Newbie clavichord build

[Gregory Chandler]

Soundboard and Bridge Question

I had already purchased the soundboard wood from Alaska Woods when Alan posted so I accept his suggestion as an affirmation. The shipping was nearly as much as the wood so I purchased three sets. The two AA quality were $8.33 each and the AAA quality was $14.14. Shipping for all three was $15.28. I sent them a sketch of my soundboard and they were able to confirm I was choosing from the correct sizes. VERY prompt communication and service. I will start with the $8.33 board.

20190317_144855.jpg

First Question: I need to reduce the soundboard thickness from 4.5 to 5 mm to a final thickness of 3 mm (plus or minus .25 according to the measurements of the original) How do others do this?

__________________

I am about half way through my first attempt at the bridge. I used an awl to poke guide holes directly from a drawing copy glued to the wood stock. It may be as Andrew noted, I am using a historical instrument drawing, not a modern woodworkers plan. That and sometimes I literally miss the mark. I think I can do better but some advice from others can be helpful. I am not pleased with the bridge pin hole alignment. They wander instead of flowing smoothly along a line. The patch of sappy wood makes it look worse. I am considering making a jig for my second attempt. This bridge is straight for 2/3 and curved over the last 1/3. I am also considering doing most of the drilling and shaping then dry heat bend or steam bend the curved part.

20190317_145030.jpg

Second Question: What kind of techniques, sequencing, tools... do others use when making the bridge?

[Andrew Seemann]

Hi,
Soundboards were (and are) typically thinned with a smoothing plane. A standard Bailey's #4 works quite well. Make sure the iron is sharp, as spruce has a proclivity to tear. You could also random orbit sand one side flat (the top) and then just do the thinning from the bottom. Then you don't have to worry about surface quality as much. The edges need to be even (no ridges) so you get good glue joint between the board and liners. Don't over-thin the soundboard. If you leave it a little too thick, it may just not be as loud. If you go to thin, the string tension may collapse the soundboard. Soundboard thinning skill is one of those things that comes with experience, and distinguishes the professionals from the rest of us.

For the bridge, make sure to use a wood with nice even grain. Pearwood, cherry, and beech were popular historically, Pearwood does't exist as lumber in North America; beech is a little on the hard side but it works nicely. Cherry is probably the easiest to use as it is a little softer than beech, but mechanically and acoustically still works well. Your bridge doesn't have a lot of curve, so sawing out of a single piece or wood is probably fine, and likely what was done on the original (historical makers didn't tend to do anything the hard way). For marking the bridge, you probably could do a pinprick to mark the hole, and then adjust as needed. The important thing for bridge pins is that the string must only contact the pin before touching the top of the bridge, or else the note will sound false, due to the wood of the bridge damping the vibrations, rather than the vibrations being transferred through the bridge pin. You might need to move one of your marks to get the pin in proper position; that whole bit about the drawing being a guide

The glue joint between the bridge and soundboard is important to do well. Clavichord bridges can be under a fair amount of sheer, and the joint needs to be tight, strong, and not creep. Historic makers used hot hide glue, but that is inconvenient and takes a lot of practice and skill to get to work right, though in theory it is the best choice. Franklin's liquid hide glue will work as long as it is fresh, Titebond 1 and 2 would probably work fine as well. White glues (Elmers, etc) have too much creep. For clamping, there are various techniques, one is drilling small holes for brads, putting a piece of softer wood or plywood under the soundboard, gluing the bridge, and then driving small brads through the holes and soundboard into the wood/plywood below to clamp the bridge to the soundboard.

One bit of overall advice, don't be shy of remaking parts, other than the soundboard, most wood going into a clavichord, harpsichord, etc. is quite inexpensive and usually not worth salvaging something that isn't turning out right. On my spinet, I have remade every single piece at least once with the exception of the soundboard, including the entire case. The soundboard, I somehow got right the first time. The first case made a nice bonfire one evening. It is one of those things that after you make a part, then you know how you should have done it. Might as well take advantage of that knowledge then

[Gregory Chandler]

The full size drawing arrived from Scotland and I pinned it to the wall near the garage. My wife says I can’t hang it in the living room. I can’t overstate how much easier working off a full size drawing is. The full size drawing also exposed some inaccuracies in measurements that are now glued firmly in place.
20190420_130855.jpg




The bridge was difficult for me. On the seventh attempt I had finally discovered a path, process and technique that addressed various failures on the previous six bridges. It still has flaws. I am pleased with the profile, it closely matches the drawing, the pins are aligned nicely except the one at the bend is tilted a bit out of vertical.
20190420_124854.jpg


