Having managed to get to the initial graduations I am finally able to start the plate tuning.
First I took stock by weighing the plates and measuring mode frequencies.
I used the same setup I used for the cello:
Android tablet to generate the signal tone
20W class D amp
4″ driver unit (used an 8″ on the cello)
Tea-bag for tea leaves
Starting point with initial graduations
The top weighed in at 88g and had perfectly matched x-mode and ring mode (an octave apart) but still a bit high as expected at this stage.
The back weighed in at 119g and is still quite a bit too stiff with the ring mode going off both ends and not closing. The x-mode frequency needs bringing down quite a bit to get to an octave below the ring mode.
After a bit of work I managed to match the back and top ring modes an octave above the x-modes at about 180Hz. The plates are still a bit heavy but this gives me something to work with when tuning the bass bar after cutting the f-holes.
I marked out the outline of the f-holes using the printed paper template I had designed and pricking through with a pin to mark the outline on the top. I then drilled the top and bottom holes with a forstner bit and opened them up to correct dimensions with the reamers I use for peg holes.
With the top carefully clamped to the cork faced form, I used a fine toothed coping saw to cut a rough outline and then carved off the excess with a scalpel.
First results are shown here but still needs a bit of fettling to get the two holes balanced.
Next I sliced up a nice piece of quarter sawn spruce on the bandsaw and sanded down to 5.5mm wide on the drum sander. I cut this to length and marked the underside profile with a pencil using a block of wood while the blank bar was gently clamped in place on the belly. Then I planed the underside profile down to the pencil lines and chalk fitted to the inside of the top aiming to get the outside edge of the bass bar 1mm inside the outside edge of the bridge foot with a 2mm slant outwards from top to bottom.
After gluing in the bass bar I tuned the bar to match the top and then gradually worked both the top and back down in frequency to match each other and get as low a weight as possible (my client wants a light easy to play instrument) with the ring tones just below F.
So to the task of joining the violin plates from the bookmatched back and belly sets. First I squared up the beautiful maple blanks and planed the two edges together to ensure a good fit. This was so much easier to do compared with the cello as the pieces of wood are so much smaller/lighter and the join quite short. I heated the pieces gently with the air gun and then glued them together with a rubbed joint which worked really well first time.
Next I planed the underside surface flat before tracing the outline from the ribs and using a washer to give the external dimension including 2.5mm overhang.
I cut out both belly and backs on the bandsaw leaving a bit of waste to be cleaned up after shaping the outside profile.
This was so quick I ran the same process for the belly.
Carving the outside profile
Next I created a set of templates for both back and belly based on the profile curves published by Sergei Muratov (the same source I used for the cellos). I drew these up in solvespace and then stuck the full size printed drawings to a piece of perspex, cut out on the band saw and then finished on the vertical oscillating sander. I finished by spraying these with a bright paint – different colours for front and back so they didn’t get mixed up.
Then to carving starting with a 3/4″ no 5 sweep chisel to get the rough shape.
I cut a 10 mm wide platform at the edge with the router using the same attachment I made for the cello work.
These were cut at the finished edge thicknesses of 3.7 mm (back) and 4 mm (belly) and make it much easier to start the gouge cuts.
Then I moved onto using thumb planes.
Here I am getting pretty close to the final profile.
I tack assembled the back and belly to the sides to cut the side overhangs down to final size – 2.75mm except for the c-bouts at 3mm. Started with the knife and finished with files with a piece of waste of the correct thickness taped to the middle of the file.
Also cut the corner shapes with a knife.
Next job is to cut the purfling groove so I made up a simple holding jig for the body first.
Then I made up a laminate 1.4mm thick for the purfling of 0.3mm black (ebonised pear), 0.8mm white (sycamore) and another 0.3mm black. I sliced this up on the bandsaw into strips about 2.5mm wide.
Then I marked the purfling groove on the front and back with special gauge set in 4mm from the edge and deepened these faint marks with a scalpel lubricated with dry soap and carefully carved out the groove with a 1mm chisel.
Next I carefully fitted the purfling using the bending iron taking special care with the corners and glued up with hot glue.
Then trimmed back with a chisel when dry and gouged the channel round the edge about 1mm deep before fairing everything in with scrapers and checking the final arching.
There is a slight blemish in the maple on the back. I was hoping I would get through it with arching but there is a tiny thin bit left and I would rather keep the proper arching than keep on scraping away – it looks like bark but it can’t be as it was not visible before carving. Should dissappear when the back is coloured anyway.
