ESP M-II type "Invaders" Superstrat

[Prostheta]

Well, that's about as close as it can get. 

So in principle, yes. I could reduce the existing blank's neck angle to 13 degrees but it would require either a great setup in the CNC to nail it first time, or several test passes to ensure we're on the mark. Equally, the option of reducing headstock thickness to 14mm eases the demands on this process. I don't know where all my Gotoh tuners go to (pun intended) and I can't find a minimum headstock thickness requirement, but based on that of Hipshot open gear tuners I wouldn't be surprised if they could support 12mm. So that way with a 14mm thick headstock I can still add in some 1mm bushing recesses, themselves cut into cast-in-place rings 3-4mm deep. Cool.

[Prostheta]

The working procedures for the neck and the body must absolutely be defined before kicking off. Let's look at the neck.

 

This is how I threw together the blank during the final weeks of access to machinery at my last workplace, with no specific project in mind. I decided upon a 14deg neck angle (no idea why now) with the taper and everything else just falling as it did. The two holes either side of the scarf were locating screws that both cinched the scarf up, stopped it rotating in place but most importantly stopped it displacing under clamping pressure. A cutoff from the blank (remember the long photo earlier?) was glued to the edge to provide extra meat for the headstock tip. It's the same Maple, but doesn't look it. The centreline is exactly 40,25mm from the front reference edge.

Tomorrow I hope to mount this onto the CNC and do one thing only; the truss rod channel and headstock access. Immediately after this, the fingerboard blank gets glued on oversize. I've calculated that there is enough material in the waste to contain a full channel around the fingerboard. It's horrendously wasteful, and in a way I want to cut off that waste and replace it with offcuts from the scrap bin. I can use Ebony for all sorts of detail work in my furniture projects. This may or may not happen.

Ideally I would do only the truss rod work since the neck will be removed from the CNC. As repeatable as this CNC is, I do prefer to minimise errors between machining cycles. In this case however, I need to face off the headstock so that the transition point from neck face to headstock face is a known value. There will always be a little movement between cycles for various reasons. In an ideal world, this neck would be machined using an auxiliary workholding bed that references on the CNC with pinpoint placement. Doing that for one neck is excessive, and a calm approach with a good eye is all this takes to spot and counter tolerance errors. That means I need to do this on a good day....

This is the sort of data that I export from Rhino as a DXF file:

This might look odd for a few reasons that aren't obvious to CNC machinists, and even then there's a few things that need doing purely based on the CNC programming software I use. The small angled red line at the lower left is a 13,5deg reference line. I'll use this to create a new working plane in WoodWOP in additon to the usual XYZ/top/front/left, etc. planes. This is the headstock plane. The headstock outline and various details are all drawn with reference to the centreline and the origin. Once the headstock plane is established, these will be dragged over to that plane. Similarly, the neck and fingerboard cross sections (green) for radiusing exist on the XY plane, and will be moved to the XZ plane for extrusion to a CAM milling surface. The truss rod and access and detailed in light blue, and the fingerboard inlays in orange. Fret lines are also in place, which will be projected onto the surface used for milling the fingerboard radius. These fall short of the edge, so we're looking at radiused blind fret slots. Cool.

Once the fingerboard is glued, the binding channels for the fingerboard and headstock can be milled in. I imagine that I'll machine off most of the waste from the headstock also. The tuner holes will be given their own 16mm diameter recesses as rings with a 6mm endmill, leaving a small 4mm central point for manual tuner bushing recessing. I'm a little paranoid about this part, since machining this out feels very subject to tolerance later. I suppose that this is just as delicate as the headstock perimeter binding, and definitely an argument for auxiliary workholding to increase precision and repeatability.

The headstock binding channel will be 2mm in width from the outline and 3mm in depth, which allows me to paint over the face and scrape back 1mm to reveal the binding. Again, this will be cut with a 6mm endmill. The headstock logos have been extended outwards as simpler pockets for excavation. These will be filled at the same time as everything else, and finally covered with a black/gold decals to mask the excess glow.

As far as pouring the binding goes, this will be a two stage process based on the differing levels of the neck and the headstock. Firstly, I'll do the fingerboard. Once that's cured enough to no longer be subject to movement under gravity, the headstock pours can be done. My head's a bit empty right now, so I think that I need to physically look at the workpiece once I've milled the channels so that any damming and taping can be figured out. It might just be as simple as a channel stopped just beyond the nut shelf, however the headstock binding comes up just under that.....hmmmm.....

