A fishy tale of sound hole decoration


The finished sound hole decoration, and sound hole

Progress on the parlour guitar has slowed a little since the Apprentice arrived, but I did find the opportunity to install the sound hole decoration this week. I had intended to work on the sound hole decoration at EWS, and to maximise the number of skills shown in the demonstration my decoration design was alternating blocks of ebony and mother of pearl between ebony banding. As it happened, I had no opportunity to do any work at EWS thanks to the level of interest from attendees. Since EWS I have been pondering the design again, and decided to revert back to my original design of a simple herringbone inlay. The corner purfling on the guitar will also be herringbone, so the matching sound hole decoration will avoid introducing another clashing element, which is particularly important on a small bodied guitar when the decorative elements are so much closer together. The previous design work was not wasted, as I was able to use the same spacing and circumferences as I would for the shell/ ebony design.


Before I make any cuts to the soundboard, I always make a test piece using offcuts from the soundboard timber. This allows me to check the spacing of the cuts and ensure that the inlay fits before I make the critical cuts for real. The two edges of the trench for the inlay were cut using a rosette cutter from LMII, after which I removed the bulk of the waste with a Lie Nielsen No.271 small router plane. The depth of the trench is critical, as the soundboard is only 2.7mm thick and the herringbone inlay is a little over 2mm high. This gives very little margin for error, and I set the rosette cutter blade to just a hair under 2mm so that a small amount of the herringbone could be removed after gluing.


With the bulk of the waste removed, I then fine tuned the trench using a 0.81mm mini chisel and small marking knife, both by Blue Spruce Tool Works. With these I am able to remove any debris from the two corners at the bottom of the trench, so that the inlay fits flush to the bottom.


Bending the herringbone on a bending iron

The herringbone banding is very fragile and won’t bend without some gentle persuasion. I bent the herringbone to the correct radius on my bending iron, going very slowly to avoid the edge banding snapping, or (even worse) the marquetry de-bonding and scattering  pieces across the workshop. When bending thin banding such as this, the trickiest part is keeping the length of the banding in a consistent horizontal plane, as applying heat to just one side will encourage it to slowly corkscrew as it bends. This will result in an inconsistent pattern once the inlay is planed flush to the soundboard. Once the banding was bent to the correct radius I levelled out the horizontal orientation by placing the banding on the flat top of the bending iron and trapped it under a broad chisel for a couple of seconds.


The banding is inlayed and ready to be levelled to the soundboard.

The banding was a good tight fit in the trench, so a final touch was to bevel the leading edges of the banding with a cabinet scraper. This eased the banding into the trench and ensured that the edges of the trench would not be bruised when the banding was inserted. The banding was glued in using Titebond Liquid Hide Glue, which had been warmed by leaving the glue tube in a jug of hot water so that it was less viscous. Ordinarily I mix my own hide glue, but hadn’t much opportunity recently. The Titebond worked well and I will certainly use it for future work. With a liberal coat of glue in the trench, the banding was placed in position and gently tapped fully into the trench with a small hammer.

DSC_0752I left the banding to cure overnight, and then planed the excess banding flush with my low angle block plane, followed by a small cabinet scraper. The scraper allows the banding to be taken perfectly flush without removing any material from the soundboard, which is critical given the thickness of the soundboard at this point – there is little opportunity for sanding or removing plane tracks once the soundboard has been thicknessed.

With the decoration complete the sound hole was then cut to a 90mm diameter, again using the LMII rosette cutter and the blade set to the full thickness of the sound board. Next I will be writing about how I approach bracing the soundboard.


Cutting the sound hole

Sounding off about… design choices


The parlour guitar has presented some interesting design choices, particularly with regards to the soundhole decoration. I don’t have the design chops of George Walker (oh but I wish I did) but I do take the design of my guitars seriously and I thought I would write about some of these decision here.