To position the bridge onto the soundboard, I made a measuring stick. Using the lowest bass hitch pin as a reference (and the full size drawing) I marked the stick where a sampling of the bridge pins should be located. I did this twice. The first time was before installing the bridge pins or gluing the bridge. The second time was after gluing the bridge, after installing the bridge pins, and in preparation for the final placement of the soundboard.
20190420_121344.jpg




I pre drilled the wrest pin block holes using a drill press before gluing the pin block into place. (I glued the wrest pin block to the frame before installing the soundboard) After drilling the pilot holes, I removed a thin slice from the pin block and am using that as a guide to drill the holes through the soundboard with a cordless drill. Because the pin block is tight against the right side case wall, I should be very near dead center.
20190423_225614.jpg


The soundboard mouldings are 6 mm x 2.5 mm. I used the band saw to rip strips, then I made a jig to hold the strips and used a hand plane to produce a group that were equal width and thickness.
20190423_204621.jpg


I used a shooting board to cut the miters. These are so small that the table saw is not the best tool. (See previous post about table saw and fingers!)
20190423_213208.jpg


I used some small sticks to apply pressure to the moulding while the glue dries.
20190423_235341.jpg




The drawing specifies oak for the balance rail. On my first attempt, I made a cutting error that I was able to patch and there was a small section of tear out on the back side of the rail. It actually looked pretty nice and the imperfections were the kind that casual observers would not notice. But I knew they were there. I had a scrap piece of 6/4 hickory that allowed a no cost opportunity for no regrets. Using the oak rail as my guide (it fit perfectly) I was able to make the hickory version without taking a single measurement. Having already worked out the various saw blade angles, it was fast, easy and fun making something with measuring.
20190428_110034.jpg

[Gregory Chandler]

It has been nearly 3 months since my last post. Partly due to the demands of spring and summer yard work, and partly due to not knowing how to accomplish a step, the progress has been slow. I haven't given up. I will attach a few photos to show you where I am as of today.
20190430_214110.jpg
I glued up a basswood panel for the keys.

20190501_200820.jpg
I made a template out of scrap to work out angles, spaces and achieve a good fit for my keywell.

20190501_215710.jpg
I used the template to trace an outline, then cut out the keyboard blank.

20190505_125412.jpg
I marked the locations of the key guide slots on the rack.

20190509_222654.jpg
I cut the slots with the bandsaw.

20190521_171509_HDR.jpg
After marking the positions of the balance pin holes on the key blank, I wedged it firmly in place and drilled through the key blank into the balance rail. I also marked the position of the key guide tongues onto the key blank while everything was wedged tight.

20190630_222120_HDR.jpg
I attached the veneer cap to the rack. Marked the locations of the rear hitch pins. Set the drill press at a 15 degree angle to drill the holes for the pins. I used a dremel tool to cut off the pins then finished them with a small file and small fine sanding block.

20190705_204919_HDR#1.jpg
Sorry the photo is rotated... I glued the rack to the case.

[Gregory Chandler]

20190707_202219_HDR.jpg
I set the balance pins.

20190713_201049_HDR.jpg
Key lever pivot pads. I used 1/8 inch square leather shoelace.

IMG_20190609_192027_158.jpg
I bought a fancy fret saw and had a proud moment when I cut out my first key. The feeling of pride turned to despair when I discovered I did not have the skill to accurately cut the keys by hand. I botched the next two keys then stopped.

IMG_20190617_235320_618.jpg
I had an old scroll saw in the barn my dad had given me about 15 or 20 years ago. I had never used it and it was in poor condition. I cleaned it up, put a new cord on the motor, and put one of the fret saw blades in. It worked! (photo was from instagram) Time to cut some keys.

20190721_212807.jpg
I have been cutting the keys out in small groups. After each group is cut, I insert the key guide and elongate the pivot hole. Some adjustments are generally necessary to allow the key to travel up and down freely. I make these adjustments with each key in isolation. After all the keys are rough cut, I will attach the key tops and begin carving.

[Alan Caro]

Gregory Chandler,

Congratulations on the progress and quality of work on the clavichord project. It takes very good visualization skills to create a keyboard instruments from plans especially without having to spent a lot of time with actual examples.

The keyboard are particularly difficult to have operate smoothly, with quick response, fast return, and importantly, that are quiet. One of the other unique qualities of clavichord is the string gauges require experimentation as the relative springiness is important to the action; if the string is too taught, the tangent will bounce off, too loose and it will fell soft and the pitch will rise. Moving the point at which the string gauge changes is a possibility especially if the string gauge was specified for A-415 and the new instrument is using a modern A=435.

