Class 6: Progessing the body

 Last Wednesday's class (12/15) had me working on all parts of the body: sides, back and top.

Sides

The sides that I glued up with the end blocks last week were ready to prepare for gluing on the top and back.

The first step was to sand the back side on a circular sanding dish (seen to the right) with the same 15 foot radius curve used throughout the back.  This will prepare the edges of sides, and
especially the blocks at the neck and heel, to hold the same curve that the back is being set to.







Following the sanding, I started to glue in the kerfing that will run along the inside edges of sides along both the top and back.  The flexible kerfing provides a surface to which the top and back can be glued, extending the glue surface past the 3-4 millimeter thickness of the sides themselves.
This is a fairly simple operation, but one, as seen in these pictures, that make extensive
use of clamps.

End piece

Edward gave a demonstration during the class of fitting an tail piece in the body.  The shape of this piece is angular (long trapezoid) so that the piece of contrasting wood can be fit tight into the cutout made in the body.

This cutout is right over the heel block; the tail piece is decorative but also used ot hide a less than perfect fit of the sides.  The side joint at the neck end will be covered by the neck itself.  The notch is cut using a pull saw, then cleaned out
with a chisel.  I'll be using a piece of Koa to match the bindings, and will probably add black purfling strips for contrast.

Carving back braces

I worked on carving the braces on the back of my guitar.  Carving the braces reduces their mass while trying to preserve their strengths, hence the use of a curve.  The braces are tappered down to 5 mm at the edges where the ends of the brace will be fit into the sides.


Carving is done with a chisel, in successive strokes until the top of the curve is set about 60 mm from the edge of the top.  The trick to get a gentle cuve that preserves mass towards the center of the  back.  My braces tend to be carved a little on the steep side.

Top

I worked further in this class on gluing in the braces on the top.  In the picture, the transverse brace is being glued on.  This brace has a 1/4" hole drilled through the center to accomodate the end of the neck truss rod.  The hole is centered 4 mm above (below) the top, so that the outside edge of the hole is just inside the brace.

I also cut and glued the bridge plate from a scrap pice of walnut.  This is glued directly under the bridge to accommodate the ball ends of the steel strings which pull back up against the top.  The plate protects the soft spruce.  The piece is cut slightly off the direction of the grain so that the ends do not pull against a single fault line in the grain.

Back fitting

Edward provided a final demonstration of how to fit the back to the sides.  I'll be working on this when I return from our break.
The back is centered over the sides and the location of the braces is marked with a pencil on the outside of the sides.  It's important to remember the cuts for the braces run straight  across, not perpendicular to the outside curve of the side.










Next a back saw is used to make very shallow cuts for the edges of the notches.  Better to cut to shallow than too deep.

















The notches are then cut out with a small chisel.  Each joint can be tested by placing a small light inside the guitar and placing the top over the sides to see if any light comes through between the back and the side.










I'm hoping to come in once over our winter break to get everything else finished up on the sides and top so that I can begin fitting my back once class starts up again.

Class 5: Muti-tasking

Half of our class was missing this evening, which made it a little easier to spread out and get a number of tasks done.  Tonight I finished out the soundhole rosette, did the layout for the bracing on my top and back, cut the X-brace for the top, and glued up the back strip and braces on the back.

Finishing the Rosette

Completing the rosette at this point was down to just bringing the strips of paper and veneer added the week before down to the level of the sound board.  The strips were riding about 1/16th to 1/8th of an inch higher than the sound board, so using a sharp chisel, I worked them down to be closer to flush.  From there, sanding with a small block at 80 and then 120 grit brought the trim level.  A final sanding, prior to later final sanding for tap tuning and finishing, was down with some 220 and a block.  The complete rosette, I'm happy to report, no longer looks like an accident.  I'll try not to go down this course next time.

Gluing up the Sides
Here's a couple of shots,  from the bottom and the top, of the sides being glued up with the end blocks.  Spring clamps were added to the heel block to bring the sides up tight.  Cam clamps were used, three for each block, to bring the whole a assembly together.
The entire assembly is now sitting in the mold in the shop until next week.

Laying out the Top
The heart of the guitar, so to speak, is really the sound board.  This is the component that provides the core vibration and response that give the instrument the bulk of its character.  The top must be flexible, but strong enough to stand up to the longitudinal forces created by the pull and compression of the steel strings.  The layout of the bracing underneath the sound board, therefore, is perhaps the single defining design.

The design used by Edward on his steel string guitars is built around a classic 'X-brace' pattern developed by Martin (and possibly others) in the 1930s or 40s.  The 'X' is the two long braces running from the lower to the upper boughts, crossing and joined underneath the sound hole.  This pattern has been transferred from a template onto my guitar.  The transverse brace runs across the opposite side of the sound hole, and numerous finger braces spread out from the primary X.