[Prostheta]

Last night, I remembered just how easily I get drawn into new ideas and projects....which was exactly the sort of thing I wanted to avoid. My "last build of 2021" was intended as a both focus and a therapeutic idea for my wandering ASD head. Blegh. I lost and I suck. The PE project is a big ticket project, and even though this one won't exactly be cheap (Gotoh trem, tuners and EMGs mostly) it's certainly a lot cheaper than the Maple top, genuine Mahogany and all that of a LP-type guitar, not to mention the pickups, paint and hardware. Still, that one must absolutely happen.

Second to last build of 2021. Is that okay? 

[Prostheta]

Onto the CNC! Headstock faced and transition point referenced. Truss rod cavity, slight adjuster relief and access routed.

[Prostheta]

The headstock was supported from underneath using a piece of Pine on one of the beams. It was enough.

I need to block sand the face of the headstock to remove that facing artifact. It wasn't supported during that operation, so a quick slap on the wrist and let's move on shall we?

 

The tuner holes aren't yet drilled to depth but will "reveal" once I thickness the headstock later. The locations around the tuners were given a centric path cut 4mm deep with a 6mm endmill, so leaving a 16mm diameter hole to be filled. Same as the binding, the exact perimeter of these can be modified by masking.

The headstock perimeter was rough cut with a 20mm endmill and a 10mm offset. The binding channel a -2mm offset (into the headstock) with the 6mm endmill. This is wide enough to clean by hand and to pre-wet by painting with resin instead of just pouring.

Yes, feel the pain of having to dig out a broken 0,5mm endmill from the middle of the headstock. The contour threw an exception (a "contour loop", which is a move smaller than the CNC wants to bother with) and I had to manually jog the spindle out. Oops, which way UP is Z again? Oh, not that one clearly....

[Prostheta]

Truss rod in place, fingerboard glued on. Nothing much to see other than the usual forest of clamps. To prevent the fingerboard sliding under hydrostatic pressure (glue acts as a lubricant under pressure) I tapped in a few tacks around the perimeter of the fingerboard blank.

Ugh, so tomorrow I have to go through the repetitive process of calibrating a new 0,5mm cutter. Unfortunately there's no automatic process for doing this that I can see with Homag machines.

[Bizman62]

3 hours ago, Prostheta said:

that facing artifact

This one? I thought it was your signature! Just waiting to be filled with some glow-in-the-dark stuff.

[Prostheta]

The pigments came in yesterday. A few initial tests to establish ratios and whether sedimentation is a problem immediately showed that I might need to rethink this choice of epoxy. Based on the size of the pigment grain and the ultra-thinness of the casting epoxy, I immediately saw the glow pigment crashing out of solution to the bottom of the pockets. That isn't terrible for the fingerboard inlays, as these benefit from a clear surface layer that is cut back during radius sanding. Cutting into epoxy with pigment in it without adding a protective clearcoat or grain-free layer over the top leaves broken and exposed grains that just deterioriate and look like ass.

The main issue is with the body binding. If glow pigment crashes to the bottom of a 3mm deep inlay pocket, imagine what it would do within a 18-19mm deep binding channel! The choices here seem to be increasing the epoxy viscosity (and gel time) such as West System 105 and/or altering the grain size. I'd have to consult more experience than I have on this one, however I am thinking that either using extra large grain size as a fill and then encapsulating that with an ultra-thin resin might be an option, although the chances of dry pockets and bubbles are enormous. The other would be ultra-fine glow pigment, which tends to have far poorer glow capacity than the larger grain equivalents. The binding channels are pretty large in volume, making the amount of glow pigment needed to perform any sort of fill pretty high. Reducing the waste would help, and certainly reduce the quantity of epoxy needed at any one time since West System can get pretty unhappy if its starts approaching thermal runaway.

As it stands, inlays would look pretty cool with the grains naturally crashing to the bottom of the pocket. The tiny individual pixels are a bear to fill without bubbles, especially since it's so hard to tell whether you've succeeded until later in the process. Black unfilled spots just aren't going to cut it here.

Oh, and the other thing that's driving me mad. I seem to have summoned some sort of offset between the XY plane and headstock plane geometry. The headstock is a few mm off the Y axis and for some reason so is the fingerboard. Not a minor disaster by any means, and I'm sort of wondering how deep my commitment can go with this project....recovering the neck is a possibility but I'll have to junk the fingerboard. First things first, identify the nature of the mistake in the first instance. No doubt it is rooted in my lack of patience and organisation....

[Prostheta]

Subsequent testing.

