Fully Braced Sound Board

Here you can see the fully braced sound board, with all the scallops cut.  I still haven't received the parts for my band-saw, so I will likely put off the back until I get the parts, and then again, maybe not.  My wife promised that we could go in search of wood stores on Saturday next -- as she put it, not my favorite thing to do -- and if I can find a suitable piece of mahogany, I will resaw it again at the school.

To tap tune it, I hold it by spreading my fingers in the sound hole, then rap it with my knuckle at the bridge plate.  I sand and scallop until I no longer hear over-tones.  None of the bracing is of a standard size, and frankly, I don't think it's that terribly important.  If you look at several guitars, they are all braced differently, and they all swear by the pattern and dimensions of their bracing.  If you are going for a certain tone (e.g. vintage Martin) I would copy the dimensions, but I am still finding my tone, and I don't mind producing something slightly unusual.


For those who care, however, the x-bracing is 3/8th wide by 1/2 inch high.  The upper traverse brace (above the sound hole) is likewise 3/8th wide by 1/2 inch high.  The lower bout bracing is 1/4 wide by 1/2 inch high.  The finger bracing is 1/4 wide by 3/8 inch high.  It's a little heavier than you might find on most guitars, but tap tells me I'm not sacrificing much in the way of response and I want the redwood to be braced a bit heavier than spruce.  It's stiff enough, but not as stiff as the spruce, and I don't want it bulging under the pressure of the strings.     

Sound Board Bracing

You can see the finished Go-Bar deck.  It took all of a half hour to build.  I cut two pieces of 3/4 MDF slightly larger than my radius dishes, nipped the corners off, drilled holes and enlarged them slightly on the top, then bolted the top and the bottom in place.  I used 5/16 threaded rods.  I could have used larger, and they would have had less wiggle, but the 5/16 are working quite well.  I didn't cut the rods, but left them at the full 36 inches.  The extra length will allow me to raise the upper deck, if I want to use it for projects other than guitars.  (My wife wants me to make some boxes for a variety of tea bags, something well crafted that could be used on the table.) 

The go-bars themselves, I made from poplar.  I had thought about ordering some fiberglass rods from McMaster-Carr, but frankly didn't want to wait.  Poplar is a stringy, hence springy, wood, so I cut some strips at about 1/8th inch thick.  Any thicker and they're too stiff.  At any rate, they work like a champ.

I have an order of precedence in gluing on the braces, and I won't outline it here.  It begins with the X-Bracing.  I begin the process by cutting them from lumber.  I'm cutting the bracing from a redwood 2 by 4 that I rescued from the Home Depot.  The first trick is the grain.  I wanted a piece with tight, even and straight grain.  It doesn't where that grain is on the billet (I'll call it a billet from here on out) but when the brace is cut, the grain should run the length of brace.  Looking at the end grain, it should be vertical to the soundboard.

I cut my X-Braces at 3/8 wide and 3/4 inches high.  They will lose a bit of that dimension when planed and sanded, but it also makes for a stiffer top.  To radius the braces, I do two things.  First, I begin by planing a preliminary radius on the brace.  Take a cut with the plane about one inch from the end, another at an inch and a half, another at two, et cetera, until you reach the center of the brace.  Do the same for the other end, then run a squiggly (technical term) pencil line down the brace.  I finish on the radius dish, as you can see in the photo, by laying sandpaper at the approximate place where the brace would lie on the sound board.  Using light downward pressure, and short strokes, I sand until the pencil lines are gone.  When I have done that, just to be sure, I put another squiggle down the brace, and sand that off as well.  I say "light downward pressure," because you don't want to bend the brace into the radius, you want to sand it into the radius.

I join the X-Brace before I glue them in place.  Layout is important.  The first step is to find the place where bridge will be on the top.  Using the LMI ruler, I measure and mark the bridge location.  I then mark the outside string locations.  The X-Brace should allow for about 1/2 inch on either side of each e string.  You want the pins that hold the strings to pass through the bridge plate, not the bracing.  The joint itself is essentially just a half-lap joint at the angle of the X.  After I have sanded the radius into it, I lay it on the soundboard, put the other brace over it, and mark the angle, then cut it with a razor saw.  I put the other brace over the other, so it covers the razor saw kerf, mark with a sharp x-acto knife, then cut the second kerf to the same depth, about half-way through the brace.   The waste between the kerfs should pop right out with a sharp chisel.


The trick is not to cut it too wide.  Better to cut the channel too narrow and take some shavings to get it to the point where the other brace slips into it snugly.  With the other brace in the channel, I mark it with the knife, cut two kerfs just inside the markings, make the adjustments for fit, and glue them together.   You don't really need to wait for the glue to dry before gluing the brace to the top, and in fact, I didn't.

You can see the go-bar deck in its full glory in the photo at left.  Here the trick is to get the brace in place with the first bar.  With the glue on the braces (and I do use Titebond) they have a tendency to slip around.  Once it is in place, then it is simply a matter of placing the go-bars so you get a nice even line of squeeze out along the entire length of the brace with no gaps.  This shouldn't be a big problem if a proper radius was sanded into the brace.  Although you can't really see it in the photo, I do have a sheet of poster board between the soundboard and the dish.  Although it isn't a big deal right now, eventually, when it comes time to adjust the kerfing on the sides, I will cover the dish with sandpaper, and so I thought I would get in the habit of protecting the show face of the sound board right from the start.