After laying out this pattern, I did several test cuts of the X-brace joint before finally cutting this joint in two pre-cut
braces.  Top braces do not have a radius along their length, even though the top will be glued up on a support that will help to provide a curve with a 30' radius.  This curve is gentle enough it does not require the curvature to be set into the  braces before gluing.



Setting Back Braces
The back curvature, however, is more defined than the top.  We are using curvature with a radius of 15 feet.  In order to give some additional support to the back, a center strip of think spruce is glued down the middle of the back.  All of the supports for the back are glued on using a support that already has this curvature.  This is one of the many unique pieces available working in a shop that new luthiers must develop.  There is a caul with the same 15' curve that will be clamped over the back strip in order to provide support during the gluing.

First the back strip is put in place with some help
from masking tape. Masking tape is a great help to lutherie, but one of only a few tools that are cheap and readily available.   In addition to the back from and caul, we are using prebuilt templates and molds for the sides along with additional jigs (such as the router circle cutting jig that got me in to so much trouble) to help move the process ahead.

Another essential element of lutherie is the use of clamps.  Shown here and below are the standard wooden cam clamps that are used throughout the process of gluing braces and many other parts of the guitar.  They can easily run $15 to $20 each, and as you can see in the pictures below, we can easily use 12 or more at a time. 

After the back strip has dried, it needs spaces
set into it for the braces.  These braces, as mentioned, already have a curve set
to their bottom and have been given a preliminary shape on top as well.  Four are glued in to the back to give strength.  Setting clamps for gluing is its own art forms could all be set at the same time.






 Below I found space for 17 clamps on the back so that the braces could all be set at the same time.

 

Another subtle skill, as in all woodwork, is getting just the right amount of glue to ensure a firm joint, but not so much that things get messy and look sloppy.  An example of the latter is shown in the close up below.

Next week I will be fitting and gluing on the top braces as well as putting in kerfing strips along the sides to create a glue surface for the back and later the top.  There is a bunch of work ahead in carving are braces and tuning up the tops to complete the assembly of the guitar body.

Class 4: Finishing soundhole and fitting sides

Rosette Recovery

Last night (12/2) I worked on trying to bring some resolution to the problem with the sound hole.  Here's the rosette after gluing and having the entire top put through the thickness sander (top thickness is a little over 3 mm now).  The inlaid wood looks nice, but inside the rosette the enlarged sound hole is still a bit awkward.  I sanded off the uneven strip of remaining spruce using a rotating sanding block on the drill press.  At this point Edward suggested that I could just leave the sound hole as is.  The larger hole may give an appropriate expansion to the volume of sound from the guitar.

Rather than take this path, I decided to add on inside the rosette.  Using several strips of fibrous paper and one strip of maple veneer, I built up inside the rosette.  After evening out the sanding so that there was no remaining sign of the old spruce, I set two layers of paper, one layer of veneer, and then one final layer of paper.  The strips were cut to be tight so that no clamping would be required.  Three applications of a quick  setting, very fluid super glue were then made to the strips to set the hole package.  When I come back next week, I should be able to sand the strips back down to the inside of the rosette.  I 'm hoping the effect will look reasonably professional and intended, as if there were a ring of ebony set inside.

Fitting Sides

The other task for the evening was to trim the sides that were bent last week so they would fit flush to each other inside the mold and can be joined by gluing in the tail and heel blocks.  Sides are trimmed by pulling them tight inside of the mold and marking where they match the midpoint (end of a single side).  Using the band saw and belt sander, the sides were trimmed close to this mark so that both would fit inside the mold when bolted together.  It is important to trim bit by bit to get the tightest fight between the two sides so that the resulting joint will be tight and help hold the sides inside the mold.

Here's what the two sides look like put together inside the mold with the heel and neck blocks put in place but not yet glued.

 

The shape of what is to come is getting clearer!  A slight curve was put on the inside of the heel block with some light sanding so that the butt end of the guitar will have a slight curve.  Edward has switched over to using laminate at this end instead of the traditional mahogany in order to add strength, especially for those guitars that get dropped onto a strap end pint.

The mahogany neck block on the other end had two holes pre-drilled to later accept the neck.  The necks will be joined flat up against the end of the guitar using two bolts.  The top of the guitar is currently face down, and the still serves as the primary registration due to the curve (15-16 foot radius) put into the back to end strength to the instrument and to better let it respond to changes in humidity resulting in changes to the size of the back.  The bottom of the neck block (up in this picture) has had a preliminary curve put into it to match the back  Later on it will be sanded down to this curve in a sanding bowl. Next week I'll be gluing this all up, sanding down the rosette, and starting work on putting braces into the back.

