'58 Explorer - 1st build

[Prostheta]

In principle, symmetry is what you're aiming for. Unless wood has internal tensions, the intention (joke intended) is for movement caused by moisture changes to be equal across the profile. This can be broken down into two things.

Firstly, the grain - or at least the lines down the length between early and latewood - should be parallel to both the centreline over the top and fingerboard glueing face along the side. If they veer to one side, that's normally a sign that the grain isn't straight enough or the cut wasn't aligned ideally. You can get away with a small amount of runout like this, however any runout is inviting twist or excess uncontrollable movement.

Secondly is the growth ring alignment across the profile. This is where the wood will move the most - radially and tangentially - and distort the profile through that movement. If you have runout in the grain as explained previously, this symmetry will change unevenly down the length so that distortion (which will happen) also changes, causing twist or other effects.

Sapele is a tough one to divine in this 3D manner. The tree doesn't grow as large and as straight as most, and there is often a lot of tension in the wood because of this. I don't like using Sapele that I haven't personally cut and prepared from lumber, as you don't get a read on might be hiding in the wood. Cutting it shows whether the wood is like a box of tense clocksprings or relaxed and chill.

More often than not, Sapele is the tweaker of the guitar wood world. In my experience, the heavier, denser stuff comes with free surprises however YMMV here. For the money, Khaya (Ivorensis or Senegalensis) is friendlier. I prefer the tone of Sapele in a neck though, especially when paired with Wengé. It just needs a higher level of selectivity to get the results.

[Prostheta]

On 11/11/2021 at 6:49 PM, asgeirogm said:

I then cut the heel to the correctish length and routed the neck angle into the heel by putting a 10mm hex key at the right length from the heel end to give me a 2 degree angle

 

 

It would be negligent of me to not point out how excellent these two steps are. A 3D-printed routing template for the corners is novel and excellent if the availability of one is there. Also, using Allen keys as to-size shims for trig-based plane routing is fantastic. Much safer than using drill bits that like to slide around the place! A couple of drops of CA or hot snot help if they still like to try and move under routing vibration.

Tips for the Makita: don't use the spindle locking button, because I don't think the casting is strong enough to take the strain when locking up a bit. Buy a 13mm and 22mm spanner/wrench specifically, and grind down the thickness of the 13mm spanner so that it fits into the spindle slots. They're safer and produce a tighter lock. The Makita only has a single-locking collet cone which applies pressure along the router bit shaft from one point as opposed to double locking collets that pinch from two locations. Vibration is higher in single locking collets because of higher possibility of cutter runout, so you don't want a cutter running itself loose!

[asgeirogm]

16 minutes ago, Prostheta said:

It would be negligent of me to not point out how excellent these two steps are. A 3D-printed routing template for the corners is novel and excellent if the availability of one is there. Also, using Allen keys as to-size shims for trig-based plane routing is fantastic. Much safer than using drill bits that like to slide around the place!

I would probably say that my favorite thing in the world is to solve problems in creative and efficient ways, so I'm very happy that you think this was good as I hadn't seen other people do these things before (even though many definitely have before me, I just haven't seen it), it just made sense to me.

[asgeirogm]

Thanks for the thorough response on wood movement @Prostheta, I really appreciate it.

36 minutes ago, Prostheta said:

Sapele is a tough one to divine in this 3D manner. The tree doesn't grow as large and as straight as most, and there is often a lot of tension in the wood because of this. I don't like using Sapele that I haven't personally cut and prepared from lumber, as you don't get a read on might be hiding in the wood. Cutting it shows whether the wood is like a box of tense clocksprings or relaxed and chill.

Seeing as I've come this far on the neck before realizing I wasn't as selective as I should have been, I'm keen to explore magnitude of the risks I'm facing here so I can make an informed decision on whether to chuck this neck and start over or continue with it. I've done some cutting on it myself to get to this point, namely thicknessing and tapering, and I've gotten a slight twist on it, so I know there is (or was) some tension in the wood.

What I'm really wondering about now is this: Let's say I finish doing a couple more rounds of faceting, let it sit for a few days, straighten everything up from the twist and then finish the carve up and let it sit for a while (days, weeks?) and it doesn't move any more. How much of a risk am I taking that it might twist more down the road (weeks, months, years)? Is it unlikely, likely, impossible to say, etc? For humidity background, I live in Denmark.

