Category: Xylophone

Xylophone Tuning (Job No: 1294)

On the surface, many aspects of my job seem to be structural or mechanical, which they are, but the solutions are always to improve the musical aspect of the instrument.  Aside from increasing its lifespan, which is almost a byproduct, the main aim is to improve playability, projection, tonality, etc.  Right at the very heart of this is the tuning.  I have been working on and off on a series of articles to explain the various aspects of temprements and tunings, and they will eventually be completed, but what I realised was that I very rarely write blog posts about this aspect of what I do.

Sometimes my self protectionism manifests itself inadvertantly with a reluctance or resistance to publicise what I am doing and how I go about jobs.  This is irrational, because the reality is that even if I were to tell someone exactly how to do something, they can’t necessarily do it.  I have experienced this many times when I am on the phone to a customer talking them through a job, or even face to face teaching my past assistants – there is an empirical feel that simply cannot be taught it has to be learned through experience.  Tuning is the same, in that there is an empirical understanding of what to do.  There are plenty of misleading internet guides on how to make your own marimba or xylophone, and if you have the time, energy and don’t mind making something shit, then they are great.  However to expect an app on a mobile phone to be capable of tuning percussion instruments accurately would be naive.

So anyway, all the notes are unstrung and arranged chromatically.  I have a bench that houses trays which contain a complete instruments note bars which can be slid out to be worked on, or be put away until the following day.  It keeps everything together so that I don’t mix up the instruments.  The work bench has a top and two shelves so that when I work on a set of notes the octaves can be separated vertically up the bench.  This means that I can tune any of the octaves chromatically, or do all the Cs, all the Ds, etc in a compact space, because I do jump around selecting various notes to do side by side comparisons.

When I start tuning a set of notes I am listening to the notes with my ears as well using electronic tuners. The tuners enable me to tune the bar to exactly the pitch I want. Qualifying what I mean by exact, the tuners I use are accurate to 10th of a cent, and a cent is a 100th of a semitone, so in other words ridiculously accurate. In practice, for wooden note bars, the tuning tolerance can be greater, so I tune to within a cent of the pitch I want.

On the first day I am listening with my ears more in order to identify any issues and inconsistencies. Sometimes there is a suppressed feel to the way an individual bar sounds, and often there are slightly discordant harmonics. Additionally xylophones in particular are prone to edge damage, so all of these factors help determine how I go about removing material from the bar. Since removing material is the only way I have of manipulating the tuning, this first day takes the longest because if I get it wrong, I can’t put the material back onto the bar.


Over the following week I return to the notes each day, making smaller and smaller adjustments until they have stabilised, at which point I seal the underside of the notes with lacquer, restring them and return them to the instrument frame or the customer.

LP Xylophone Notes (Job No: 1242)

wpid-screenshot_2015-01-30-08-17-382.jpg.jpeg

Out of the whole family of keyboard percussion instruments, xylophones suffer the most with going out of tune.  There are two reasons, first the note bars seasoning, second from being played.

Due to atmospheric changes wooden note bars absorb and release moisture, as the moisture leaves the wood, it takes a bit of matter with it, so the note will always go out of tune even if the instrument is not played.  However this is more evident in marimba bars which are wider, thinner and more extensively arched, whereas xylophone bars are more chunky and tuned to fifths, or should be, but more of that later.

When xylos are played, hard beaters are used, and these damage the surface of the note bar leaving indentations.  Therefore nothing harder than the note bar should be used to play it, so no hard plastic beaters if the piece has lots of fortississimo.  Furthermore the edges of the bars take a battering, and they become frayed.  When the bars are tuned, I have to remove this loose material, which further affects the tuning.  Eventually I run out of wood to tune the bars, or the internal structure of the wood has been softened or split thus killing the sound, so new bars need to be bought or made.

One last bit to finish off, there is a trend now in new instruments towards thinner bars which have been octave tuned as well as using cheaper woods like padouk.  Wrong, wrong, wrong!  Thinner bars don’t give the stacato sound of a xylophone, octave tuning just sounds wrong – a xylophone has a jarring sound, that is what it is for musically, and padouk is even more prone to damage and splitting.  So not only are they not xylophones, but more like piccolo marimbas, they will not last very long – years instead of decades!  Finally, because the bars are so thin, my job of retuning them is severely limited.