When laying out the soundhole decoration for full sized guitars I try to follow several rules of thumb. The outside circumference of the decoration should meet the corners of the fretboard, with the end of the fretboard falling over the top section of the decoration. The length of the fretboard is calculated as being the number of frets desired for the particular guitar plus an extra fret’s worth of fretboard material (so for a 22 fret guitar the fretboard would be 23 frets long).

However these rules of thumb do not all work for the parlour guitar due to the smaller body size. The parlour guitar has a 90mm diameter soundhole (compared to 100mm diameter for a full sized guitar) and 18 frets. But setting the diameter of the soundhole decoration large enough to meet the corners of the fretboard would be in danger of visually swamping the smaller body of the guitar. After drawing various mockups of the decoration, I decided to extend the fretboard so that the end meets with the edge of the soundhole. This allows the soundhole decoration to meet the corners of the fretboard, but on a tighter 106mm outside diameter which remains proportionate to the size of the body.

The decoration itself is a 4mm wide band comprising 0.6mm ebony banding on each circumference, with 2.4mm inlay in between the ebony. The main inlay blocks are eleven alternating ebony and mother of pearl pieces (the ebony blocks are shaded black in the above drawing).

I will be cutting and installing the inlay at the European Woodworking show on 12 and 13 September, so stop by my bench at the show and watch it unfold in real time!

Just a pretty face and thin body

I prefer to leave thin jointed surfaces clamped up over night to allow the glue to cure fully before further work, and so I removed the parlour guitar soundboard from the thin panel gluing jig the morning after it had been glued up. Because I place panels face down in the gluing jig, there can be a slight step on the rear of the join if the panels are not of consistent thickness, so before surfacing the show face of the soundboard I removed this step from the inner face using a block plane.

The final surfaced and thicknessed soundboard

The final surfaced and thicknessed soundboard

When surfacing soundboards and backs, it is imperative to get the work as flat as possible, as well as achieving a clean surface free from tear out, and this involves constant checking of the work with a machinist’s straightedge. In the past I have used a No.5 jack plane set to a fine cut, followed by a cabinet scraper to remove any hint of plane tracks, but this time around I decided to give the No.8 jointer plane a go, and the extra width of the blade allowed me to achieve a flat surface quicker than I would have with the No.5, and left only a little work for the cabinet scraper. For difficult timbers (particularly the backs and sides of acoustic guitars, which use exotics like rosewood) I use a toothed blade, and I may have to look into getting one made for the No.8 for this purpose. Because the soundboard panels are bookmatched, it is important to change planing direction on each half of the work, so as to avoid tear out, and the cabinet scraper is invaluable for cleaning up the glue line as it seems to resist tearing out while working against the grain.

The surfaced yellow cedar panel

The surfaced yellow cedar panel

With the full panel surfaced, I marked out the shape of the soundboard using my plywood template, and then cut to shape (staying bang on the pencil line) with a fret saw.

Using a half-body template ensures that the final soundboard will be symmetrical

Using a half-body template ensures that the final soundboard will be symmetrical

The thickness of the soundboard is critical for the sound of the finished guitar, and with the outer face now surfaced I was able to start thicknessing the soundboard from the inner face. The purpose of thicknessing is to achieve the right balance between a thin soundboard which will be responsive and resonant, while remaining strong enough to withstand string tension without imploding. The advantage of handmade guitars over mass produced machine produced instruments is that each soundboard and back can be worked to the optimum thickness for that particular piece of wood, with the bracing further adjusted to reflect the properties of the soundboard and back.

Measuring the thickness of the soundboard

Measuring the thickness of the soundboard

Using my large calipers. I mark the thickness of the soundboard at 2″ intervals, and then work the high points down until the the soundboard is down to a nominal thickness, using block and bench planes. This is a very precise way of working, and with planes set for a fine cut I know that as soon as the pencil numbers have been removed from the soundboard I have reduced the thickness by 0.1mm of material. Once each number has been removed, I remeasure everything and write the new thickness of the soundboard at the same 2″ intervals. Once I am close to the nominal thickness, the numbers stop being as important, because I am testing the wood by flexing it and listening to the sound it makes when tapped with my thumb. The need to constantly remeasure means that thicknessing a soundboard will take a couple of hours, but the end result is worth it. For a yellow cedar soundboard I work to a nominal thickness of 2.7mm, and this soundboard ended up just a hair under that as the timber had good stiffness and tap tone, which allowed me to remove just a tiny bit more material.


