished before removing the tabs. To finish the top side of the paddle, I need to need to sort out the fingerboard transition into the paddle camber. To sort out the fingerboard transition, I need to figure out (most of) the details of the nut.

Being uncharacteristically perspicacious (try saying that after ...), I had -- in true hand-tool neander fashion -- located equivalently-distant tiny knife nicks in both sides of the neck before I put the fingerboard on. This was one of the reasons for the knife-cut nut line: to get those nicks accurate (they were at approx. the 12th and 20th fret positions).

Now that the fingerboard is flush to the neck, locating those nicks allowed me to use an engineers' square and transfer them to the top side of the fingerboard. I could then use calipers across those nicks to get the exact width of the neck at those locations. I could then draw freshly-sharpened pencil lines between those nicks (across the neck), set half the distance on the calipers, and use the calipers as accurate compasses to locate the middle of the neck on the pencil line. Doing this in the three locations gave me three dots, to connect, which I duly did and ... voila ... instant accurate centre-line.

Because pencil on ebony is hard to see in the grooves of them-thar fancy protractors, I took white masking tape and spent some time and effort running it exactly down the centre-line. This ought to make locating the fret positions easier and exactly perpendicular to the centre line.

I was relieved to find out that the original nut line matched-up perfectly, however my guesstimate on where the ground/sanded camber of the paddle head-stock would end-up seemed to suggest that the current nut location was very slightly too far back, so I moved it a mighty 1mm (5/128") forward (we have enough wiggle room for that), and re-cut the nut-line from the new centre-line (this also ensures that the frets will be parallel to the nut-line).

I had thought about various ways of cutting the nut, mostly involving router set-ups of various kinds. I don't have a 6mm / 1/4" flush-cut bit, though; I do have a 6mm plunge bit and some template bushings or I do have a router 'trough' jig I could use (which would be hyper-accurate and relatively safe after a hyper-fussy set-up process). But here's the thing: routers work fine until they don't, and I didn't want another miscalculation or oversight to further throw me into arrears. My hand-tool chops, on the other hand? Well -- not to blow one's own trumpet -- but they're not too shabby; I've had a lot of practice -- and you usually have more time to react-to/fix something going wrong. So, in conclusion, I decided to do it the long-but-peaceful way: Chisels. In ebony. Going cross-grain. 'Better put the kettle on.

At first I thought I wanted to fully embed the nut in the fingerboard, so that's how I started ...

... but I soon realized that the sanding drum I was using for the camber was never going to reach more than a few mm up the height of the fingerboard regardless, so micro-housing(dado)-cutting didn't make too much sense as most of the end wall was going to disappear anyway. So I pared down the back wall as I went, until I got to a point where it was about the 'right' height. In truth, it could have been higher, but it worked fine enough.

I didn't get many pictures as I was just chopping, slicing and paring my way down -- pretty pleasant evening, in fairness. This was as I was starting to get near the bottom:

Testing the fit/flushness of the nut (here I'm using a Tusq blank, but I may use brass in the final thing):

After doing some final tweaking with needle files and slim chisels:

I then took it back to the HOBSS (yeah, not-quite-cool kids ...) and carefully finished the transition camber.

Not too crusty.

In other thrilling news, I weighed the neck (before I cut the nut) and it came to 718g. The packet of machine heads (including wrapper) came to 268g and the fret wire (including packaging and excess) came to 65g. So that's a max total of 1.051Kg (or 2lbs 5oz).

Back in post #52, the body came to 2.52Kg (5lbs 8-and-3/4 ounces) which, combined with the neck info comes to a grand total of 3.571Kg (7lbs 14oz) -- without pickup and neck routs, and without the neck carve.

However: even if we take that as 'the' number, it means we can lose nearly a kilo (964g / 2lbs 2oz) to the body hardware and it'll still come in under 10lbs. That's a lot of leeway (I don't think it's that much), so I'm optimistic about bringing it in under 9.5 lbs (4.31Kg), and outside-chance-hopeful-but-realistically-sceptical of bringing it in under 9lbs (4.08Kg). That last one will depend greatly on how much weight comes off in the carves / routs and the total weight of the finish.

But still, not as boat-anchorish as previously feared. Cool.

 

[Quills]

More text than pics today.