The largest grain size - also the most powerful (XL-ZZ1150) - has a large grain size of around 220-280um. Pretty much like fine sugar. This stuff absolutely crashes out of suspension in epoxy. especially when the epoxy is heated to assist in bubble elimination. In addition, it is feasible that it can introduce more bubbles to the epoxy during mixing. It must be tried of course. I have a small amount of Z-Poxy 30 left, which for the most part is a reasonably close analog to how West System 105/205 would work.

My first step of the process is to measure out the correct amount of resin into a graduated measuring cup, mix in the appropriate pigment and only then add the hardener. This would perhaps be easier with West System 105/205 since that is a 5:1 ratio, allowing for easier estimation of pigment by volume. After mixing in the XL pigment around 3:1 resin/pigment, I added the hardener leaving this somewhere around 15% pigment to epoxy. This is very difficult to accurately measure with small quantities during tests. The final mix felt a little grainy with too much separation of the coarse pigment and the epoxy. I added about 10% of the M-ZZ620 used in previous tests as a "filler".

Testing this in a flat black-painted 3mm pocket in Maple, the coarser pigment crashed out the epoxy when heated as you'd expect and settled at the bottom of the pocket. The medium-grained pigment stayed mostly in suspension until the epoxy started to catalyse and gel up. I'm not certain if this would represent success in a larger volume casting, however it does seem to imply that thicker epoxies are the way to go. Thinner/finer might yield a similar result with fewer bubbles, however the differentiation between pigment and epoxy by casting height sounds like it would be pretty problematic.

I don't particularly want the binding to have the clear look that the thin casting epoxy creates. A more homogeneous look would be preferable, even if that seems a lot harder to achieve at this stage. The only other idea that seems to be feasible given the processes I'm using is another hybrid approach; casting the binding using XL grains in the first pass, re-cutting a finer channel at a layer immediately adjacent to the finish cut and casting a second chemically-bonded layer with a different pigment content. Essentially a core binding of coarse high-power grains (that might not even charge at that depth) with a finishing surface layer of weaker finer grains.

I'm open to options on this one, guys. The best guidance I can find before I get input from Tom is from hobby casters who simply state that finer grain size (10-15um)

This is an exceptionally-useful table of info on the products that chaostrade.eu sell, and you can see exactly why I chose to buy pigments from them....most companies don't offer technical differences between pigments to help choose the right product.

https://chaostrade.eu/Comparison-table-of-luminescent-pigments

So according to the available green pigments, that target of 10-15um grain size seems a terrible compromise between glow power (this is absolutely needed!) and the grain size that apparently doesn't crash out of suspension.

Perhaps a good idea for a test would be raising the ratio of pigment to epoxy and seeing how that affects sedimentation; would this increase the shear forces within the epoxy enough that it started exhibiting more thixotropic properties? That is, resists movement of the grains under gravity enough through binding within the mixture? I'm certain that rising bubbles will disrupt this and cause movement and separation between the two, but hell, that's why we test, right?

[Prostheta]

Interesting point of note. The higher fraction of glow pigments to epoxy in this test make the unlit inlays far more visible during daylight. The initial casting epoxy test (low fraction content) is close to invisible by comparison. Revisiting this test with the same approach as the second test using Z-Poxy 30 seems worthwhile, even if the higher viscosity might make filling the finest details problematic.

Sedimentation and separation doesn't appear to have been a problem at all with Z-Poxy 30. The remnants in the measuring cup (always keep these to test for how that batch is chooching) have consistent appearance through the depth (around 5-7mm) and hold a consistent glow charge. This is exactly what we want from the binding. This leaves the decision on whether to buy more Z-Poxy or go with West System, which is marginally less expensive by volume. I don't think Z-Poxy comes in larger packs than 8oz that I can find.

[Prostheta]

This is how the inlay test runs look. There are faults, but acceptable at a push. I want to perfect them.

Headstock areas that'll be masked using decals:

 

12th-15th-17th-19th fret inlays

 

Full board (hard to capture):

 

The big problem with inlaying such detail into Ebony is its tendency to chip out easily, which definitely showed itself in the invaders' eyes. The B-grade boards I picked up might be even more prone to that, who knows? I'm wondering whether CA might stabilise Ebony enough to make the milling work, that is, mill close to the islands within the pockets, stabilise with CA and then mill fully. The other option would be to remill the inlays later on, however I'm having some doubts about whether the workholding strategy (or lack of one) I am using has caused problems with alignment. It really shouldn't as I normally work to 0,05mm during my working-work with the CNC and calibrate cutters to 0,01mm. Losing 2mm of accuracy over less than a metre is highly unusual, and I suspect that I'm looking in the wrong place. Checking the CNC program this afternoon showed that all of the geometry aligns correctly, without any planes being offset with respect to a common centreline. I absolutely need to know why this went wrong, otherwise I'll quietly go batshit!