[asgeirogm]

39 minutes ago, Prostheta said:

Tips for the Makita: don't use the spindle locking button, because I don't think the casting is strong enough to take the strain when locking up a bit. Buy a 13mm and 22mm spanner/wrench specifically, and grind down the thickness of the 13mm spanner so that it fits into the spindle slots. They're safer and produce a tighter lock. The Makita only has a single-locking collet cone which applies pressure along the router bit shaft from one point as opposed to double locking collets that pinch from two locations. Vibration is higher in single locking collets because of higher possibility of cutter runout, so you don't want a cutter running itself loose!

I bought the router as a part of a set and it had these kind of spanners, and I always use them to tighten, so I'm good, but I appreciate the tip regardless

[Prostheta]

Cool, I don't think mine did. Using a 22mm spanner on the end nut and the 13mm on the spindle is much better than stressing the router's casting with that red spindle lock button. Mine started to wear before I shifted to two spanners, and I've read of people cracking the router housing more than once. I'd prefer for that not to happen! Funny that you seem to have gotten yours with an offset base. I didn't know that those were even available for that model, which is why I made my own. Did that come as part of the kit?

[Bizman62]

4 hours ago, asgeirogm said:

I have a question regarding attempting to prevent twist in a less-than-perfectly-straight neck blank.

You don't have the #4 option which is bookmatching i.e. you split like above and turn the pieces sideways - actually that gives four alternatives as shown below:

[Prostheta]

Remember the three shrinkage directions.

Tangential > Radial > Logitudinal (TRL)

• Tangential is along the rings or "around the tree".
• Radial is "outside into the tree centre"
• Longitudinal is "top to bottom of the tree"

In terms of looking at arranging wood for your profile, bear in mind that woods shrink and expand more tangentially than radially. Not a magnitude, but several times more. Longitudinal is usually a couple more magnitudes than either of these to the point of being negligible. This would make the uncut piece of wood in the above illustration convex on the top surface if it dries; the longer length of the growth rings can shrink more by percentage, pulling the edges in and back. The opposite is true if it becomes more moist.

This is why I think quartersawn and wide-ringed flatsawn are great to work with. Movement does not distort the profile, it changes it more or less proportionally. When arranging pieces, I try and keep shorter sections of growth rings where possible, and place wood nearer to the centre of the tree as being removed by preference as they have smaller radii growth ring structure. If I were to choose from one of those four, I would choose B or D. For B you are removing more smaller "more distortion prone" radii rings. With D you are reducing the length of the longest source of radial movement. Beyond that it becomes an aesthetic consideration on whether you want the show face to present with growth rings mostly perpendicular to the finished surface or more parallel. One will have more flake structure (Maple for example) whilst the other will be more of a "cathedral grain". Figured woods absolutely benefit from this consideration.

Another argument for QS vs. FS is that movement is along axes rather than being a weird 2D distortion. A neck can become thinner (QS) or narrower (FS) which do not cause problems with geometry as much. There are a few schools of thought on this, but generally the bottom line has always been "good clear straight grain".

[Prostheta]

Useful reference:

https://www.wood-database.com/wood-articles/dimensional-shrinkage/

....whilst this helps illustrate how orienting growth rings affects cosmetic presentation of structure on the show faces...

 

https://sherwoodlumber.com/moisture-content-in-lumber/

[killemall8]

3 minutes ago, Prostheta said:

Useful reference:

https://www.wood-database.com/wood-articles/dimensional-shrinkage/

....whilst this helps illustrate how orienting growth rings affects cosmetic presentation of structure on the show faces...

 

https://sherwoodlumber.com/moisture-content-in-lumber/

Ive never seen this explained before. Very useful!

[Prostheta]

20 minutes ago, killemall8 said:

Ive never seen this explained before. Very useful!

I learnt a lot of this hands-on from my degree work. We regularly took in trucks of wet wood into the steam kiln, plus did practical experiments of weighing and quantifying wood by volume, drying it in an oven to pretty much zero moisture and examining the changes in geometry, dimension, volume and weight. The came in especially useful when making slab-top furniture, because too high a turnaround from taking in wood at transport moisture (typ. 12-15%) to putting it into production leads to big trouble in the product. Equally, having large amounts of wood on stock drying is dead money and dead space if it isn't turning around quickly enough. I turned that into a motherfunkin' artfully tuned machine. Knowing how the wood works and what is optimal translated to having enough reliably-dry wood to satisfy demands of throughput whilst not having shitloads choking up all the space and gathering dust. Funny, because that was with the least professional, most dumbass employer; just myself and him. I guess you have the ability to reconfigure methods quickly at that size, and to his credit he took on the important points and made them happen. Subsequent bigger "professional" companies mostly order wood in at transport moisture and put it straight into production in the order of a couple of weeks thanks to terrible misguided implementation of zero-stock JIT manufacturing. Putting together a white Oak desk and having to cover it in plastic with a tub of water isn't a solution....it's offsetting the problem.