So lets get into the job.  First I get a plank of Honduras Rosewood that will do the job.

wpid-wp-1422608145557.jpeg

This is then ripped to give me the width.  I got two lengths out of this plank which is enough to do all the notes.

wpid-wp-1422608240959.jpeg

These batons are then cut down into the lengths required for the note bars, resulting in lots of sticks of wood.   I then plane to faces to get them flat and square.

wpid-wp-1422608379040.jpeg

I can then thickness them to the correct width and depth.  I now have a pile of equally sized sticks of wood.

wpid-wp-1422608526871.jpeg

The next job is to drill the holes for the note cord.  It takes ages to set the drill up so the notes are held in the correct place, and flat on the bed of the vice.  Getting the angle right is the easy bit (if you have the right equipment).  However once everything is set it is quick to drill all the sticks.

wpid-wp-1422608572835.jpeg

I can now put the radius on the edges of the bars.

wpid-wp-1422608880353.jpeg

Now I finally have note bar blanks which can be varnished and tuned.

Premier 600 Xylo (part 2 of 2) (Job No: 1187)

This blog post follows on from 1187: Premier 600 Xylo (pt 1)

After all the xylo notes have been varnished, they get tuned and resealed.

The last part of the job is to service the frame.  The note pegs on these Premier xylophones are solid rubber mouldings.  Because they are natural rubber they do age and become brittle, this process can be slowed by removing UV exposure.  Just like people do to prevent sunburn, this can be achieved simply by covering up with a blanket.

I do have a diminishing supply of these note pegs; like most parts I have for obsolete instruments, I strictly control the distribution.  The last few are to be used to repair as many instruments as possible.  They are not to sit on someone else’s shelf just in case they need them for their particular instrument in the future – I deem that selfish, and it won’t happen on my watch!

Premier offered me the moulds to make these parts, however besides storage problems (they are massively heavy), the cost of making the parts was prohibitive.  Essentially the moulds are too old compared to modern techniques.  At some point however, (when funds allow,) I will invest in the equipment I need to make alternative spares using different materials.

Getting back to the Xylophone; besides a few broken note pegs, the frame was in excellent condition, and just needed a thorough clean.  Once the notes were back on, it looks like a new instrument.
wpid-wp-1415005454153.jpeg

Vintage Premier Xylo (part 2 of 2) (Job No: 1188)

The first part of this blob post is Vintage Premier Xylo (part 1)

After the note bed is completed, I make the frame that the instrument sits on.

The first thing I do is make the top and bottom transoms, which are either end of the legs.  The top transom is a known length as it is defined by the note bed, the only decision to make is where to put the hinges.

The bottom transoms have the casters attached.  I make these 10mm longer than the upper transoms at the low end of the xylo, so they are 5mm wider than the instrument on either side.  This is so that it is the very bottom of the instrument that hits a wall or is a positive contact point for tying into a van.  This is also when I calculate the caster swing, and the bottom bar fixing points, and decide on the width of the high end legs.

After the bottom transoms and fixings points are made, I can join them by adding the bottom bar because I already know the length of the instrument.  The only decisions is where to weld the bottom bar in the horizontal plane, front to back.

The leg length is a matter of mathematics – I have been told how high the customer wants the instrument, so I make it to the correct height.

So now I have a note bed, two sets of legs and a bar for the bottom, in other words a complete square, I can assemble the instrument, and put in a brace to keep it square.  Depending on the instrument, I use one or two braces.

All the metal work can now be sent off to be powder coated.

In the interim, I finish the note bed, by putting in the note pegs.
wpid-wp-1412497901333.jpeg

And the original badge.
wpid-wp-1412497969607.jpeg

Once all the bits are back from painting, I clean everything, and put it all together and put the notes on.  Below you can see the figure in the end boards.
wpid-wp-1412498076674.jpeg

Finally below is the finished xylophone.
wpid-wp-1412498173634.jpeg

Vintage Premier Xylo (part 1 of 2) (Job No: 1188)

It is always nice to receive a pile of bits and asked to turn it into a usable instrument.

wpid-img_20140911_104959.jpg

The xylo is actually complete, including the frame, however requirements and expectations have moved on somewhat since this was made.

The first thing I need to do is get the note bed repaired.  All the joints are loose, and it looks horrible, so I will disassemble it down to components, and strip the finish off.

wpid-img_20140911_122118.jpg

wpid-img_20140911_141238.jpg

In the first image, it can be seen that the notes originally sat on felt strips with the note cord running along the outside of separation pins.  This is rubbish, it’s going to be really noisy, so I will substitute the pins for note pegs which means drilling bigger holes.