Besides the fraction of a mm placement of the tangent necessary on a fretted instrument, the distance between the tangent and string is also important; too loose and it's too quiet, too far and it's difficult to play easily.

Tangent placement is critical as it on a fretted instrument it will vary according to the temperament. Many 17th C. clavichords will have been meantone, whereas later 18th C. will have a version of equal temperament. For the tangent placement, mount the tangent of the lowest note of each string pair to it's plan location, and with a tuner, hold the next higher note's tangent in place according to plan, set the tuner for that note, pluck the string and move the tangent to where it's in tune in the temperament you're using. It's possible to bend the tangents some to tune the fretted notes, but if they're bent too far, the top surface is not in full contact to the string and it can sound a bit weaker or can buzz.

Thanks again for the progress reports and I look forward to the next installment.

Alan

[Gregory Chandler]

Alan,

Thank you for all the information. I have a few questions...

How far does a key travel before the tangent makes contact with the string?

I assume I should set the tangents at a depth so every key has the same amount of travel?
20190731_010855.jpg
My plan has a chart that shows that the distance from the levers to the tops of the tangents vary by 3mm from the bass section to the treble.

The plan also shows the distance from the top of a representative tangent to the string as 6mm, but does that distance vary?

I probably have a month of more of carving and key detailing before I am ready for stringing but would you recommend starting at the treble end and do one string pair at a time with the associated tangents before moving on to the next string pair? Or would you do all the stringing then work on the tangents?

I have never seen (other than photos) or touched a clavichord, and my drawings are just drawings, not instructions. The suggestions and insights you and Andrew have shared are so helpful. Thank you!

Greg

[Alan Caro]

Greg,

The distance from the tangent to string is progressive, generally being closer in the bass and farther in the treble. This is another indicator of the role that the string tension plays in the action. The long and often brass strings in the bass are not as taught as the treble strings, so when pressing the key, the looser tension makes the apparent key depth larger and it feels less responsive. This happens on harpsichords too, but harpsichords are generally scaled further down in the bass : (that is the doubling of string length for each octave reaches a but lower and as the key pressure on harpsichords does not affect the volume of sound produced, it's not as noticeable. I'd suggest following the plan notes- that seems about right. I've played quite a few modern clavichords that increased the tangent to string distance up to 9-12mm to try and increase the volume, but I find them difficult to play.

The feel of the action does take quite a lot of getting used to. No other keyboard instrument has that sensation. And of course, the quite amazing proportional control over the volume and the ability to produce vibrato and bend notes for effect isn't found on any other stringed keyboard instrument. This is the reason that learning the clavichord was so highly prized as it required an evenness of touch pressure not required on an organ or harpsichord. Clavichords are really amazingly elegant in terms of effective simplicity, considering the expressive control of a simple keylever on a balance rail with a vertical blade of brass as compared to up to 130+ parts in one note of a grand piano action.

Not a lot of sound, but one quickly tunes into it. I attended a recital series with 120-150 people in a 12th C. church in northern Italy and the requirement to not make the slightest sound really drew the audience into participation in the event. Clavichords have an odd introspective and calming effect.

In my view, recordings can help provide a further insight into the appropriate sound of a less familiar instrument like a clavichord.

I recommend, particularly, this one of music by Froberger, performed by Thurston Dart:

https://www.youtube.com/watch?v=wILVMuon6r4&t=595s

Very expressive music that at the time could have only been realized on clavichord.

And Dart's Bach French suites, here introduced by Dart:

https://www.youtube.com/watch?v=pzI7rK7hwNc&t=858s

Christopher Hogwood also recorded quite a lot on clavichord:

https://www.youtube.com/watch?v=gcyPrOVYRVk

Igor Kipnis, who I met in Princeton in 1973, and interviewed on the radio in 2000, recorded quite a bit on clavichord in addition to his usual harpsichord:

https://www.youtube.com/watch?v=e-bV-Sm4aWs

> listen at 27:45

A lot of CPE Bach seems very suited to the clavichord and CPE was a promoter of it for learning:

https://www.youtube.com/watch?v=1zOt-oBekls

The clavichord videos by Win Winters- which include surprisingly successful selections of Beethoven (Beethoven did mention having a clavichord and there's is one in the museum of the Conservatoire de Paris that I've seen, claimed, probably spuriously to have been his) and


What are your plans for the case finish?

Alan