Wood will do what it will. It's a relentless creeping doom that you need to work with, not against

 

[asgeirogm]

3 hours ago, Prostheta said:

Funny that you seem to have gotten yours with an offset base. I didn't know that those were even available for that model, which is why I made my own. Did that come as part of the kit?

I 3d printed it I actually need to make a V2 some day, the base is not quite stiff enough so I need to add some braces on top to prevent flexing

[asgeirogm]

17 hours ago, Bizman62 said:

You don't have the #4 option which is bookmatching i.e. you split like above and turn the pieces sideways

Well the neck blank I have is much closer to being quartersawn than flatsawn, so I don't think this option applies for this particular blank. the problem is more that the grain just isn't as straight through out the length of the board as one would like.

@Prostheta Thanks for all this knowledge, this is really useful info!  

16 hours ago, Prostheta said:

I turned that into a motherfunkin' artfully tuned machine

Nice, sounds like you have a knack for supply chain management  I enjoyed that anecdote a lot btw.

[asgeirogm]

I feel compelled to ask again since I didn't get a response, and I get that since I guess someone might be hesitant to answer since if you say "it's not likely that it will twist later" and then I complete the build and then it twists, I would hold you responsible. Well, that certainly won't be the case, I just don't know enough about how wood moves over long periods of time so I would love to learn more so I can try to make the best decision here.

So, I really would love it if someone could share their feeling on how likely it could be that the neck would twist somewhere down the road after the build is completed, given that it stopped moving during building after being given ample time to stabilize. I just have no feeling if it's likely since it twisted a bit before, or if it's unlikely since I live Denmark and the wood will have moved how it wants to move already, or if it's impossible to say since it depends on the grain in that particular piece of wood, etc.

Any input would be most welcome, and also taken with a grain of salt I just want to learn.

[Prostheta]

Sapele has a higher tendency to move, both because of the nature of the wood not being as straight as most and also by it having a higher incidence of internal tension. The internal tensions will release as you work the material, a lack of straightness will show itself as odd movement such as twist or other warping. Unfortunately, there is no good way of predicting this other than knowing how likely it might be with the materials you choose to work with. Sapele can remain dead as an arrow straight (a good arrow, anyway) and clear Maple can twist. The less you "know" the material from source to finished product, the less chance you get of having a read on what it might do.

Example; I was ripping a 52mm board of Sapele a couple of years back on the table saw, and the material was parting like a Y which pushed the cut material from against the fence into the side of the blade a fair bit. The piece even made big deep cracking noises that sounded like booming sea ice in the wood. "Reaction wood". Sapele seems to have a lot of this.

A little bit like this, but imagine booms as thick sea (not lake) ice flexes and fractures.

 

I was out here on the frozen sea at Santalahdensalmi just off the coast of Laitakari.
https://goo.gl/maps/qxAWaFc26GCLS7MWA

[asgeirogm]

@Prostheta I guess the subtext of what you said is that it might well twist/warp down the line, it might not.

I think I will do some more facets over the next few days and see how I feel about my chances before either chucking the thing in the garbage or finishing it. One thing that is putting a mark in the "don't chuck it" column is that I'm not sure I have a really good replacement candidate ready, so I might have to source a new neck blank from somewhere, and I'm really trying to see if I can make use with what I already have. Let's see...

[asgeirogm]

Continued faceting

Couldn't see any more twisting/warping so I just finished the carve and the headstock transition. I will then do tweaks and more sanding once the fretboard is on and all that (I can already feel that I would like the shoulder a little less pronounced, at least on the bass side)

I then decided to just flatten the twisted guitar top, so I clamped down my wonderful steel beam with 120 grit DS taped to it and just sanded until everything was nice and flat

After some sanding it was easy to see the corner that had lifted away from the previous plane

Done!

Here are a couple of pictures that show how the grain is not completely straight

I've decided that unless the neck suddenly turns into a propeller on me, I will keep the neck. To be honest, after carving it out and loving the feeling of holding that in my hand, I've become emotionally attached to it, so I really don't want to chuck it in the bin If it twists on me later (and I can't fix it with a level and a recrowning), then at least it will be pretty on the wall.

