ESP M-II type "Invaders" Superstrat

[Prostheta]

So looking at how the body looks made up in Rhino leaves a few blank spaces that I think need tying into the overall theme.

Firstly, the pickup rings. Two options present themselves here. One would be to take silicone moulds of existing rings and pour some glowing resin rings. The other is to pour a block (or rough blank) and mill the rest out. This also provides the option of making rings somewhat more custom. I'm unsure how at this stage, but I'm sure that inspiration will strike.

Thoughts? Just about to start working on the inlays....

[Bizman62]

Just now, Prostheta said:

Firstly, the pickup rings. Two options present themselves here.

Three... You should be able to make them in a similar way to the binding. With your CNC skills carving the moulds to a scrap piece should be a no-brainer as you already should have the digits for the angles on the top...

[mistermikev]

alien archeologist lizards - cracked me up there lol

afa rings... i like the idea of pouring them in the sense that I think this could be a useful process to know.  I also think that cutting them might be problematic given the material.  could combine the two ideas and mill something out of scrap and then make a mold from it.

[Prostheta]

Buying moulding silicone is expensive, or at least an expense that I'd rather not have to extend to. I've ordered a larger pack of casting resin than I think I need, so I hope the glow powder and pigments will run to doing the whole batch with tests. 

[Prostheta]

I'm thinking that inlays like this would work nicely.

[Bizman62]

Wohoo! Pure arcade!

BTW have you played the Excel spreadsheet version of Space Invaders or a couple of other classics?

[Prostheta]

17 hours ago, Bizman62 said:

Three... You should be able to make them in a similar way to the binding. With your CNC skills carving the moulds to a scrap piece should be a no-brainer as you already should have the digits for the angles on the top...

 

I missed this post....

That¨s pretty much what I thought. There are options. I've no idea how "grabby" this resin is, which means it might break or jump out if cutting slightly too aggressively.

[Prostheta]

This is definitely working for me.

@Drak - a Gotoh tremolo would be perfect for this, especially since the price is pretty damn good. A lot of people say they're better than FRs as well....any input on that?

[Prostheta]

Lights on!

[Prostheta]

It seems that there's something a bit fucky going on with the internal curve of the lower horn. Nothing that can't be fixed of course, but worth mentioning as a "spotted, will fix" point.

[Prostheta]

A logo and name are mandatory. Just not entirely sure how I'd go about doing this in glowing material unless I screen printed it.

 

 

Time to get building, methinks. I'll discuss how I will approach various specifics on the CNC later.

[Prostheta]

It's definitely a good stepping stone towards my old Mirage. I wonder if the new owner might ever come by this thread and get freaked by how much a previous owner missed his/her new/old axe?

 

[Drak]

The Sonic Crayon isn't up and running yet, which the Gotoh trem is appointed to, so no, no comments on the Gotoh in particular as yet.

[Prostheta]

Thanks Drak - I'd appreciate your input when you've got any opinions on that unit. I'm presuming that it's a straight OFR swap, at least in terms of the routing?

[Prostheta]

Okay, a bit on how I'm going to approach building this. The thread isn't intended to be a detailed thread on how-to with CNC, but a lot of the processes will be outlined so the concepts aren't a simple "so I mashed the keyboard and pressed the button, tada, secret guitar appears, so there" but at least convey the basics. I'm lucky in that I can work on this project using a very capable 5-axis industrial CNC via my workplace, and generally challenging myself with complex pieces (more often, relatively simple processes that end up needing complex solutions) always increases my ability through hands-on experience. CNC is very much not a "push button, fire and forget" thing, and needs a whole other batch of skills in order to make a project succeed. I'm still learning, and hopefully always will be. When I stop learning, I stop being good and will likely be bored to hell....

Primarily, there are two approaches to CNC. One is where the processes are geometrically intentional, and the other is "dumb go/no-go" where the CNC simply cuts most of the material that isn't meant to be there and mostly leaves only the stuff that is. A bit like how copy carvers reduce unwanted material rather than eliminating everything. This is often the main difference between 3-axis 2,5D CNC machining and simultaneous 5-axis, as many things with the latter simply aren't possible with the former. For example, a forearm contour based on the section of a cylinder can only be approximated with 3-axis motion but cut by exact geometric pathing with simultaneous 5-axis (I know, 4) motion to reveal a smooth cylindrical finished surface.

My design work within Rhino is done in way that preserves all sketches and fundamental geometry used to produce the 3D model. It's this data that I use to develop a working CNC design, not the 3D model. Often my data is still 2D, with the 3D part being the remit of the CNC programming software. That bit is simply the "bonus proof" that helps me gauge a number of things, such as the feel of proportions, visual weighting and even shadowing between planes, etc.

An example would be that forearm contour. A 3D object imported straight into CNC software loses a lot of valuable context. Choosing the set of surfaces representing the forearm contour from a baked 3D model and asking the CAM processor to cut those approximates the original intention and specificity. It'll likely still be very very accurate and needing maybe a bit of 240 grit sanding to smooth out, but is still only representative that being geometrically-true. Being able to specify a process that cuts a cylindrical section based on a vector and radius is vastly better for a long list of reasons, not simply "accuracy".