[Prostheta]

Hmm. After a beer and sauna, I'm sure that I can recover this neck as-is. The fingerboard will have to go, and the replacement will need kiss milling to ensure that the locating is on the mark. The headstock is 1mm shifted to one side, so I'm sure I can recover this without it being detrimental to the finished neck.

I still need to try and identify the source of that discrepancy. 

[Andyjr1515]

Interesting stuff, @Prostheta  

[Prostheta]

4 hours ago, Andyjr1515 said:

Interesting stuff, @Prostheta  

 

Thanks Andy. I wanted to try and do a guitar as much by the numbers as possible without significant investment, and look where I ended up. Again....

[Prostheta]

I think I should cut my losses, or be wary of overextending myself with this build. The glow binding is nice in principle, but will soak up a lot of money in materials. I'll pick up some green ABS binding and reduce the glowing stuff to inlays.

[Bizman62]

Awww... And just when i was going to ask if the consistency of the binding could be achieved by casting several layers as they advised back in my childhood with a product called "valumuovi" - "casting plastic" especially if you wanted to have objects inside the cast to be at various heights.

[Prostheta]

There's two reasons for the change in tack. The first being that the cost of both the glow pigment and the epoxy will significantly increase the overall project budget. This was intended as an easy throwaway idea with scope for experimenting with a few fun features. The second is that it also require significant experimentation/testing that might not necessarily reduce overall risk in the workpiece itself. That also pushes up cost as "pioneer ideas" always suck up all money in the vicinity. To put it into rough numbers, a kilo pack of West System 105/205 epoxy is about €60-70 on its own, Z-Poxy 30 would need 2-3 packs at about €25 each. The glow powder seems to benefit from being in a higher viscosity epoxy like these in a high-ish concentration of about 4:1 which increases the difficulty in eliminating bubbles or settling into finer details.

If the budget supported this approach and the proof testing required, the binding channels could feasibly be poured in layers, yes. Essentially, mixing the glow powder into a measured batch of the resin and portioning that off into individual batches before adding the appropriate amount of hardener to each one subsequently. The size of each batch is a variable that can't easily be guessed at without full-scale testing and/or commitment. I'd imagine that minor sedimentation would make the different pours have visible boundaries of strong/weak pigmenting which might look pretty cool similar to faux ivory binding, however that isn't the objective. The ideal would be consistent flat-coloured binding that is both greenish in daylight and a strong glow in the dark. The logo inlay pocket tests confirmed this can work in wider shallow pockets, but perhaps less so in fine mm scale details.

The casting epoxy is fantastic at flowing into fine details but sediments out so much (and quite likely does so even with the finest of glow pigments) that it doesn't carry a flat colour of its own. The only solution that I could think of here would be to coat the internal faces of the inlay pockets with a colour for daylight visibility, and accept that the powder crashing out of suspension is a given. The inlays look great when lit, but the powder settles so much that the inlay is difficult to discern otherwise.

 

10mm pocket drilled into scrap African Blackwood, casting epoxy with as much pigment as I think it can take before becoming chalky and shear-resistant.

 

T+1 - Poured and before heating to eliminate bubbles.

 

T+3 - heated and letting bubbles rise out.

[Prostheta]

Over the next few hours, this will sediment out depending on pigment ratio and pocket depth. The more places the pigment can settle under gravity, the more of a clear(er) layer of casting resin is left on the surface. I think this is an achievable way of doing the fingerboard and headstock, however I am edging towards the idea of painting out the pockets with a green paint prior to adding this pigment/epoxy mix in. Not sure if that will look as good in practice though.

[Prostheta]

T+14 - fine bubbles are still emerging, pigment is noticeably settling in the warmed epoxy with a clearer layer being visible on the surface.

 

Tests allow me to work outside of the recommended ratios of epoxy to pigment to see how this affects things. This is maybe 50% pigment to epoxy, much more than the 30% in normal guidance. It's also worth noting that these tests are being done with an aqua pigment (M-JNW450) to save the green pigment.

[Prostheta]

T+4hrs and both pockets are into their gelling time, working through to an initial cure. 1-2 days for basic curing and a week for extended glass-hard curing.