One thing I forgot to mention before is that I made a mistake when sanding the neck pocket, I clearly didn't sand it as evenly as I though I was, so I made it too loose further into the body, so when I realized that I hadn't sanded it enough further front in the pocket, so I did that and discovered to keep the neck snugly in the pocket, I needed to move the neck forward in the pocket by 3 millimeters. Annoying, but I could live with moving the neck pickup and the bridge forward by 3 millimeters. This was a few days ago, I just put the neck into the pocket again and clearly either the neck or the pocket (or both) have been moving a bit so now to have the neck snugly in the pocket, I need to move it forward roughly 6 millimeters, and I think that is too much, so I will have to create a couple of shims and glue them in the pocket on each side and route that again and sand until the neck fits snugly in the correct place. I will on doing that until I'm sure the neck isn't going to move any more. I will also need to reroute the 2 degree neck angle into the neck as after sanding the front flat, there is a 0,5mm difference on the end of the heel from one side to the other. I will also sand the front of the headstock a little bit to move the neck break down a bit.

Edited November 17, 2021 by asgeirogm

[Prostheta]

The main thing is that you build. Every project you work through will improve everything you do after that point, so in most cases it's irrelevant whether the project is completed fully or not. The difficult part is deciding whether you want to be on an ongoing journey-of-many-builds or simply want to complete one or two, and that's it.

All looking good though. Keep moving!

[Charlie H 72]

Nice save on the twisted neck. Somebody was posting here a few weeks ago that Gibson used to glue all of their necks as you have mocked up there - with a gap at the heel end for a tight fit. That 6mm might be more than Gibson used, but it could still come out in the wash. However - neck pocket shims aren't too difficult to make and can be practically invisible if you match up the grain right. 

Good idea to re-rout that neck angle though - 1/2mm on the heel could cause a significant difference at the nut. 

And like Prostheta said - just keep building! The work is really clean and you are dealing w/ the unexpected very methodically. I think this is going to be a good looking and good playing guitar. 

[Bizman62]

4 hours ago, asgeirogm said:

I need to move it forward roughly 6 millimeters, and I think that is too much

That's only half a fret at the dusty end. Shimming the sides of the neck pocket is a solid option as it's side grain but putting a block between two end grain pieces may not work that well as a glue joint in end grain isn't too good. If you feel like the neck is going too deep into the body, you can make the body 6 mm shorter - just slab your routing template on the body 6 mm towards the bridge and reroute the shoulders. No one would ever know!

[Bizman62]

28 minutes ago, Charlie H 72 said:

Somebody was posting here a few weeks ago that Gibson used to glue all of their necks as you have mocked up there - with a gap at the heel end for a tight fit.

Note that Gibson uses a narrow tenon instead of a full width like here. A gap at the heel in that case helps getting the seam tight at the heel. The heel may also be carved a bit concave for an even tighter fit. Many acoustics are built the same way.

[Charlie H 72]

Ooh the concave heel! Cool! I love "sprung" joints like that.  It's worth googling "Gibson tenon gap" and see what's out there - Here's one. I think its an epiphone but same idea. Though this guitar has more tenon north of the neck pickup

 I

Edited November 17, 2021 by Charlie H 72

[Prostheta]

Easing the shoulders also allows for "flossing" of the glueing joint with sandpaper for a razor-tight join.

I can't speak to Gibson's reasoning for the gap at the back of the mortise, however many builders that use tapered tenons into a matching mortise fit the neck at the back and use a clamp configured as a spreader to push the neck into place and tighter into the taper. Aria Pro II/Matsumoku went one further with this and drove a couple of screws through pre-drilled holes in the mortise once positioned, which applies additional clamping pressure. A nice touch.

[asgeirogm]

56 minutes ago, Bizman62 said:

Shimming the sides of the neck pocket is a solid option as it's side grain

To me that's by far my best option as it allows me to put the heel again all the way back into the pocket. 

45 minutes ago, Bizman62 said:

A gap at the heel in that case helps getting the seam tight at the heel

I had seen many LP builds and I was wondering why all of them had a gap, it had to be intentional. Now I know, thanks for the info

[asgeirogm]

34 minutes ago, Prostheta said:

Easing the shoulders also allows for "flossing" of the glueing joint with sandpaper for a razor-tight join.

Could you elaborate on that?