The exception to this would be complex shapes such as carved tops that do not easily strip back to basic 2D geometry. This minefield of worm cans can be discussed another time!

Often I see a lot of threads that involve CNC work, and people request "the G-code" or even a 3D model. At the very least, the latter can usually still be used to derive basic geometry to generate a good working set of processes, and hence specific G-code. G-code is generally the end product to drive that specific machine, those specific cutters and into that specific workpiece. Outside of one person's very defined set of parameters, G-code quickly becomes meaningless. This might not be the case with smaller CNCs and simpler processes, however it's best practice to generate your own solution from the fundamentals rather than seeking a pre-micronuked roadside café meal. The results are always better if you cook from fresh ingredients. Unless you suck at cooking, in which case you need to learn a bit more cooking and get with the game man 

I know this is borderline preachy stuff, but from hereonin the actual CNC stuff that is very specific to my circumstances becomes less meaningful than the fundamental working ideas that underpin it. Make a fire and a man will be warm for the night, set him on fire and he'll be warm for the rest of his life, etc.

I've been a bit mentally drained the last two weeks, so I have no idea how that ^^^^ all reads to you guys. It makes sense to me, but hey.

[Drak]

4 hours ago, Prostheta said:

I'm presuming that it's a straight OFR swap, at least in terms of the routing?

It is not, actually. The Gotoh is ever-so-much slightly larger all the way around.

I had already cut my cavity and I'm reasonably certain I'm safe, but just barely.

If i knew it going in I would enlarge all routes by just a hair.

I think the back route is the same, its the top routes that come into question.

I did extend the top/rear cavity (where the fine tuners end) to the dead-maximum I would normally go.

There is a review on YT that covers the differences, I think, and the slightly larger size is mentioned (Gotoh vs. OFR)

You can't tell just by looking at it, the difference is so small, but <just enough> that I would give a touch more space to the routes.

[Prostheta]

Hmm. I think it's worth getting a unit in an quantifying that first. Thanks for the heads up! I like tight, but close to the limit when working blind is a little pucker-inducing.

[Prostheta]

A little on the workholding strategy as I've been rolling it around my mind. Primarily the CNC I work with uses vacuum consoles to hold workpiece, all of which have varying sizes and ability to retain a workpiece. The smallest consoles are intended to hold in precise areas rather than small workpieces, so I am thinking that an auxiliary bed might be necessary, and in a lot of ways easier. A board made from heavier material such as 19-25mm (3/4" - 1") MDF works well for this even though the material loses vacuum through its porous nature....especially when cut into.

The CNC uses hydraulic referencing pillars that raise during workpiece mounting, generally intended for the alignment of sheet goods. Something a bit larger than a guitar body still works very well, and has excellent repeatability. That is, taking the workpiece off and replacing it between processes doesn't have problems with placement being different each time.

I haven't entirely formalised the exact working processes yet, so partially me writing this down is doing so. I have already planned that the body will have three processes carried out on it; the binding channel as a top-down process, a second to rout the outer profile, forearm contour, pickup routs, control locations, neck pocket and tremolo recesses. Finally, the body will be flipped and held in the auxiliary workholding board for the rear routing. The body blank without any through-hole routing is just about large enough for a couple of large vacuum consoles to retain it securely, however that's not entirely ideal for the heavier operations plus reducing the footprint with the full perimeter cut is likely not possible.

The perimeter cut can be done to a depth that is sufficient that the partially-machined blank can be mounted face down into an MDF sheet with a matching mirrored recess of known depth. Since this is aligned and cut by the CNC already, repeatability should mean that cutting the rest of the perimeter depth from the rear should match up perfectly with the front. The workpiece needs some form of holding within this MDF which I haven't yet fully considered. Perhaps a bolt through the trem rout or holddown clamps either side of the body that are removed and replaced between two cycles of perimeter routing.

[Prostheta]

The auxiliary workholding bed can itself be cut with the CNC which aids with placement. The internal pocketing operation to produce the mounting recess has a perimeter that absolutely matches the body outline. This may need to be eased by say, 0,1mm or a little more as the cut edges won't be perfectly smooth, creating a slightly obstructive fit. This is something that just needs to be fixed in the mix, so the pocketing operation for the bed will involve maybe a couple of back-and-forth adjustments to the offset of the cutter making the pocket.

[Prostheta]

I'm going to take my first steps into silk-screening a logo in order to lay down a glow pigment based headstock design. I imagine that this will be a simple rectangular frame that is mounting using the tuner holes, with something like a 230 mesh count. I haven't calculated the pigment grain size versus the mesh count yet or decided on the medium. This is a side quest, but definitely one I'd like to try out. Again, another string in the bow.