I started by marking out the thickness profile contours and then drilled on the bench press to a few millimeters thicker than this to make the rough thicknessing quick and safe.
Then I roughed out with a large gouge.
Now I have to build a caliper/nail tool for marking the finished thickness before plate tuning.
And it works really well – I have made it big enough so I can do a cello with it.
Nearly down to the initial thicknesses before plate tuning.
I split some willow to make the blocks and set them at 32mm high at the bottom tapering to 30mm high at the neck and after getting them good and square, glued them into the form which I supported on a few pieces of plywood to get the form more or less into the centre of the blocks.
Then I used the perspex template to mark the outline on the blocks ready for carving.
When the glue was dry I carved the blocks down to the marked outline with a chisel and finished with a round file. The split willow carves very easily but because they are so small this is so much quicker to do on the violin compared to the cello – although you have to go slowly to get the accuracy required!
Preparing and fitting the violin ribs
Next I thinned the ribs down to 1mm with a combination of the drum sander and then scrapers. The figuring has to be matched up so the stripes all point upwards when looking at the back. Next the pieces were cut to size and carefully bent after damping in a wet tea towel and then finally glued to the blocks using pre-shaped cauls.
Then I planed the edges of the ribs down to make sure everything was perfectly flat.
Adding the Linings
I thinned the willow linings to 2.1 mm on the drum sander and cut them 10 mm wide on the band saw (aiming for 8 mm finished depth) to leave a few mm proud for trimming back flush to the ribs.
I was able to re-use the peg clamps I made for the cello and just substituted a thinner pivot piece to get a parallel clamp on the 1 mm ribs + 2.1 mm linings.
Now to move onto the back where I have some gorgeous maple to play with!
I had already decided to try my hand at making a violin after 2 successful cellos when someone in the orchestra actually commissioned me to make them a violin on the strength of my cellos!
I already had the Harry S. Wake book on violin making from the cello work but bought another book on amazon by Juliet Barker which I found I could not put down once started because it answered a whole load of questions that had been floating around my mind from the process of building the 2 cellos. Interestingly I also discovered that she only wrote the book because of encouragement from David Dyke who supplies me with my instrument wood! – small world.
As with the cellos I decided to draw my own plans – not because I think I can do better than Stradivarius – but more so that I understand the basics of how the shape can be produced, have something I can repeat without access to 3rd party plans and so I have a clear reference point if I need to adapt the design in future.
I used the solvespace parametric drawing package again referring to the Strad design in Harry Wake’s book for key dimensions. The result is within a gnat’s whisker of the strad outline but achieved very simply.
Here is the inside mold outline as a result and the solvespace file if anyone wanted to adapt it.
After cutting out the perspex and carefully sanding to get an exact profile, I used the template to mark out the outline of the inner mold on a piece of 18mm plywood. After careful sanding down to the exact dimensions, cutting out the recesses for the corner/end blocks and adding a few coats of varnish it looked like this:
Really exciting day – I just collected the wood from David Dyke down in Sussex to start on cello number 2.
Procuring the cello wood
Front and back plate preparation
And here are the belly and back book matched pairs getting glued up after spending hours getting a perfect joint – lovely flame on the maple back.
Managed to do the back without clamps but the spruce always seems to move a bit when heated up for gluing so I just needed a slight clamp pressure to get it tight.
The belly wood has a wonderful ring to it and great grain even and straight structure.
I also started to thin the maple ribs down to size on the drum sander but not to finished size so that I could scrape a finish on what will be the outside.
The cello ribs
Will have to scrape them next because I don’t want to see any sanding marks on the finished instrument.
And here are the blocks cut ready to to glue into the form. The end blocks are in spruce, corner blocks in willow as it is more forgiving to carve.
Note I have added some removable sections to the central form at top and bottom to make removal easier after gluing the linings.
Just need to carve the blocks to shape and chamfer the linings.
Carving the cello back plate
I built a simple jig to use with my small router so that I could cut the edges to the right thickness (5.5mm all around, 5mm in the C-bouts and 6mm at corners) with the top firmly clamped down onto the workbench.
This is an enormous improvement to the approach used on my first cello and as well giving a really consistent edge thickness, it gave me something to aim at when rough chiselling the back profile.
Edges cleaned up on the drum sander and slots for purfling marked up.
I used my new pillar drill to good purpose to drill out depth guide holes across the whole back (the old drill didn’t have a wide enough throat to reach the centre of the instrument).
Now to start the long process of thumb planing the inside to give the correct thickness profile and plate tuning to get the right mode shapes and frequencies.
Carving the cello front plate
Now to start on the spruce top. I got to this point in just a day.
Came across a resin inclusion when carving out the top. On scouring the internet this is apparently quite common in good tops (1 in 10 maybe) and no problem – just have to to shape a plug to fit the cleaned out hole.
Then a final bout of scraping to get a perfect profile before marking out the purfling channels and deepening with a scalpel.
Next step is to cut the channel and fair everything in before hogging out the back side of the plate.
Cello plate tuning
I thinned the plate until I got good clean mode 2 and 5 shapes but before going further on the top I have to cut the f-holes and fit the bass bar.
Tuning the bass bar
Started with a bass bar 11mm wide and initial depths set according to the following table:
Proportion of length
Then progressively reduced the depth to get the changes of mode shapes shown in the gallery below.
Reducing the ends helped to close the ring-mode shape but also reduced the x-mode frequency which I was generally trying to maintain. Reducing the centre of the bar predominately reduces the ring-mode frequency but I also had to thin the edges of the top and the centre of the upper bout between corners to try and get the ring-mode down to an octave above the x-mode.
A final tap tuning allowed to identify and remove any thick or high stiffness spots both on the top and the bass bar.
The final bass bar ended up only 18 mm high in the centre with X-mode of 58.2 Hz and ring-mode of 121.5 Hz. I then thinned the back to match those frequencies – much easier – and got 59.2 Hz and 120 Hz.
Making the box
Started by gluing the sides onto the back after first re-enforcing the sides with strips of linen.
The Cello Neck
The neck blank I cut back in March had settled a bit so I squared it up with the plane and then marked up and cut it out on the band-saw.
Next step is fitting the neck to the body.
I did a drawing of the fingerboard profile to give an even break over the strings at the bridge of 21 degrees and then an action over the end of the fingerboard ranging from 4 mm at the A-string to 6.5 mm at the C-string. This showed I needed a 2.4 degree slope on the neck towards the bass side. I have heard of 1.8 degrees being used so ended up going for 2 which still looks quite a slant!
Marked up the heel with this 2 degree slant narrowing the heel down to 27 mm at the bottom with enough material to give me a 22 mm overstand on the A-string side.
Before fitting the neck I will cut and shape the fingerboard and stick it in place temporarily to aid in getting the set of the neck right.
This consisted of cooking up Boiled linseed oil and rosin at 180C for a few hours in the ratio of 1 part linseed oil to 1.5 parts rosin by weight until the mixture achieved a sticky toffee like consistency when cooled. I used my trusty mini fat fryer to do this safely in a double boiler arrangement using cooking oil instead of water. I then added turpentine so that the resultant cooled varnish had a golden syrup like consistency.
Then I ground in 10g of Kaolin (china clay) to 30ml varnish with 7ml of rabbit skin glue and added water until I had a nice creamy consistency that could easily be applied with the fingers and then ragged off to give an even finish.
Then into the UV cabinet for a day.
I added a few coats of shellac to make sure it was all sealed properly, cut back with 600 grade wet and dry and then a final wiped on coat of shellac before applying the first colour coat and drying in the UV cabinet.
Next I mixed up some oil pigments – 1 squirt of burnt umber, 1 squirt of burnt sienna and, a bit less than half a squirt of alizarin crimson and a little boiled linseed oil and then smeared on by hand until the colour was the right depth and fairly even and then back into the UV cabinet.
Wasn’t entirely happy with that – a bit streaky on close up – so rubbed most of it off with a turps soaked cloth and put it back on with several much thinner layers so as to leave no marks.
Result has a nice warm glow and enough variation to look interesting!
It will darken up a bit more after another 3 or so colour coats.
One more colour coat and then a couple of clear coats before cutting back and polishing.
Curing outside in the sun on one of the few sunny weekend days this year! – Just like the old Cremona makers used to do.
It all looks a bit shiny at the moment but that will go once the finish it has been cut back – aim to get a nice satin sheen.
Clear coats next.
Applied 2 top coats and then cut the varnishing back with micromesh down to 4000 grit and then finished with rotten stone and linseed oil. The cello has a wonderfully rich chestnut sheen to it.
Glued the neck on after removing the keeper, and then fitted the pegs, saddle and endpin.
Just finished the instrument and delivered to a very happy customer (George) who christened it “Beryl” in time for our weekend concert with the Stratford Symphony Orchestra!
I just got back from an exciting afternoon with my good friend George Shilling at his Bank Cottage recording studio in the Cotswolds. George kindly spent some time recording my new cello “Anna” (after a visit to the pub).
You can find the resulting pictures and recordings here.