Before I glued in the bridge plate, I did scallop the X-Bracing.  The scallops on the ends of the braces start at about 1 and 1/2 inches from the sides of the guitar.  The secondary scallops are positioned so they line up with the bridge plate.

Cutting a scallop takes a little practice, but not much.  Start with the sharp chisel.  (Mine this morning were a bit dull, so I will need to attend to it.)   For the end scallops, start at about 1/2 inch back from the end.  With the back side of the chisel up, cut on either side of the brace to create a peak, then chisel then scoop the peak away.  Go back a bit on the brace, and repeat the process, cut away the sides until you have a peak, then scoop the peak away.  For the secondary scallops, at the center of the scallop, cut a kerf to the depth of the scallop.  Then do as above, just working from the center kerf to the mark.   

The next step is the bridge plate.  It is made from a block of (I think its cocobolo) that I found in the garage many years ago when we were living in Michigan. It is a beautiful piece of dense wood, and this is the first cut made into it.  I had in mind that someday, when the spirit moves me, I will make a morris style plane with the remainder of the block, but for the moment, I will let it serve as a bridge plate.  It is about 1/8th inch thick, and positioning is important.  Although I plan to make a pinless bridge, the bridge plate should be below the bridge and the pins should pass through it (not the braces!).    You can see the shaping from the photos, and the reason for the particular shape will be obvious when I install the next set of braces.  

Rosette

I finished the rosette, almost.  There is still a little tweaking that I need to do at the joints of the wafers, but it is essentially what it will be.  The tile at the top is there primarily as reinforcement and fill.

The first step, of course, is layout.  To do this, I used my template, which has already be scribed, and traced the outline on the sound board.  I am going for a parlor-type guitar, which I interpret to mean a steel string guitar with classical like dimensions.  It also suits me because I play finger-style, and find myself switching between classical and steel strings, so some continuity in the dimensions works for me.  Consequently, I am using a 25 and 11/32nd inch scale length.  There are all sorts of ways of coming at this, but I have found that one purchase has repaid itself several times.  Get a fret rule from LMI.   Again, I am going for classical like dimensions, so I placed the 12th fret line at the top of the sound board, and marked the location of the 22nd fret on the center-line (the glue seam).  With my compass set at 1 7/8th inch diameter, I scribed the circle for the sound hole.

Once I had the tiles set out in the working scrap, which you can see beneath the rosette, the whole process became rather simple.  I set the router so it would ride within the channel that I had cut to fit the tiles, set the depth to just a whisker (about 1/64th of an inch) shy of the total depth of the tiles, double checked (actually, quadruple checked) to make sure that I would not be cutting through the soundboard at that depth, and then made the cut, testing first where it would be covered by the fretboard.  Once the initial cut had been made, I eased it out and in using the precision adjustment on my router jig to accommodate the tiles and the perfling strips.  When I was finished, I gave the rosette channel a coat of shellac to prevent bleed out from the glue.

I put all the tiles and perfling in place, leaving slight gaps at the seams so I could lift them out, one by one.  Beginning at the bottom of the rosette, I would lift out one tile, run a bead of superglue along each of the perfling strips.  I then replace and quickly position the tile between the perfling strips.  I kept going I had worked up both sides of the rosette.  I wiped away any squeeze out, and pressed the tiles into place with my fingers until the glue set.    

Before I cut out the sound hole, I did a couple of things.  First, using my template, I transferred the outline of the guitar to the back of the sound board.  That was a simple matter of using the 1/4 hole for the router pivot, and a similar hole at the center of the sound hole in my template.  When sanding the back, make sure that you never sand away the lines that they cannot be retraced.  I also leveled the rosette.  Nothing fancy there, I just used my trusty card scraper, newly sharpened, taking light passes until the rosette was level with the sound board.   I then sanded it using 120 grit sandpaper on my sanding block.  A word of advice, never sand without a sanding block.  It's amazing how quickly one can cut shallow hollows sanding without a block.  Mine is just a scrap of cherry with a cork face.  I wrap the sandpaper over it and hold it in place as I sand.

To cut out the sound hole, I swapped the 1/4 inch bit in my router to a 1/8th inch bit, made the adjustments so there would be a lip of redwood approximately 1/8th inch wide, made an initial cut, then cut through.   Some have jigs to round over the edge of the sound hole.  I don't.  I just use a light touch with some hand held sandpaper.  I can't tell the difference.
 

Prepping and Joining the Plates

I re-sawed the plates for the sound board at the college before the Christmas break.  I ran them through the thickness sander until they were approximately the right thickness, then used my jointer to prepare the joint.  In the absence of a jointer, one could use a shooting board, and I know that some luthiers prefer the hand method regardless.  I do have a jointer, however, and it works perfectly fine.  I take the two book matched plates, and run them over the jointer together and I get a tight joint that stand up to the light test (holding the joint together and trying to see light through it).  My secondary test is that it squeezes out glue uniformly along its entire length.  That too tells me the joint will be tight.   