First job was to make a depth stop for my barely-used fret-slotting gentleman's saw. I bought this a few years ago to help with refrets, but I've scarcely needed it. A little zona modelling saw and a home-made slot cleaner/scraper jerry-fabbed from a stanley-knife blade, combined with standard x-acto and marking-knife appendages, have done everything I've needed them to do. But ... I'm going to be cutting the fret slots by hand on this build so it's finally going to see action. I'm going to need some practice, though, just to get the feel of it. I'm used to gentleman's saws in general, but not this specific one.

Anyway, some 3mm baltic birch ply, some M5 hardware, a bandsaw, a jack plane, a drill press, some chisels, about 20 minutes' worth of time ... and 'twas done. There are people who say, "you don't need a depth stop, just use a few layers of tape". Maybe, but tape bends, especially over multiple operations (and the people who say that all seem to have over-deep nut slots, so whatevs ...).

(Despite the reflected light making its best effort to confuse the camera, I can assure you that edge is straight. If I didn't mention it; someone would.)

After that, I didn't feel like doing any real work (mess-wise). So I decided to measure and preliminarily mark the fret slots. I'm probably going to do the fingerboard radius after the neck carve (don't quote me on that ...) but, either way, I'm going to do the thing where you cut the slots into a flat fingerboard, then shape the fingerboard, then recut the radiused board using the existing slots. The tabs are coming off soon (after I cut the rough neck taper), so it makes sense to do this now (or, at least, it doesn't not make sense to do it now) as we'll be into radiuses and carves pretty quickly after that.

My original plan was to use the dial calipers to measure the distance of all the fret positions within their range (200mm) and then switch to a ruler for everything outside their range, but I realized that this didn't really make sense as, if you use the same yardstick for everything then, if anything's out, it's at least consistent with itself. Using two measuring devices introduces an extra layer of potential mismatch.

So I carefully clamped a ruler along the centre-line opposite the tape. I lined the 10mm mark up with the nut line; I even broke out an x-acto blade and a 10x jewellers' loupe to make sure the very edge of the middle of the nut line bisected the middle of the 10mm mark. I then added 10mm to every measurement I needed to mark. This was to stop any weirdness near the start of the ruler affecting the distance.

I then carefully set about marking the positions on the tape using a pencil. I'd round the value from the StewMac fret calculator (thank heavens for that thing) to the nearest 0.1mm and try and place the finest line I could on that point. It's basically impossible; you can probably get close to somewhere around 1/8mm by intent (5 thou), anything below that is for machines. But close enough is fine. I sharpened the pencil every three or four marks to keep the line thin. When I was fairly convinced the mark was good, I'd put a little black dot under it with a fineliner. If I made a mistake (I forgot the extra 10mm a couple of times), I'd put an 'x' next to it.

When that was done, I went through again with a different known-good ruler and this time measured the distance from fret-to-fret (the original distances were nut-to-fret). They were all good, except for one part where one mark was a hair sloppy high and the other was a hair sloppy low (which was a hair too much sloppy) so I erased and redid those. I also discovered (with the help of visor goggles) that some of my tape work yesterday wasn't quite as good as I thought it was, so I re-laid that. All this double and triple checking is worth it. It's very easy to screw this up.

When that was done I put a fresh blade in my knife and, very carefully, using one of them-thar fancy protractors, lined-up and cut the fret marks into the board, with increasingly heavy cuts starting from light scoring marks. I've no idea how the hell you would do this without one of them-thar fancy protractors, or something that functions in an equivalent way. The fingerboard is wide enough that a slight angle change will throw things noticeably out. But I went slowly and, eventually, all 22 slots were done.

Incidentally; the stewmac 700mm-scale-length fret position pdf file I have is dated from 9th of May 2022. Shows how long I've been messing with this (and there was a long period before that when it was going to be 685mm scale instead of 700mm).

I call this still-life "Afternoon (wot I did wiv it) (2024)":

Art shot:

 

[Flying B]

To save me trawling back through the thread (it is slightly on the wordy side) were you planning on using the rustoleum spray over a water based dye?

 

[Quills]

To save me trawling back through the thread (it is slightly on the wordy side) were you planning on using the rustoleum spray over a water based dye?