[Prostheta]

After a little reflection, I've decided against using silk-screening on the headstock. Primarily the reason is down to consistency; the printed logo needs several applications including a white basecoat, and won't match the power of the binding's glow. So, the only other viable solution is to mask a cast logo using a decal. Essentially, excavating the majority of the logo footprint and casting that the same as the binding, flattening and applying a waterslide decal mask in between layers of paint and clear.

Silk screening is a nice technique and great for laying down inks and oils, however glowing pigments benefit from weight of application that I don't think can be reliably done without losing detail and/or chancing misalignments between applications.

[Drak]

Just thought I'd give you some measurements since it seemed to be concerning to you. You can see the differences are small, but they're there none the less, depending on your router template allowances. I don't think Gotoh would knowingly manufacture a part that would not fit its biggest competitors' pocket by 1 to 2 mm, that would seem silly and non-profitable.

And I don't know what different manufacturers' locking trem pocket tolerances are, I just know mine, which is a 2-part template set (3 including the rear), from StewMac from ~25 years ago...maybe other manufacturers allow more tolerance in their pockets...could be.

Baseplate width OFR - 90.5mm / Baseplate width Gotoh - 91.5mm (1mm difference)

Baseplate length OFR - 37mm / Baseplate length Gotoh - 38.5mm (1.5mm difference)

Full front to back of entire block assembly, not including fine tuners, but measured to the back of the fine tuner shelf,

The front of the baseplate nose to the back of fine tuner shelf OFR - 58mm  / Gotoh - 61mm (3mm difference)

In the YT comparison, they believed the Gotoh string lock screws are slightly longer. Which would add to the overall final length and their biggest complaint, IIRC. Which would mean (possibly) that you couldn't precisely intonate the low E (possibly the A too). As if you moved the saddle back to the proper intonation point, the back of the string lock screw would protrude beyond the recessed shelf pocket (if using one, not everyone does). So I removed the low E string lock screw from each unit and measured that too.

They were correct... OFR screw - 44mm / Gotoh - 48mm (4mm difference)

I checked for thread compatibility between both units, and they are a match.

So if the Gotoh screw wound up overextending the recessed shelf when the saddle was set at the proper intonation point, you could just buy an OFR screw to save a few mm's. Or adjust your initial placement of the baseplate location forward by a few mm's. Or not have your low E perfectly intonated but close, with everything else intonating properly.

So overall length, baseplate nose to back end of screw, is 8.5mm longer on the Gotoh, that could wind up creating a problem if you didn't make the recessed shelf long enough, which I did extend mine by a little bit.

 

[Prostheta]

I really appreciate that Drak, and what you detail lines up exactly with the information I am able to find online. I don't think I'll be going to the lengths of replacing the string locking screws (no pun intended) as extending the recess is hardly a huge modification to a new build. Unless it looks like ass visually (doubtful) I think it's going with the specs from the Gotoh unit ass-is.

Those units are damn cheap as well. €140 or so from Thomann, however I don't think that comes with the locking nut and they don't seem to stock that.

[Prostheta]

The design as drawn may need very marginal alteration as the neck blank I am using is already shallower than the 21mm I set out. This makes the neck pocket shallower and the heel thicker. I'll compensate for this by introducing a relief plane onto the back of the heel so it slants towards the front and the treble side. Instead of a neck plate, this one will be getting black ferrules, black oxide bolts and maybe a glowing surround around the ferrule holes.

[Prostheta]

So, the first step with the neck is to get good referencing since there will be multiple operations done over 2-3 cycles. This neck blank had been made so that the central Maple laminate tapers from end to end, meaning the outer reference edge isn't parallel with the centreline. This never helps. I marked the centreline and placed the neck over three vacuum consoles and referenced the edge against the pillar stops. Measuring the distance from the leftmost pillar to the centreline and then adjusting the neck at the rightmost pillar to match should put it parallel to the X axis. A small piece of scrap wood was cut into a long triangular wedge, and I used this between the rightmost pillar and the neck blank to edge it away from the pillar until the centreline measured the same at both pillars. The vacuum was then engaged, the workpiece held in place, the pillars dropped and we're good to go. A simple 600mm trimming operation along the X axis right to left (a climb cut) produced a clean reference edge. There's still a lot of meat left on the neck both in width and length, but I need that for holding the neck with the vacuum consoles. I also cleaned the glueing surface where the headstock scarfing meets the neck with a simple pocketing operation. I'm tempted to take off maybe 0,2mm across the entire face of the neck blank now we're referenced flat, and then do the same with the headstock to produce a surface colinear in the Y axis. There's still a good 18mm or so of material thickness in the headstock and I much prefer to take that off the front than the back where possible so the scarf joint is closer to the headstock, not in the middle of the neck. The scarf measures 13,9deg at the moment, so I'm going to do a little calculating to see whether that 18mm supports me edging that back towards 13deg. If not, I'll surface the headstock at the natural angle of the scarf and use a bit of the additional thickness to recess the tuners. Maybe with glowing perimeter rings....