The lower ratio pour is showing shadowing around the periphery plus a couple of open bubbles on the surface. I didn't heat this like I would a "real" pour, as normally you'd prick these within 30-60 minutes so they level out. The pour with the higher fraction of glow pigment has a larger amount of fine bubbles trapped at the surface and very little shadowing from sedimentation. Realistically I think hitting something between these two would produce the brightest lit inlay and most visible unlit, however it's risky. An inlay that doesn't get a clearcoat later on (as in the case of a fingerboard) won't have minor imperfections such as cut surface pigment grains or bubbles removed. I'd accept the result on the right, with a little more concentration from the left. Interesting how the test on the left didn't sediment out as much. I am thinking that at some point the sedimentation produces a dense lower layer that supports the material on top from crashing out further. I couldn't cut an inlay such as that on the left as the exposed cut grains would look like ass.

[Prostheta]

Alright, well this one needed a bit of a reset I think. So I've got two QS rock Maple blanks on the way from Espen, one of which I'll cut and laminate with Bubinga for this build, the other I'm keeping to one side to recreate my old Mirage (separate project that this is sort of the precursor to). Those'll be here in a week or so.

In the meantime I need to do some work to fill up a page of decals from Rothko & Frost. I'll also be sourcing green binding for the body, neck and headstock from there plus maybe some binding for the other Mirage and the PE.

This also gives me the opportunity to reset the specific parameters of the build such as the headstock angle that was "baked" into the blank that I had laying around. I'd still very much like to use that blank, and I may still do that in spite of having new Maple on the way. It really depends on whether I can recover the 2mm oversize binding channel offset around the headstock and make that into the 1,5mm required by the binding. It's a lot of work to recover, and I'm unsure of success. By resetting, I regain control over this.

[Prostheta]

So here's the proof of the tests. The top one is the lower percentage of glow pigment, the lower was something like 50%. Note the pinholes in the surface that ended up capturing some of the dust from filing/sanding back. It seems like 30% is the mark to hit.

(edit for clarification, or at least bringing the test results together)

The epoxy to pigment ratio is a band that seems to top out around 30%, and the band itself is dependent on how deeply the epoxy is being cast. The more epoxy and less pigment, the higher the likelihood that the pigment will crash out and the deeper it can sink. The lower strata of the epoxy will end up with a far higher ratio of pigment than the original mix and would resemble the lower sample if you cut that far back into the epoxy. Higher ratios reduce the amount of pigment crashing, but introduce new problems; pinhole bubbles not migrating through and out of the casting adequately due to higher viscosity or shear tension, and surface pigment causing a poor finish when cut back.

The poor surface finish of exposed cut pigment grains is likely less of an issue if the final piece is being clearcoated. This would effectively "resubmerge" the surface grains and makes the finished item look as though there is the bare minimum of epoxy over the glow pigment. If anything, this would produce the strongest and most visible glow, plus the easiest to charge. The pinholes are far more problematic in that sanding dust or other fine debris can be impossible to remove. I haven't got compressed air to try and blast the stuff out same as most, so it has to be taken that this is maybe not achievable. If it were, and a clearcoat could fill those up, that would be the high standard of glow power and maybe through-consistency.

Eliminating bubbles seems to be the absolute decider here for clearcoated items. For exposed castings, the right ratio of pigment to epoxy with normal heat bubble elimination. In principle this would mean that I could try the original binding and detail ideas that require deeper castings using more viscous epoxy and higher fractions of glow pigment, however there would be a very significant increase in trapped bubbles that would require pressure to crush. Vacuum elimination would disrupt any homogeneity within the epoxy/pigment mix, increasing crashing, pushing epoxy out of the cast with bubbles or at the very least reducing consistency within the casting. It would also require a not-insignificant amount of epoxy, pigment and equipment to achieve, if it could be achieved at all.

[Prostheta]

I've been playing with the heel somewhat. After seeing several photos of high-end modern bolt-on ESPs with this sort of heel arrangement, I figured it would be a nice aspect to add in. The heel is cut back with a flat angle that terminates within the upper cutaway with a radius. Since I prefer ferrules over neck plates, these can place themselves anywhere within the angle or even across the transition. In addition to this slope, a more typical corner cutaway has been added. The geometry of this appeals to my eye as generally I'm not a bit fan of easing within cutaways themselves. I could modify the heel shape however this is the basis for a future repro of my original Mirage, so I'm aiming not to get too far off that path.

[Prostheta]

Hopefully this illustrates the geometry of the forearm and rear comfort contours. They're all based on sections of a cylinder, which makes executing them much simpler as CNC operations.

[Prostheta]

Comfort contour: