My good friend, recording engineer and cellist George Shilling was recently involved with Tony Iommi’s latest creative work with the Birmingham Gospel Choir called “How good it is” where George and Anna provided the cello accompaniment.
Birmingham post article complete with video is here
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.
Having completed the rather rewarding finishing process I can now set up the instrument ready for stringing.
The first step was to build my own peg shaver as the cheap Chinese one I had bought together with a peg reamer and end-pin reamer was way too small for the pegs I have.
I found a piece of maple off-cut and cut it down to the length of a peg and drilled it slightly larger than the thin end of the peg reamer before reaming out a hole with the thick end matching the thick end of my pegs (about 14mm). Then I planed away one face until I had a nice parallel slot showing about 5mm wide. Then I glued another capping piece over the slot and re- reamed the hole. Finally I fitted the blade from my Chinese shaver together with a very simple fine adjusting mechanism.
The result is shown below.
I also built a very simple jig from 2 pieces of hinged oak to sand a finish onto the pegs.
Then I reamed out the pre-drilled holes in the peg box (using a few reverse turns at the end) so that I had a little more than the target 21mm between the edge of the box and the decorative rim of the peg to allow for the wood to compress a bit.
I then set up a simple jig to hold the pegs while I cut them to length and drilled them. I marked the hole positions a few mm from the centre of the box towards the thicker end of the peg and drilled the holes using a 2mm bit and then chamfered the openings.
I rough cut a 13mm square sction from a piece of close and straight grained spruce taking care to ensure no runout and then planed it down to an 11mm square section. Then I carefully planed it down systematically to an octagon and so on till I had a perfect cylinder 11mm diameter.
Actually fitting it could be quite an adventure with the fitting tool and retriever I purchased!
Yes – quite an adventure but eventually got the knack – certainly got my money’s worth from the sound post retriever! The f-hole was only just wide enough to fit the post through though – must remember to make it a millimetre wider on the next one – took me 2 hrs of frustrated attempts! Phew.
I filed down the ebony nut blank to fit the fingerboard profile and also added a 2mm T-extension as some makers have suggested this improves the look and I needed it anyway to get back to the designed dimensions. I tack glued the nut to the end of the fingerboard with a few drops of Titebond.
Nut ended up proud of the fingerboard by 1mm at the A string to about 1.4mm at the C string.
String spacing was set with dividers at 8mm and slots cut with a fine saw and then nut files before lubricating the slots with graphite (pencil lead).
Filed down the ebony saddle blank to match the curve on the bottom of the cello and marked /cut the top with a sharp knife to fit. Glued in with hot hide glue
I built myself a simple jig to hold the bridge during fitting:
I used a piece of carbon paper to highlight the area on the feet to remove.
Then marked the bridge height using a marked up stick on the fingerboard to give 4mm action clearance over fingerboard for the A, up to 6.5mm for the C string and cut the bridge profile on the band saw giving it a few millimetres extra to allow for final adjustments.
Next I used a small plane to get the thickness profile right before hand carving with a sharp knife to get the bridge to target dimensions. Finally hand sanded to get rid of any tool marks and marked the string positions with dividers (15.8mm string centre to centre) and cut the slots with fine saw and then nut files. Ordered some parchment to protect the A and D slots.
Finally got to the exciting bit of fitting the strings and seeing what the result of 9 months labour sounds like!
Held the bridge in place with the bridge holding jig I made earlier while I fitted the tailpiece and started stringing it all up with a set of large scale. Everything seems to work ok – need to do a bit more on the pegs to get a more even grab between the two ends and the bridge needs a bit of tweaking to bring the action down to target and also a coat of linseed oil and the parchment protectors for the A and D strings – but it sounds wonderful! Very responsive, lovely sweet tone, good depth and well balanced across the strings and right to the top of the fingerboard. Quite a strong wolf note on F but a 7g eliminator on the G string sorted that out.
Will be taking it to our cello-nite on Saturday night to try out on the felli celli!
Cant wait to see how the sound develops as it gets played in.
I have decided to call her ‘Anna’ – I think she deserves a name after all that work!
Link to all the specifications for the finished instrument are here.
Link to some better pictures and recordings made with it by my good friend George Shilling.
After much reading on the Internet I decided to go for the following approach:
I found an excellent supplier of ready made varnishes and varnish ingredients in Germany at Hammerl and ordered the oil varnish and a few extra ingredients.
I then set about designing a UV light box for both giving the white cello an initial suntan and for drying the oil varnish coats.
I settled on a design that I could assemble and disassemble easily for flat storage when not in use.
The base is 1/2″ ply 610mm x 400mm into which I routed some 6mm grooves to take the hardboard sides, front and back. The sides are braced with vertical triangular cross section struts at the edges and onto which are mounting UV fluorescent fittings (4 off 18W 600mm UVA fittings on each side).
The box is 1500mm high and a top piece of similarly grooved 1/2″ ply and with a small USB fan to extract heat/varnish fumes holds it all together.
All the inside surfaces are coated with aluminium foil.
I lightly glued a piece of scrap on the neck to hold it straight and clean during tanning/varnishing. Then the cello got a final wipe down with water and a final scraping all over before going into the light box for a week or two depending on how fast it darkens up.
Here is the result after 10 days.
I then tried fuming using household ammonia (only about 9% but seems to work quite well). Here is the result after 10hrs.
Next I sealed it all with a few coats of a 1lb cut of shellac.
And then the 1st coat of coloured varnish. This is incredibly thin and only a tiny amount was enough for the whole cello.
Then back into the light box to dry.
Once dry (I left it 2 days) I rubbed down with wet 2400 grade micro-mesh and then used a mixture of burnt umber and burnt sienna artist colours rubbed in with my fingers and wiped off with a kitchen towel until the colouring was even.
Another 2 coloured coats but still looking too light so had another go smutzing with artists colours.
Next step is a final coloured coat and then 2 clear coats.
Next I cut it back with 3200 grade micro-mesh and water and then pumice and linseed oil on a felt pad. Still left some corduroy grain on the top. Finally polished to a nice matt sheen with rotten-stone and linseed oil.
Then popped off the keeper on the neck before gluing back the fingerboard.
The big moment has finally arrived and time to glue on the back.
I modified the form so I could dismantle the ends to allow me to keep the form in place while the back was glued on.
Before fitting the belly I reamed out the end pin making sure the pin stayed centrally aligned inside the body and leaving 1mm gap between the pin body and the ribs.
Then I fitted steel pins to the belly where it meets the end blocks to make sure everything stayed aligned during gluing.
Heated everything up with my paint stripper air gun and then some weak hide glue and all came together nicely.
One seam had a small gap so popped it and re-glued.
Then I used a knife to round the edges and finished off with 400 grit sandpaper.
I marked the side of the neck where it should meet the top (280mm from the nut) and with an upstand of about 17mm on the bass side and 19mm on the treble side (due to the 1.5 degree cant on the fingerboard).
Then marked the mortice position with a knife and carefully sawed out the sides of the mortice and chiselled out the waste. When I got close moved to a file and chalk to get a snug fit with the fingerboard centred over the body and projecting a height of 85mm at the bridge position. This was definitely one of the trickier operations but I kept moving slowly with lots of trial fits and got there in the end.
Here it is assembled before gluing:
And after gluing ready for finishing.
Before I did anything I roughly cut out the neck profile to give the wood a few months to settle.
Now with the plates nearly done I levelled the top and squared the sides (they had actually moved appreciably in the intervening 4 months) before cutting the profile more accurately on the bandsaw. I then filed/scraped the outer scroll outline and drilling the peg holes on the drill press before sawing the scroll sections and then started carving.
After much filing, carving and scraping I got here:
Next I rough carved the fluting with chisels
I bought a pre-shaped blank which reduced the amount of shaping required substantially. First back to the drawing board to get the profile right. I started with the string profile over the end of the fingerboard with a 20 degree break between strings and string heights over the fingerboard from 4mm (A string) to 6.5mm (C string) as I am going to use Larsen steel strings on this.
Fingerboard length will be 5/6 of string length = 580mm but added a few millimetres for luck.
Then extrapolated to get the bridge profile.
Plans are here: fingerboard profile
1st job was to plane the edges straight and take the twist out of the blank.
Then made a template for the fingerboard profile at the end and started planing away with the jack plane to get a rough shape. It is quite hard to check for flatness using a straight edge so I just laid it on a flat surface with a light behind and looked underneath for the dark spots.
Once the rough shape was there I used a scraper to fine tune and get the 1/2 string diameter scoop measured with feelers.
Ended up with 8mm sides with a slight scoop on the edges to keep the wall height in the middle.
After cutting the outline of the neck root ready for fitting to the body I tacked on the fingerboard with a few dabs of weak glue – just hope I will be able to break the joint later!
After fitting to the body I carved and filed the final neck profile and finished off the chamfers on the scroll. Note the chalk on the neck joint from fitting to the body.
I had to buy a cheap 25W amp and a hefty 8″ driver unit to excite the resonant frequencies in the back and top.
I then used a free app on my Android as a variable frequency sine wave signal source and tea leaves to see the patterns!
There is a lot about plate tuning on the Internet with several systems used. I found two particularly useful articles:
Here is the mode #5 (bell-mode) on the back.
Here is the mode #5 (bell-mode) on the front
and mode #2 (x-mode) on the front
The back is still a little high compared to the top but I will come back to that once I have cut the f-holes and fitted and tuned the bass bar.
Having cut the f-holes and glued the bass bar I then re tuned the top to bring the bass bar into balance.
Both Modes #2 and #5 had gone up with the bass bar fitted.
This is how the mode #5 shapes developed as I cut down the bass bar:
I must admit I wasn’t sure at first that this was actually #5 as it looked pretty weird!
And the mode #2 looked like this:
At this stage I had got the mode 5’s on back and front equal at 130Hz and the mode #2s to 60Hz on the front and 64Hz on the back so a little more to do still.
Final adjustments made.
| Mass | Mode #1 | Mode #2 | Mode #5 | |
| Back | 669g | 30Hz | 57.6Hz | 122Hz |
|
|
|
|||
| Belly | 490g (without bass bar) | 28.4Hz | 58.4Hz | 120Hz |
|
|
|
Having finally finished the back it is time to move onto the top.
First job was to glue the two book-matched halves together as on the back and then levelling the assembly and marking and cutting out the outline.
After building two profile templates (one longitudinal and one transversal at the centre line) I power planed down to the rough centreline thickness and started with the thumb plane. I can’t believe how much easier the spruce is to work compared with the maple on the back. In just a couple of hours the shape is nearly there!
And then mark and cut the purfling slots before gluing in the purfling and cutting the channels.
Then shape the inside after drilling some guide holes.
Plate tuning of back and top is next before cutting f holes and fitting bass bar.
The f-holes were marked by pinning the plan to the back of the belly and pricking through with a knife before joining up the marks with a pencil and adjusting the shapes by eye.
The holes were then cut with undersized forstener bits and reamed out to the correct size (14mm and 18mm) with my peghole reamer and my endpin reamer.
F-holes then roughly cut out with a coping saw.
And then carved to size and matched using a very sharp knife and a scalpel. I undercut the holes so they were bigger on the inside than the outside.
Then I cut the vees and glued on the bass bar.
I have 2 beautiful pieces of book matched maple to make up the back.
I started by getting my jack plane (from Axminster) into good working order. This had never worked particularly well and after closer inspection I realised why – the blade was slightly bowed and so the burr did not rub off properly. Well after an hour or so of flattening against my diamond sharpening (stone?) I managed to get it flat and properly sharp and then joy, – producing cigarette paper thin shavings and both halves flat and true in no time.
I used the heat gun to warm everything up, made a new batch of glue and tried a rubbed joint – worked first time and no clamps!
Next was levelling the underside – boy is that maple tough but after about 3 hrs with the jack plane and a few blisters got it done.
Then clamped the ribs to the flat surface to mark out the outline and used a washer to mark the overhang before cutting out on the bandsaw and finishing off with the vertical spindle sander.
Next marked up the thickness of the edge – 5.5mm except for the corners (6mm) and the c-bouts (5mm) before starting all that chiseling! Boy, is that maple hard work!
I built some thin plywood templates to help with the arching profile. There is a great article by Sergei Muratov that gives a good starting point for the arching curves. I adapted these slightly to give an arch height of 25mm on the back.
I used a power plane to get a rough arch height and then a gouge, small plane, and thumb plane.
Finally finished profiling the outside of the back – this is a long week! For anyone actually reading this blog you will by now have realised that my weeks are “virtual” this week started at the beginning of June and now July is nearly over! This is the result of actually having to “work to live” and in a job that has me trotting round the globe regularly so this virtual week is so far 7 regular weeks long and have still to shape the inside of the back!
Now for the purfling – I made up the purfling from a 1mm piece of sycamore sandwiched between two 0.5mm pieces of ebonised pear all glued up (hide glue of course) and clamped between 2 cauls. Then sliced it up on the band saw into 3mm strips and after a bit of cleaning up with a plane all ready to go.
I managed to find a tool for marking the purfling groove on ebay but had to grind the blades down a bit to get them 2mm apart (the width of my purfling) and ground down a very small chisel (I had in my toolbox) to be slightly under 2mm.
Glued in purfling using a syringe to fill the pre-heated channel with hot hide glue, tap down with a mallet onto kitchen roll to stop it spraying everywhere and clean up with hot damp towel.
After gluing all the pieces of bent purfling having taken care to try and get “bee sting” at the corners I dug out a channel over the purfling with a chisel and then faired the channel into the body with scrapers.
Now I turn the back over and marked out the desired thickness profile in pencil.
I built a simple rounded stop to put under the drillpress and then drilled holes (1mm over-thickness) all over the back to provide a guideline when hogging out. Then using a fearsome wood carving attachment to the angle grinder (from Axminster) which cuts through maple like butter, I rough hogged out the back to the bottom of the previously drilled holes.
Then it was back to the thumb plane and more blisters to take it down to final thickness prior plate tuning.
Having finished the form the next step is to cut and glue the blocks for the top, bottom and wings.
I cut these from a piece of spruce, cut to length to fit the tapered ribs (120mm at bottom to 116mm at top), flattened the ends and started playing with hide glue for the first time.
Looked at the electrically heated glue pots available for an astronomic price (>£150) and then found a perfect alternative – mini deep fat fryer for £12 from Amazon (thermostatically controlled). Filled with water – not oil of course – and did a quick experiment to find the setting that gives 135F and then full steam ahead (not literally) with the glue in a jamjar inside the fryer . Anyway, works an absolute treat and had all 6 blocks glued to the centre layer of the form in no time.
Also bought a heatgun from Screwfix for £10 for undoing things when I got it wrong and got to try it out immediately!
Next the blocks were marked with the desired profile and carved/filed down to size.
I had some nicely figured maple rib blanks that came about 4mm thick so I started by thicknessing to about 2mm on the drum sander before bandsawing to width (with a few mm spare), scraping the outside to give a good smooth finish and then final thicknessing (the inside) on the sander to 1.5mm.
Then I used my bending iron to shape the C bouts after giving them a thorough dampening with a water spray.
Before gluing up needed to make sure the ribs had thoroughly dried out and shrunk back to their dry size in order to avoid any subsequent cracking.
This was not quite as easy as it looks – as soon as you put a clamp on they move! Had to take a few goes at this with several clamps used to hold it in place before squeezing up the pre-shaped cauls. Also worked much better when both surfaces were pre-sized with glue and pre-heated with a hot air gun.
I made up some cauls to hold the ribs against the form to stop things moving around while the ribs were being clamped up.
After levelling off the ribs I removed the top layer of the form to give room for fitting the linings.
The linings were cut from willow sanded down to 3mm thick and cut into 20mm widths on the bandsaw. These bent really easily after spraying with water. Mortices were cut into the corner blocks with a saw and then finished off with a small 2mm chisel and cleaned up with a knife.
I read on the internet how to make some very simple but effective clamps using rubber bands cut from an old bike wheel inner tube.