With the soundboard thicknessed, I was then able to cut the rebate in the neck which accept the soundboard when the guitar is assembled.

The rebate cut into the neck

The rebate cut into the neck

This joint needs to be cut precisely, as the top surface of the neck and soundboard need to be completely coplanar. To achieve this, I set my marking gauge a hair less than the full depth of the rebate (2.5mm instead of the 2.65mm thickness of the soundboard) and once I reached my baseline I checked the fit using the soundboard, slowly sneaking up on the final fit with a very fine cut. To cut the rebate I bevelled the two outside edges with a chisel to prevent spelching, and then used a Lie-Nielsen to remove the waste.

The soundboard seated in the rebate

The soundboard seated in the rebate

Shooting from the (hip?)

A couple of days before Grace was born, I pressed the new iteration of my thin panel gluing jig into service gluing up a soundboard for the parlour guitar. Although I started this guitar build with the neck, I normally prepare the soundboard and back first, and one of the things I love about building acoustic guitars is that very early operations have a decisive effect on the appearance and sound of the final instrument. Preparing the soundboard is one such critical task.

When jointing a bookmatched soundboard you have a choice between two edges for the joint, and this choice is critical in how the guitar will ultimately sound. I always joint the plates so that the tight, close grain is at the centre of the soundboard, with the grain becoming wider as you move towards the edge of the instrument. This ensures that the soundboard is as stiff as possible, which in turn allows the bracing to be reduced for a more responsive top (conversely a soundboard which is less stiff will be less able to withstand string tension without heavier braces – acoustic guitar building is a constant balancing of various factors).


With the tight grain side of each board identified, my very first step is to take a couple of light swipes with a smoothing or block plane over the face of the board, just over 1 inch or so at the edge to be jointed. This cleans some of the fur off the board and allows me to see the grain properly. I then count the grain bands from the edge until it is possible to identify on both boards where the joint should fall. This is essential for ensuring that the finished soundboard has symmetrical grain radiating outwards from the centreline. I mark the joint position using a fine (0.3mm) mechanical pencil, and then joint each plate.


To joint the plates I use a long grain shooting board and a No.8 jointer plane (although I have used a No.5 in the past, and experienced o difficulties in getting a good joint). A lot of woodworkers seem to think of long grain shooting boards as being training wheels for the inexperienced, but for thin stock where it is not possible to balance the plane on the edge in a consistent orientation, the long grain shooting board is invaluable. My long grain shooting board is nothing fancy, just a piece of 5mm mdf glued to a wider, 25mm thick plywood base. Because the joint for a soundboard is rarely at 90 degrees to the top edge of the stock I don’t have a full width fence on this shooting board as you would for an end grain shooting board. Instead I have a small piece of marine ply which catches the top left corner of the soundboard, and I hold the stock in position with my left hand. Similarly, the plane does not ride against the edge of the top layer of the shooting board to maintain a straight cut, instead I keep the plane in a straight cut by varying where I place the pressure. My right hand holds the side of the plane, with my forearm on the rear tote. I then attempt to plane a concave curve into the stock (which thanks to the length of the plane, is impossible), pushing in with my hand on the first part of the cut, and then transferring pressure to my forearm on the second half of the cut. With a fine cut, and constantly moving between the two boards, this produces a good solid joint with no gaps pretty quickly (this soundboard took me about 90 minutes from laying out the joint to glue-up).

The test for a good joint in thin stock isn’t with the straight edge. Instead, I hold the two boards together and try and rotate them across the joint. If the joint is tight then it will lock in place under minimal hand pressure and the plates won’t slip, nor will any light show through. If more work is needed then the plates will not “lock together” and will constantly slip off each other.