Class 3: Anatomy of a Screw Up

Thoughts on Making Mistakes

The plan for the night was to finish routing the channel for the soundhole rosette and then to set and glue the rosette into the soundboard.  The other activity was to bend my sides.  Things started out well as I routed out the channel in three passes with the circle cutting jig set to a depth of about .075" inches, a little bit shy of the thickness of the maple piece.  For each pass I first routed a channel in the MDF work board, then moved over to the top of the soundboard.  The first pass went through the middle of the channel, the second to the inside radius, and the third to the outside radius.  After the second and third passes in the MDF, I tested the fit of one of the rosette pieces along with two strips of purfling.  After the third pass everything looked to fit well.

The only remaining step was to rout a final circle through the thickness of the soundboard to cut out the actual sound hole.  After adjusting the depth and radius, I made the final pass, but it was clear from a little way in that something was wrong.  Turns out I forgot to turn the locking nuts on the jig, so as I cut with the router, the radius of the cut kept increasing.  The result: a spiral soundhole that cut up to the edge of the rosette channel.  After at least 5 passes with the router to cut out the rosette and dig the channel, on the very last pass I forgot a basic step.

Here's a closer shot of the piece that came out.  This should be a perfect circle.

Here's where it also gets interesting working with Edward.  His first instinct was that I might need to start a new soundboard, but after closer inspection he thinks it can be saved.  For now, what I'm trying to do is to glue the rosette into the channel, surrounded by the two purflings.  After it has glued, we will finish removing the remaining wood on the inside edge of the sound hole until the outside purfling becomes the edge of the sound hole.  At that point, we may look at adding an additional strip of wood to give a little more reinforcement to the sound hole.  It should look quite nice and may result (?) in a little more volume from the guitar.

I appreciate Edward's approach to problems.  Generally, mistakes are  looked at as part of the course of developing skill rather than as evidence of imperfection.  While we are trying to make the best guitars we can in the class, by being careful and using the skills we have, it is understood that the experience of learning and developing new skill is part of the journey and that the completion of a single instrument should not be seen as the end of our travels.  This attitude which runs through his classes is probably what has made this process so enjoyable for me.  It stresses the chance to continue developing rather than the completion of products.  This is not to say that there is no interest in completing the projects, but the feeling of growth is in many ways as important.

So, I took Edward's advice and continued using what I had - here's the rosette being glued in, covered by a piece of wax paper and a caul to compress it evenly.  The edge of purfling where the channel has no inner wall is held against a few finish nails tacked into the piece of MDF.  I'll have to see what it looks like after the break for Thanksgiving.

Here's a shot of a similar rosette done by another student.  This is how mine was intended to look.The strip of spruce on the inside of the rosette is what will be missing.








Side Bending

I also bent the sides of my guitar and the binding strips.  I think they look pretty nice:

The first side took some additional manipulation to get set evenly in the mold, as a result it was essential bent twice.  This might be why there are some compression marks along the waist.  Edwards thinks that these will mostly sand out.  More adventures.

It was also interesting to contrast my experience in working on this project with a conversation at lunch with a friend who's late father was a serious luthier.  Apparently he was a woodworker and a machinist with much passion for the craft.  The type of craftsman who honored his tools, sought out his own materials and rarely took shortcuts.  Almost all of his work was done without power tools - planes were apparently his passion.  Although, he apparently did not make his own rosettes (another lesson here?).  I would love to see some of his work. 

However, sitting at lunch with my friend and his brother, both of whom were brought up in their dad's shop, neither could profess any interest in woodworking on their own.  Apparently their father had always viewed their efforts as instructional opportunities.  Praise was never available.  As a result, neither ever developed any enjoyment for the activity.  To put it all simply, attitude seems to be important.

Class 2: Rosette and Preparing to Bend Sides

Soundhole Rosette

I continued to work through a couple of iterations of the radii I would need to define the inside and outside curves of the rosette. 

These were laid out on paper and then on the piece of spalted maple.  The maple was taped with double sided tape onto a piece of MDF in order to be cut with a router with a circle cutting jig.  The router is equipped with a downward spiral flute bit - the downward spiral helps to ensure that the edges of the cut are not torn upward.

Afterward, the resulting circle had to be carefully separated from the MDF in order not to break the ring; inevitably,  however, it did break into a few pieces.  Once broken, however, it is easily reassembled.  Additional 'grain' can be added by using a pen to color the would around the fractures, but I'm skipping that step.

Here is the ring (now in 4 parts) ready to be laid into the soundboard next week.  In order to define the channel into which the pieces will be inscribed a channel will be routed in the sound board.  Layout of this channel has to also allow for room for an inside and outside strip of purfling as well as the glue joints between the wooden rosette ring.