It's one of the options I was / am exploring for sealing the dye in, specifically if you wanted a poly-based finish. Obviously it is (some type of) poly-based finish, so you could just use that itself, but I was looking at maybe using one or two thinnish coats as a sealer to allow using whatever poly you wanted over the top (in the same way you can use nitro and acrylic sealers for those top coats).

Since then, I've discovered that white spirit / mineral spirits doesn't actually move the dye that much, so provided you don't rub too heavily and apply too much at once, you can 'probably' use wipe-on oil-based poly straight without too much damage (the first one or two coats of that will seal itself). I've got a test that I haven't finished which is rustoleum vs nothing against oil-based poly and wipe-on. That's still to be done. I left the rusto curing a few days ago and haven't gotten back to it yet, but I will.

Off-topic but, for anyone else reading: water-based poly (water-based *anything*, in fact) definitely requires a sealer.

 

[Flying B]

Yebbut I was wondering specifically which dye you were using? I'm looking for something to use over the feibings leather dye which is alcohol based.

 

[Quills]

Yebbut I was wondering specifically which dye you were using? I'm looking for something to use over the feibings leather dye which is alcohol based.

I'm using Keda, which comes as powder or liquid. It's an aniline dye and you can use both of them with both water and alcohols (denatured/meths or isopropyl) -- which is why I was exploring sealing options. I'm not sure what Fiebings is -- it might be aniline, it might not.

Big D uses Mohawk Pre-Catalysed sanding sealer with Angelus dyes, which are also alcohol-based and, from what I can gather, similar to Feibings -- but I think that's specifically with one eye on a nitro finish. Theoretically, it should cause a problem as it has spirit-based solvents, but it doesn't because, if you apply the first coats lightly, they flash off before you lift any of the dye.

What you use as a sealer depends as much if not more on what you want to put over it than the solvent that's in it. I know people who use Chestnuts' acrylic sealer over water-based dyes (despite the fact that the sealer is water-based), because, applied thinly, it cures fast enough not to cause problems -- apparently (don't take my word for it; I've never tried it). Obviously, they're doing this with an acrylic-compatible top coat in mind; you probably wouldn't want to use poly over an acrylic sealer (the stearates in the sealer might cause problems).

In your case, you'd 'probably' be OK with an alcohol / spirit based sealer as long as you applied it thin for the first couple of coats. The key thing here is: Test it. There's a lot of contradictory info on the web and figuring out what's true might take some experiments. What was the top finish you were looking at?

 

[Flying B]

What was the top finish you were looking at?

That's the million dollar question! I do have a compressor and a cheap gun but no proper spray booth, so needs to be something water based. If the rustoleum product was good enough that could be used to both seal and top coat.
My old work had a proper spray setup so I used a/c lacquer which was awesome. Sealed and top coated and polished up after 48 hours. Evil stuff though!

 

[Quills]

That's the million dollar question! I do have a compressor and a cheap gun but no proper spray booth, so needs to be something water based. If the rustoleum product was good enough that could be used to both seal and top coat.
My old work had a proper spray setup so I used a/c lacquer which was awesome. Sealed and top coated and polished up after 48 hours. Evil stuff though!

My experiments so far show that, whatever it actually is, the rustoleum seems to work. One of the reasons I haven't updated the thread yet is that there hasn't really been any news. Everything seem to be working -- both water-based and oil-based are sitting over it just fine (so far). I did give it three days to cure before applying anything else, though, so that might be part of it.

Like I say, you've probably got more choice than you think as spirit-based sealers will probably work, applied thinly enough. One of the big things is, whatever you use, spraying the sealer coat. As I've found, the rubbing action does a lot of the work in actually pulling the dye up. If it just sits there, it doesn't move around too much. If you wanted to go with acrylic lacquer, I think I'd just apply rattlecan sealer (something like the chestnuts' if you wanted water) and then either rattlecan lacquer or water-based acrylic over the top. It's very doable and shouldn't tint the finish. One of the downsides of the rustoleum poly is that it is yellowing. It does change the colour of what you put it over. More on that in a future update.

 

[Quills]

@Flying B , One of the reasons to maybe test if you can use a spirit-based sealer, though, is because water will raise the grain more, even if it doesn't raise the dye. Whether it raises the grain significantly will probably depend on how much you pre-raised it (which is over-promoted as a technique, IMO, but can help) and how much you spray on. My apologies; I forgot that part. There's a lot to bear in mind with dye-sealer-finish combos ...