With a nice tight joint established, both plates are placed face down in the gluing jig on top of a sheet of grease proof paper (to avoid sticking the plates to the jig), and the moveable beam is adjusted so that there is roughly 10mm of space between the fixed beam and the edge of the plate. Glue is applied to both sides of the joint (I use Titebond Original) and a rub joint used to achieve the initial adhesion. The wedges are then hammered into the slots between the plates and the fixed beam of the jig. Finally, a piece of scrap is clamped on top of each plate to prevent the plates from springing out of the jig under pressure from the wedges. The soundboard was then left in the jig overnight to cure.

Over the Wireless On Tour: European Woodworking Show 2015


I am very pleased to announce that I have been invited to pitch up my bench at the European Woodworking Show at Cressing Temple on 12 and 13 September this year!

I will be working on the parlour guitar and demonstrating some lutherie techniques over the course of the weekend. I also plan to have both my Anarchist’s Tool Ches and Esmerelda in tow.

If you are planning to attend EWS 2015 then please do take the time to stop by my bench and say hello; it will be great to meet new readers and say hi to some old friends.

Nipping at your heels, part 2

IMG_0501It seems like a long time since I started carving the heel for the parlour guitar (last August in fact!). In many ways carving the heel is a great demonstration of the benefits of dividing a task into coarse, medium, and fine stages, as taking a three stage approach promotes a very efficient way of working. Although many woodworkers naturally gravitate towards fine work (with those lovely wafer thin plane shavings) there is a real place for coarse work where quick removal of stock is desirable, before becoming progressively finer as less material is removed.

In my first post I wrote about the process for roughing out the heel shape, using a coping saw and chisels to hog out much of the waste. Now, with the rough shape established, I started to fair the curves and remove corners left by the chisels using a series of rasps. At this point in the process I was still using a relatively coarse rasp (9 grain 10″ cabinet maker’s rasp by Auriou) to remove the majority of the chisel marks. One of the key challenges when carving the heel is to avoid touching the centre line once the sweep of the heel is established, as this will encourage the centre line to wander and re-establishing it accurately will be the devil’s own job. Treating the centre line as sacrosanct and avoiding hitting it with any tools makes for a much less frustrating experience.

DSC_0085At this stage, maintaining the symmetry of the heel is critical, and there are a number of techniques to make this easier. Using raking light across the workpiece will throw the profile of the heel into sharp relief, and highlight any areas where more work needs to be done to keep things symmetrical. I keep a bright IKEA gooseneck lamp near my bench for this task, as it is easier to move the lamp then to unclamp and re-orientate the neck as I work different areas. Secondly, a piece of the backing off some double-sided sticky tape, held against the neck will highlight if the curves transition at different rates on each side of the centre line. Finally, much of the work on the heel is done looking down on it from behind the heel block (as with the photo at the top of this post), and it pays to occasionally stand at the headstock and sight down the neck towards the heel.

DSC_0659With the shape coming close to the final design, I switched to a 7″ 13 grain modeller’s rasp to remove the marks of the larger 9 grain rasp, as well as a 6″ 13 grain rat tail rasp to make very fine, localised, adjustments. The heel shape is now finalised, and as the 13 grain rasps leave a very fine surface all that will be required is some gentle finish sanding prior to spraying the lacquer.

DSC_0662Once the heel was carved I relieved the interior of the heel block. This was a simple task of marking out the waste area of the block, and making a number of relief cuts into the waste before chiselling it out. The corners were rounded over and softened, again using the 9 grain and 13 grain rasps. In the past I have spent more time carving a gentle curved relief into the heel block, which is more aesthetically pleasing but less practical if you then decide to fit a battery holder to the block for any internal pickup or microphone system. As no one will see the heel block (apart from the readers of this blog, obviously!) I decided to go for a more utilitarian heel block shape and save myself the heartache of trying to attach a battery holder to a curved surface.