Edward estimates that each glue joint adds .005" to the required channel with, and the purfling strips measure .060" across.  The channel width, with two purfling strips and associated glue joints, must be a total of .130" wider than the ring.  Measurement of the ring is taken off of the MDF on which the ring was cut out.  The measurements are then increased the required  amount and then laid out on the top.  A dry run to sizing the channel will also be done on the MDF board on which I am working.  The ring is about .080" thick, so we need to get the router bit depth a little shy of that (about .070") so that the rosette sits a little proud of the soundboard prior to thickness sanding.

Bending

Half of the class bent the sides for their guitars today.  I'll do mine next week, but got them set up with the binding strips in preparation.

Each of the two sides was taped up with a binding strip on the top and bottom so that the bindings can be bent together with the side in order to have the same curve.

The bending process begins with soaking the wood for several hours.  Each side is then heated in a sandwich made of the side, encased by two pieces of sheet metal, encased by heating blankets (see the first picture in the sequence below).  These blankets can get north of 500 degrees, so the sides heat quickly and give off steam.  After heating for a while, the whole package is firmly forced into the mold then clamped together for the sides to cure.

Here's the sequence, demonstrated by Edward

Edward uses the following timing in this sequence:

1. Heating blankets are plugged in and the side is heated for 2.5 minutes.
2. While still heating (use gloves), the side is forced into the mold.
3. At the 4 minute mark, the heating blankets are turned off, so as not to overheat and burn out over time.  LMI recommends using the blankets with a rheostat or other control so as to not overheat and burn out, but Edward uses these timings based on full throttle.
   
4. After a 30 seconds, (4.5 minute mark), the blankets are turned on for another 45 to 60 seconds.
5. At about 5:15 or 5:30, the blankets are unplugged and the sides are left in the frame for an hour before unclamping.

All three sets done last night seemed to unclamp and come out fine.  Next week, I'll be bending my sides and finishing the channel for the rosette so that it can be glued up.

Design details

Soundhole Rosette
This weeks homework is to come up with a design for the soundhole rosette.  The rosette will be cut out of a piece of spalted maple and surrounding by purfling strips.  So the design is a fairly simple affair: working out the width of the rosette and the offset from the soundhole.  The proportions I'm looking at are shown below (not to scale)

The soundhole diameter is about 3-3/4", the rosette will be offset a quarter inch and measure just under 1/2" across.  I'm wondering about breaking it into 4 quarters (as shown above) with additional purfling strips.

Headstock
I've also been thinking about the headstock design.  I'd like to use another piece of walnut for the plate and try to due some type of inlay either in wood or shell abalone.  It will have to be something simple, maybe a very basic compass rose.  I'll need to work up some designs, although this is probably a couple months out.

First Class

Initial Thoughts
Tonight was the initial class.  The first half was spent on an orientation to Edward's shop and teaching

process.  From there we started talking about design issues and the role of materials - wood - in affecting a design.  We discussed some of the physical qualities of wood - it's response to moisture, tension and vibration and the importance of how wood is cut.  Guitar woods need to have straight, continuous grain, so boards are almost always quarter saw (here are some nice images of a quarter sawn log).

The art of building a guitar is to bring the materials together in a design that is just strong enough to resist the tension of the strings (up to 150 lbs/psi for a steel string guitar) without falling apart.

For me, on this project, the other design key is size.  The OM guitar that I'm starting on has a smaller body than a dreadnought, emphasizing the treble and less bass and making it more appropriate for finger picking.  I'm hoping that the walnut also helps to bring out a strong voice for the treble while still providing a solid bass.

Materials

Here is a preview of the woods that I am using for this class.  The top is Engelmann Spruce which has consistent color and grain.  For the back and sides I am using Peruvian Walnut which is a relatively inexpensive tone wood with consistent grain and should bend easily.  I'm also hoping to get strong treble tones from it.

The bindings will be Koa, although I debated using some darker Rosewood to get more contrast with the top.  Spalted maple will be the basis of the soundhole rosette.

Top, back and Side Layout
Layout began with tracing the provided template onto the back and top.  The template is registered off the center to ensure that the jointed seam in both the top and bottom becomes the center of the top and side pieces.

Here are the back and top with the OM shape drawn onto them.  These were then rough cut on the bandsaw.

First cuts made and it already looks like a guitar.

Next the sides needed to be laid out.  The trick is to understand which end will go where (butt or neck) and to preserve the bookmatched grain.  The sides are marked up as shown in the diagram below.

The pieces are then taped together with the butt ends and neck ends together so that a single template can then be laid across both.  Key is to making sure the top edge is kept together.  This edge will match to the top of the guitar in one of the few right angles in the project.

Once laid out these pieces are also cut to depth following the curve of the template.  The initial pieces are now all roughed out


Next week we will be working on cutting out the rosette for the soundhole and preparing to bend the sides.  Homework this week is to figure out the sizing and location of the rosette in relationship to the soundhole, which has a diameter of 3.75 inches.