 

[MPU]

I like to use cutting tools for shaping instead of dusting tools. That’s why your use of chisels, planes and Iwasaki rasp would be my choice too. I’d add a scraper to the list of tools. It takes a bit more time but hey, why rush if you don’t need to.

 

[Quills]

I like to use cutting tools for shaping instead of dusting tools. That’s why your use of chisels, planes and Iwasaki rasp would be my choice too. I’d add a scraper to the list of tools. It takes a bit more time but hey, why rush if you don’t need to.

Agreed. Also: sin in haste, repent at leisure. The advantage of the power options is that they take a lot of stuff off very quickly; that's also their disadvantage; you can easily end up with a forced redesign very quickly (as I did on the headstock). I prefer the control and flow of the bladed hand-tools. I will likely be using scrapers a bit more toward the end; a lot of the stuff already-done was the 'roughing' (even though some of it's quite smooth).

 

[Quills]

Quick update. Like: super-quick.

As some of you have probably guessed by my absence, this thread is destined to become one of those slow-grow "WBO2024 that appears for a while after the close-by date" threads. There's no way I'm going to get this finished in time for the deadline.

Unfortunately, real life has gotten in the way. This is partly because work has just hit one of those super-hectic phases that happen every so often, but it's mostly because we have family with an emergency (nothing too serious, fortunately), and a couple of extra, somewhat-small, house-guests staying with us. This is putting a major crimp in my ability to sneak out to the shop -- and a major crimp in my remainder energy levels -- but, on the bright side, I'm winning about 30% of the King of Tokyo games.

I'm also still not entirely happy about the neck, so am humming and hah-ing about whether to rebuild it in maple or real-ish mahogany. Either way, it's not going to get done by next Tues.

Good luck to everyone getting it done. I'll be checking-in sporadically.

 

[Beej]

There might be a Summer Build Off too. Which could be a chance to finish this off under the Assembly category. It could possibly be starting on Thursday June 20, 2024 at 1600h PST or something. You never know for sure.

 

[Quills]

There might be a Summer Build Off too. Which could be a chance to finish this off under the Assembly category. It could possibly be starting on Thursday June 20, 2024 at 1600h PST or something. You never know for sure.

Great point. Ta, Beej.

 

[Quills]

Gzzz ... zzz ... zzzt ... crackle ... fizz ... ZUMMMMMMMM!

Igor! It's alive!

AKA

I'm back and it's still a long post (ain't a darned thing changed ...).

As I mentioned above (and elsewhere), real life got in the way of this build for a while. Everything's fine; I've just been supremely busy, but it's about time for a franchise reboot. I finally got to the point last week where there was enough leeway to start work on this build again but then I realized we were probably coming up on a Summer-Build-Off, so that's what's about to happen; a winter, scratch-build, Not-a-bass challenge is about to become a summer, partial-build, Not-a-bass challenge.

Clearly, I don't need to prep any templates as ... ermm ... already done ... but that doesn't mean we can't get on with things in the meantime.

Numero uno.

Firstly: Marginally sad news, but I've decided that the neck I was working on above needs to die. There are a couple of reasons for this. One is that the headstock readjustment owing to the router mishap bothered me; not enough to end the project, but it was a 'gripe'.

The second was to do with the wood. As you'll be aware if you've read the above (I don't blame anyone for not doing so; most of this is for the benefit of my future memory), the choice of blank for the neck had to be changed mid-build owing to a concealed grain orientation contributing to a twist issue in the original Utile blank. At the time, I did it expeditiously as I was still trying to meet the WBO deadline and my initial decision felt like the right thing to do to stay on pace, especially as I believed that there was a suitable replacement in my wood stash.

However, very shortly after working with the new blank something started to bother me, which was that the wood felt unexpectedly light, and certainly a degree lighter than the other projects I'd made from the same set of boards of the same species. I set my concerns to one side and got on with cutting the nut and fret slots, but there were a couple of places where little pieces of ebony (from the nut cut) had gotten under the board while I was chisel-chopping and had created little indents in the neck blank (mostly my own fault). They were nothing that wouldn't have steamed-out, but they were sufficiently large and noticeable that it again reinforced the notion that the wood was lighter and less tough than I'd previously thought.