Every rose has its [something that is a woodwork joke]

Despite the recent radio silence on this blog I’ve been hard at work cutting the inlay for the parlour guitar. The inlay process will be covered in an article in Furniture & Cabinetmaking later this year (which is why I’ve not written more about it here as yet). But for now here is a sneak preview of the finished inlay; rose in sterling silver, mother of pearl, and green and red abalone. All set into the ebony headstock veneer. I always enjoy inlay work, and I am quite pleased with how this design has turned out.

Headstock (shoulders, knees and toes, knees and toes) part 2

Having flattened the headstock veneer in my previous post, I marked out the centre line (ensuring that it followed up from the centre line on the neck) and layed out the headstock shape using the ply template I prepared last year. How I approach the shaping work depends on the design of the headstock, particularly whether there are any curves or just straight lines. As this guitar has a straight sided “paddle” design I worked the top first and then moved to the sides. All the shaping is done before the headstock is bought down to final thickness. DSC_0429The excess material from the top was removed with my Bad Axe carcass saw. This crosscut saw leaves such a fine finish that I was able to cut very close to the final line meaning that I only had minimal planing to do for the top; there is a reason this saw is one of the go-to tools in my workshop. I did a small amount of fine tuning on the top of the headstock with my low angle block plane, as this left a finish ready surface, even across the end grain. The advantage of planing the top edge before addressing the sides was that any spelching on the far corner of the top fell far outside the edge of the headstock – had I approached the sides first any spelching would have fallen within the final dimensions of the headstock. DSC_0433Most of the excess material off each side was removed with a coping saw, and then taken down to final dimension with the low angle block plane. The key when shaping a headstock is to ensure that the sides remain 90 degrees to the face of the work, so I keep a small engineer’s square to hand for constant checking. DSC_0444 There is something quite wonderful about planing the side of a veneered piece, and watching two-tone shavings pile up.   DSC_0438The curved transition from the nut to the headstock was roughed out using a 13 grain modellers rasp, and fine tuned with a 13 grain rat tail rasp. I use Auriou rasps, and find that they cut aggressively yet precisely, allowing me to work right up to the line and requiring minimal clean up. Again, the curved sides of the transition have to remain at 90 degrees to the front of the neck, so plenty of checking with an engineer’s square. The headstock is now ready to be thicknessed and for the slots cut for the tuning pegs. Stay tuned for more!

Headstock (shoulders, knees and toes, knees and toes), part 1

DSC_0408 With the truss rod installed, I was able to put my router away and return to the (much less terrifying) hand tool work I prefer. Next up was fitting the ebony veneer to the headstock. The bottom edge of the veneer had some significant splits, so I trimmed that edge with my Bad Axe carcass saw. Fortunately the veneer was significantly longer than I needed, so even having trimmed the split edge the veneer was plenty big enough for the headstock. DSC_0410As the veneer is only a couple of millimeters thick it is very flexible. My bench top is not flat enough to keep the veneer perfectly flat while planing the gluing surface, so instead I stuck the veneer to a sheet of 25mm thick ply wood using double sided sticky tape. Double sided tape is one of the secret tricks in my workshop, and is invaluable for working thin stock, holding gluing cauls in place, and securing router templates (on the few occasions I use my router). Once secured to the ply wood, the gluing surface of the veneer was then prepared with a Lie Nielsen No.212 Scraper plane. This plane is invaluable for the work I do, as it scrapes difficult exotic timber leaving a finish ready surface with no tear-out, and is one of the go to tool in my workshop when dealing with timber that would tear-out under a more conventional plane. DSC_0411 The bottom edge of the headstock veneer needs to fit against the back edge of the nut with no gaps. To do so it must be fitted in two directions: straight along the width of the headstock, and a chamfer planed into the thickness of the veneer so that the full thickness of the veneer buts up against the nut. Effectively, the chamfer is the angle of the headstock, plus 90 degrees, so that the resulting edge is perpendicular to the surface of the neck. The best way to fit the bottom edge of the headstock veneer is to mount it on the headstock, projecting over the surface of the neck, and to shoot it using the neck to support the plane. This method ensures that as the veneer is straightened along the width, the correct angle is also planed into the thickness of the veneer by virtue of using the headstock itself as the reference surface. I use my Lie Nielsen block plane for this operation, as the short length of the plane makes it easy to balance on the neck, and the sides of the plane are ground to 90 degrees of the sole which guarantees the correct angle on the chamfer. Not all planes have their sides dead on 90 degrees to the sole, so do check before you carry this task as there is a risk of planing a chamfer with some funky angles if you use an older plane. And from a different angle… DSC_0414 To glue the veneer to the headstock I clamped a 300mm steel rule to the neck so that it provided a reference edge against the back of the nut. The veneer was clamped under a caul to spread the clamping pressure, and left over night for the glue to cure.DSC_0417The veneer was broken out of the clamps once the glue had fully cured. I then flattened the face of the headstock, again using the scraper plane. To check my progress with thin stock where I want to remove as little material as possible, I often draw across the entire surface to be planed. As the lines disappear it becomes clear where there are hollows in the surface, and where the high points are. The white pencil in the picture below has been with me since Totnes (8 years ago now!), and I suspect it has very few sharpenings left! DSC_0422 The headstock is now ready to be shaped, which I will write about next time.