I used the lattice jig, pictured, left to join them up.  It worked precisely as advertised.  I glued the seam (using titebond) rubbed it until the glue began to set, then put it in the jig.  I wrapped the ropes in a criss-cross pattern loosely over the jig, then the loose ends I wrapped around one end, then drove the wedge shaped slats under the criss-crossed ropes.  It pulled the plates together sufficiently, and kept them dead flat while the glue dried.   You'll note there is no wax paper under the slats, but the slats themselves were well coated with wax before the glue up. 


The jointed plates are pictured at right, and you are looking at the exterior side.  After all the fussiness with the making of the rosette tiles, it was almost shameful how easily the plates went together with the jig.  You will note that there is a slight difference in color on either side of the center line.  The plate on the right has oxidized some, but as I do the finished sanding, that difference should disappear.  The next step will be the leveling of the top and the inlay of the rosette.  If all goes according to plan, I will tackle that in the coming weekend. 

Cutting the Rosette Tiles

Here's how I do it.   I have a "precision" router set up.   You can see it in the photo, and I will admit right up from there are a few design flaws in it.  First, as you can see, the router base itself is attached to a plexiglass base.  This actually works rather well for the operation I am about to describe, in part because the base itself is wide enough that it covers the round of tiles.  There are three slots cut in the plexiglass, two for the adjustment mechanism, one for the pivot pin.

The adjustment mechanism is basically two steel rods that pass through the hold down and the pivot.  The two rods are epoxied into the pivot.  The two carriage bolts pass through the plexiglass and serve three purposes: (1) sliding in the slots, they serve as a gross adjustment, (2) they lock the steel rods in place, and (3) they lock the adjustment mechanism in place.  It isn't perhaps as visible in the photo, but but there is a saw kerf cut through the holes for the rods.  The carriage bolts pinch the rods when the wing nuts are tightened.  That is the first design flaw.  I should be able to loosen the rods without loosening the adjustment mechanism, but I have discovered that the "precision" mechanism can be adjusted by loosening the two carriage bolts just slightly, holding it in place with my thumb as I make the adjustments to the screw, which serves as the precision mechanism.   

The "precision" mechanism is simply a 1/4 by 20 steel rod that passes through a bolt inset into the adjustment mechanism into the pivot, where I inset a washer and a bolt.  To make the radius smaller, the "precision" mechanism is adjusted clockwise, pushing it out or closer to the router.  To make the radius larger, the "precision" mechanism is adjusted counter-clockwise, pulling it in or further from the router itself.  It works reasonably well, but there is some play in it, and that constitutes the second design flaw.  Someday, I may break down and buy the set up sold by Stew Mac.

The first step in making the rosette tiles is the template jig, the second is cutting the slices that will be used for the inlay, and the third is shaping the pieces to fit the jig.   My sound hole is 1 and 7/8 inches in radius, so I scribe a circle of that radius on a scrap of 3/4 MDF.  I also scribe a line at 2 and 1/8th for the inside diameter (ID) of the rosette, and then 2 and 5/8th for the outside diameter.   I scribe two copies of the rosette.

The next step is to use the rosette template to cut a couple of circles that will produce a template that is the mirror image of the rosette channel.   You can see it in the photo above.  I set my router to cut an OD channel, then an ID channel, leaving the mirror image of the rosette channel proud in the center.  This will be used to shape the inlay pieces.

The inlay material itself is a pile of wafers cut from the end grain of the padauk at about 1/8 thickness.   Normally, I would use my band saw for this, but I'm still awaiting parts, so I use a cross cut sled and a stop, and grit my teeth at the waste.  I put double stick tape on the proud rosette channel, then tape the wafers to that.  Make certain that they are secure.  If they wobble around you will not get precise cuts.  You can see that pictured above.   

I then cut the ID and OD with the router set up.  It pivots on the pin pictured.  It will take several passes, so again make sure the wafers are well secured with the double stick tape, and that the router base does not catch on an edge.   Cut the OD and then the ID with the router.  For the second batch, I leave the router setting in place for the ID, and cut the ID then make the adjustments to cut the OD.  For the third batch, I leave the router setting in place for the OD then make the adjustments to cut the ID, alternating until there are enough shaped tiles. 

The next step is to cut a channel that will fit the rosette tiles.   I consider this practice before I cut the channel in the soundboard, but it also serves a purpose as a jig to get the angles correct on the ends of the pieces.  I will cut a kerf through the rosette channel on the table saw, then use the kerf as a guide for trimming the ends with my gents saw, using the kerf as a miter box.  By the by, to get the pieces perfectly sized, I cut the rosette channel to fit the narrowest piece as precisely as possible, then make adjustments to the other pieces with sandpaper.  Be cautious not to round over the ends if you have already cut the miters, or sand to fit before you cut the miters.  Also, it always a good idea to have several more tiles than you need.  Because they are cut from end grain, they are fragile, and if you break one (and I broke two) you don't need to go through the whole business of the router set up again.