Some playing around with the other boards revealed the answer. The catch is that the other projects in which I'd used the same wood had come from a different board in the set, and it appears that natural variation resulted in that one specific board just happening to be atypically denser than all the others.

I'd previously thought that the wood was Bossé, precisely because of its density being too high for Spanish Cedar (and there was a period in western Europe in the 70s and early-80s where a lot of Bossé was sold as Spanish Cedar), but now I believe that the boards very likely are Spanish Cedar and the one which I interpreted as Bossé was just abnormally dense. I've since made a spice rack from one of the other boards and it's similarly Spanish-Cedar-ish to the neck.

People do make acoustic guitar necks out of Spanish cedar but, even then, some players are 'iffy' about it and I'm not particularly happy about using it for a baritone where it may well see extra-thick strings used in Drop C at some point -- and with it having a much longer neck than an acoustic to start with.

TLDR: I don't like the replacement neck blank so that has to go. At some point I'll split it open and recover the truss but, in the short term it's not necessary as I have another one and a different neck blank. The only major losses are time and a decent piece of ebony, but I can live with that. Rebuilding the neck will therefore be part of the SBO.

Numero dos.

With that out of the way, I'm going to pick-up (pun slightly intended) on posts #68, #72, #73 and #76. In those posts, I started testing the viability of using Rustoleum (EU)'s Polyurethane Finish (henceforth RPF (no affiliation)) as both a sealer or final finish on guitars.

I made two test boards: one that had water-based 'poly' and oil-based polyurethane (Rustins) as a base and gradually sprayed RPF over the top; and another that used the RPF as a base and had the polys gradually applied over the top. This was to see how compatible they were and how long was needed for the under-layers to cure.

I can report that, as of this morning, all the combinations seem stable and compatible with each other, despite me rubbing at them with an eraser, some sandpaper and a wet sponge to see if anything would give. Provided you leave a few days for the polys to cure up, and give them a hint of tooth with some light sanding before applying the top coat, they all function well.

Numero tres.

This left us (in post #73) with me preparing a mostly-fully-dyed board to see how the RPF would work as a sealer and to what extent it would tint the dyed surface. As a recap, this was the board after preparation:

Since the last update, I then sprayed half the board with RPF, to see how it looked. It came out like this (RPF on the left, bare on the right):

... and here it is in different lighting conditions (in my micro-shop):

As you can see (and as we established in the earlier posts), the RPF has a distinct yellowing effect (which you'd expect from solvent-based / oil-based poly). It's not unpleasant in its own right and you could certainly use it either as-is or intentionally for the effect and it would still look decent. You could even deliberately bias the substrate very slightly to the pink side so that the addition of the yellow would pull it back to the 'middle', but that would need careful experimentation for each colour you wanted to try it with to get the balance perfect. You'd also have to bear in mind that the yellowing/ambering will probably increase over time.

So, with that result in mind, I decided to test out Chestnuts Acrylic Sanding Sealer (Aerosol) (henceforth CASS for the sake of politeness (also no affiliation)). This is a product that Chestnuts sell as a base for their acrylic lacquers, although by many accounts it can be used under a lot of other acrylic-based topcoats as well. It's water-based and should therefore not yellow as much, but would probably mean that you'd have to use some type of acrylic finish over the top.

Chestnuts want you to use their own finishes (obviously), but I've become rather taken with the matte version of Rustins Quick-dry Varnish (henceforth RQV (also no affiliation)), which is an 'acrylic-based emulsion' that refreshingly makes no pretense at being polyurethane. It comes out a bit thicker than the milky nature of typical water-based 'poly' but levels-out very well and leaves a surprisingly smooth finish. So: part of these investigations are to see whether you can use CASS+RQV as a water-based sealer-topcoat combination over aniline dye, given that we already know you can use RPF either alone or with whichever tested poly you like (albeit with the additional oil/solvent-based yellowing effects).

So I took the blue testpiece above and sprayed CASS on the other half of the surface, leaving a gap in between them for reference. Here it is again, with two medium coats of RPF on the left (as before) and three light-medium coats of CASS on the right:

And here it is again in the natural workshop light:

As you can see, it very slightly yellows the surface -- nowhere near as much as the RPF -- but it's more that the extra translucency, sheen and chatoyance increase the contrast of the piece.