All Trussed Up

Now that the new workshop is set up I’m making progress on the parlour guitar build, and fitting the truss rod was the next stage of the build.

For this guitar the adjustment nut for the truss rod will be at the heel end of the neck, which means that there is no need to cut a truss rod cover in the headstock veneer. I find myself leaning more towards this orientation of the truss rod for acoustic guitars, as I find the absence of a truss rod cover leads to a cleaner and neater looking headstock.

DSC_0398 Orientating the truss rod this way does have some implications for the build process. With the slipper heel method of construction the heel block is an integral part of the neck, and a rebate is planed into the top of the heel block to accept the soundboard. As a result, the truss rod slot has to be cut deeper at the heel block end than the but end of the neck, so that there is sufficient clearance above the truss rod for the rebate. As a general rule cutting the truss rod channel 3mm deeper at the heel end is sufficient for allowing the rebate to be cut into the heel block.

Now, I’ve made no secret of the fact that my least favourite tool in the workshop is my router. That being said, there are times when a router is the best task for the job, and cutting the truss rod slot is one of them. To allow the router to cut the necessary slope I first trimmed two pieces of pine to length and using my No.5 jack plane, planed an incline in both, so that one end of each was to 3mm thinner than the other opposite end. Both pieces were then secured to the fretboard gluing surface of the neck using strong double sided sticky tape. This established the ramp on which the router would travel.

DSC_0395 Taking shallow passes I routed a channel using a 10mm wide router bit, with the router edge guide registering on the side of the pine ramp. The channel ended up 13.1mm deep at the heel block and 10.1 deep at the nut. The traditional “U” shaped truss rod is 10mm square, and the extra 0.1mm is to account for the fillet that is glued over the truss rod.

DSC_0400 I prepared the fillet by ripping a strip of yellow cedar off a soundboard off-cut, and planing to width. The open side of the truss rod was then covered up with masking tape to prevent the ingress of glue, and the truss rod was glued into the channel, with the open side faced down. The fillet was glued on top of the truss rod, and the whole assembly clamped underneath a large clamping caul to ensure even pressure and to avoid the clamps marking the flattened surface of the neck.

DSC_0401 The assembly was left clamped up over night to allow the glue to cure, and then broken out of the clamps. I then planed the fillet flush to the neck surface using my low angle block plane. I really love the fragrant scent of yellow cedar shavings, and it is a smell that always takes me back to the time I spent in Totnes. Just lovely. At the heel block the fillet currently lies below the surface of the neck and so couldn’t be cleaned up. This doesn’t bother me though as I will clean up that part when I cut the rebate for the soundboard in a couple of weeks.