So that got me curious: Does the same degree-of-effect show up with Red? So I prepped another board (ahem ... the other side of the same board). I didn't go too carefully with this as I wasn't trying to create a 'wow' surface, just get enough colour in to perform the test. Here is the dyed board before spraying:

And here's the dyed board after spraying (again, RPF on the left, CASS on the right):

Truth be told, I slightly screwed-up on this one as in my haste I forgot a crucial step, which is gently sanding back the top layer with p600-ish paper or 0000 steel wool after the final dye application but before the finish. This does multiple important things, like getting rid of the excess dye on the surface, and denibbing any lightly raised grain; it also slightly thins the finish to increase the contrast. You can see on the left that the RPF hasn't had uniform application as the slightly raised grain has differentially absorbed the finish.

However, that doesn't change the colour result much, which is that the RPF still yellows more than the CASS (the image doesn't show this well; it's clearer in real life), but that it's less important with red. In fact, if you watch Big D's youtube channel at all, applying a yellow wash over red finish is often used to make it 'pop' (as it slightly orange-izes the red highlights). It would have been better to see the effect with a proper sand-back, but it still gave me the information I needed.

This led to another question. We know from our previous experiment in post #68 and #76 that, with RPF as the sealer, you only need a couple of 'heavy-ish mist coats' to effectively stop water-based varnishes from raising the dye: So how many spray coats of CASS would you need to do the same job?

I knocked up another test board -- this time with straight black and red (cyan) -- with different sections that had increasing numbers of coats of CASS. Instead of using heavy-ish mist coats, I used light mist coats this time, as I wanted a bit more granularity in the results. Here's the test board before the test:

You can see the slight difference in tone between zero coats and one coat but, after that, the gradation is barely discernible. The other thing here is that the board was sanded to p120 (with a ROS) but the grain raise (from the Chestnuts sealer) was relatively minimal. It did increase with each additional layer. By layers 5 and 6 it was more noticeably raised, but was still nothing that you couldn't knock down with a light brush of p600. If the surface was already-prepped to p600 (instead of p120), as it would be after a final-dye-knock-back, I doubt it would raise the grain much at all, especially if the surface had already been grain-raised and sanded-back before dyeing.

The protocol was similar but not quite the same as the last one. I took fairly-damp kitchen towels and rubbed the top-end of the red and black lines 32 times each in each section using medium-heavy pressure each time (as a musician it's easier to count regular bars). I then took some of the same water-based poly as last time and blobbed it on the lower-end of the coloured bars in each section and left it to dry, just to see to what extent the standing water-based poly would pull up the dye.

The results were as follows:

You can see discernible rub effects on blocks 0-2, slightly on 3, and very slightly on 4. By the time you get to 5 any visible effect seems negligible. You can see finish leaching on blocks 0-3, and 3 is somewhat light; after that, I can't really see any.

You can get a different / better view of the rub effects by looking at an angle:

Again, the rubbing and leaching seems to have mostly gone by 4 or 5.

This is borne out by the tale of the rags. Like before, these are in pairs: one red and one black per layer, 0 coats on the left, 6 coats on the right:

You can see that some residue is still being pulled up on layer 4 and very mildly on layer 5, but this had very little effect on the surface appearance of the dye.

Bearing in mind that: (a) there was no final-dye-knock-back before the sealer (which would have taken most of the surplus off the surface); (b) 32 medium-hard rubs is probably two orders of magnitude harder than anything a normal finishing process is going to do; and (c) the mist coats were lighter on this test than the RPF test, I think it's a fairly good result: 4 light mist coats would do it; 5 if you want to be sure. In practice, fewer heavier coats might be more effective.

I think that a reasonable protocol would be:

• One light mist coat as a fixer;
• Another light mist coat to back up the first;
• A medium coat to consolidate the first two;
• Optionally another medium coat to make it pretty dependable.

That would put you up in the 5-6 ballpark. You could then do a light knockback with p600 and add whatever topcoat you needed (I think ....... ). However: This needs to be tested to be sure. I'm already pretty confident of what RPF will do; it'll work fine with most poly topcoats and is less prone to dye raising as it's oil/solvent-based; the unresolved questions are with the water-based CASS method.

So the next job is to basically make a full small-scale workpiece: Dark underlayer; sand back; at least one lighter top layer (maybe more) with appropriate sand-backs afterwards; then the prototype CASS protocol above; then several thinnish layers of the RQV acrylic finish to see what we get.

I'll get back to you when I've found out.

 

[Quills]

The experiments are finishing

Or

8 infinitesimally different shots of the same piece of wood

Following directly on from the above, I made a test piece to assess the viability of using Chestnuts Acrylic Sanding Sealer (CASS) in combination with Rustins Quick-Dry Varnish (RQV) as a water-based, non-yellowing, home-brewable alternative to poly sealer and poly topcoat. For scale reference, the piece is pretty much dead on 203mm / 8 inches long.

Unfortunately, I didn't have too many highly-figured test pieces left -- they're either spoken-for for future projects or I'd need to sand back one of the others; so I grabbed a bit that was flamed, but not heavily or overtly so. The test is a good demo of the extent to which you can exaggerate/pop even relatively light grain, but it's also a demo of the fact that there is an upper limit. Better figure makes for better results.

I don't show the dyeing process here as you've seen it before and this one wasn't super complicated but, for the details-curious, it was: one dark layer (0.8g keda dye powder into 120ml water*, 7:1 black-blue mix); after drying, sand back with p320 sandpaper to expose the figure; then one layer of lighter-blended, slightly-desaturated blue (0.6g into 120ml*, 2:1 black-blue); after drying, flatten back lightly with 0000 steel wool.

I was going for a desaturated used-denim-y kind of look and it worked. Here was the test piece after flatting back the dye:

I then hit it with the sanding sealer as deduced in the previous test. Experience is worth more than test results sometimes as, in spraying it onto figured wood, I realized that you didn't have to go that light; the solvent flashes off very quickly. I still gave it the protocol in the above post, but the first couple of layers weren't that mild; they were a medium-heavy mist. followed by two other medium layers.

After the final layer, I lightly flatted-back again with p600 in one gentle pass, just to take the high spots off. The sealer is quite soft and thin (as evidenced in the previous tests), so you don't want anything too coarse or aggressive or you'll go through it.

Post sanding sealer:

I left it about 42 hours and then coated it with 1 light-ish and 3 medium-heavy coats of RQV (matt), applying it with a brush, leaving two hours in between coats as per the PDS. After the second and third coats, I hit it very lightly with p600 and cleaned the debris off before applying the subsequent coat. It came out like this:

It looked great and felt very smooth -- the RQV self-levels reasonably well -- however, the brutal scrutiny of a direct back-light revealed some predictable issues. I know people recommend foam brushes for water-based acrylic / 'poly' and they're right; it would probably save some flatting work at the end as traces of the brush stroke remained in the final layer (possibly from the previous ones, too). There were also some tiny bubble remnants and dust. Backlit shot:

It's actually not as bad as it looks in the picture (except for a couple of spots, you can only just barely feel it) and is sufficiently scant that some of these low-angle light shots took quite a bit of effort with a phone camera to target the intended detail. Unless you get down really low, it'd be good enough for many home-brew projects, but not if you're trying to get nearly-passable as factory.

The other issue is that, as you get lower, it's not really that matte, either. It's certainly not glossy, and the varnish was well-mixed, but the typical guitar matte finish is done with abrasives and finish, not just finish, so onwards we go.

I waited the PDS-recommended 3 days to let the acrylic 'fully harden' then took it out to the shop. I worked it with p600 to get most of the hills out, then p800 as the valleys got shallow, and finally worked it in circular motions with 0000 steel wool. It came out pretty great -- and definitely more like a 'matte' finish.

Outside, above:

Outside, backlit / low-angle:

Inside, above:

Inside, low angle:

You can still see some of the scratch patterns from the steel wool on this final shot, but it's a lot more consistent and, to the touch, it feels perfectly smooth. From all other angles it looks pretty-much perfect, but raking back-light reveals all. Some of those scratches are artifacts from not thinking and doing a straight up-and-down motion at the end, when I should probably have kept it to circular patterns, bit it still works.

In all truth, I'd be happy with it as-is for a matte-finish, but I'm going to experiment with higher grits and scotchbrite to try and find the right matte balance.

TBC.

(* These weren't the actual amounts, just the relative ratios. Using that much dye would be a massive waste on a work-piece this small.)