The consequence of this is that the rails need to be moved further apart, because the note cord now runs down the centre of the note rails as opposed to the outside edges.  The benefit is that the instrument will be that little bit wider, which is no bad thing.

wpid-img_20140912_194610.jpg

Above you can see new end blocks between the note rails which have been glued and pegged in place.  I don’t often use nails or screws when making a frame.

wpid-img_20140912_19461022.jpg.jpeg

Zooming in on the photo, it can be seen that the note rails are a bit arbitrary in length. There are several reasons, but essentially it is because it was made wrong in the first instance  I have tried to reach the best positioning of the notes, so they run parallel to the ends.  The ends need to be parallel so I can make the base frame square.

wpid-img_20140913_170443.jpg

Underneath all the clamps, I have added onto the outside of the instrument a spacer and and end plates.  The end plates will make it look nice, and protect the notes, and cover the top of the legs.  The spacer is just another bit of wood (the offcut from the board after the end plates were cut out), but it will give me something to work with when I make the base frame.
wpid-img_20140913_191256.jpg

I mixed up some stain to colour the end plates to somewhere near the colour of the note bars, so that the instrument will hang together aesthetically.  It will also bring out the figure in the wood.  It is now ready to be varnished whilst I make the base frame.

The second half of this project in in Premier Vintage Xylo (part 2)

Premier 600 Xylophone (Job No: 1187)

I have an old Premier xylo in for refurbishment.
First job is to refinish the note bars which have water damage in the varnish. So the bars are stripped of all the old varnish

wpid-img_20140908_125911.jpg

Day 1: There are three ways to remove varnish: heat, abrasives, chemicals. Heat poses a high risk of scorching the wood. Abrasives alone get clogged up, so a lot of material expense unless very coarse abrasives are used, but they have too much effect on the dimensions of the note bar and thus dramatically affect the tuning. My preferred method is a paint stripper, they are pretty much all water based now. It is a sticky messy job, and time consuming, but effective. After scraping off as much varnish as possible, the note is washed in soapy water to neutralise the paint stripper, then left over night to dry.
wpid-img_20140909_100143.jpg

Day 2: Start sanding, and keep sanding way past the blister stage, way past the point of boredom. This is the tedious part, concentration is needed to remove the last bits of varnish, then they need to be cleaned up of the whitish residue, and any surface imperfections removed. I do this all by hand, its the only way. Any light weight sanding machine (orbital or palm sander) will leave little circular marks that look awful – and I’m not employed to make stuff look awful. A belt sander will be way too aggressive. I do have two machines that I built for use when re facing note bars, one of which takes off 0.07millimetres each pass, even this is too aggressive for re-finishing! At this stage I also repair any edge damage.

wpid-img_20140910_143749.jpg

Day 3: Yep still sanding and feeling insane! However they will be finished today.

After all the prep work is complete, I do a primary tune. This is one cycle through all the notes to bring them back up to pitch. It is easier to tune things that are too sharp, but more invasive bringing notes up from flat (which need to be made shorter). I do this at this stage so that the ends of the bar can be varnished, without the need to re-seal them when I come to properly tune the bars later.

Once tuned, I wipe off all the dust, then start varnishing.

wpid-img_20140910_182828.jpg

I do all sides except the face on the first run, then leave them overnight.

Day 4: De-nib the notes, which means lightly sand with fine abrasives to remove any dust particles that landed on the wet varnish, thus creating a “nib”. Then I varnish the faces with the first coat.

Day 5: De-nib, varnish bottom and sides.
Day 6: De-nib, varnish face.
Day 7: De-nib, varnish bottom and ends (not sides).
Day 8: De-nib, varnish faces and sides. Have a celebratory beer, the varnishing is complete. I then put them to one side for as long as I can to let the varnish cure prior to tuning.

This post continues

Premier Xylophone (Job No: 1079)

An old Premier xylophone in for tuning and repair.
20121031-104534-AM.jpg
There are two main design issues with this instrument, and a further complication:
The note rails sag in the centre mainly because of the joint, but fundamentally because thin wall tube isn’t strong enough. The simple solution is to weld the joints in place to create a fixed rail.
The second design problem is that the resonators act as a structural component of the frame; this is just silly, doesn’t provide strength and is an example of how instrument manufacturers fail to understand the rigours of life as a xylophone and how they are used.
The complication is the round tube on the lower transoms which hold the castors; again this is just a silly idea which makes caster renewal needlessly difficult.

So this frame wobbles all over the place, and investigations show that whoever repaired it in the past didn’t understand the forces in play.

Newtons Mechanics: p=mv

p is the momentum or force, m is mass, v is velocity. Ignoring units of measurement a 60kg xylophone being wheeled along a corridor at 3mph (slow walking pace) will triple the mass to 180kg.

Put the castor at the bottom of long lever (frame leg), and that 180kg is multiplied again.

Understanding these principles explains why bolts shear, frames break, castors fall apart on commercially produced instruments.

However, understanding the principles also gives the solution for me to make the repairs that will last. I am happy to sacrifice the lifespan of a caster to save a xylophone, that is easy to repair, use stronger casters.

Back to the xylo.
First job is to make new lower transoms that are fit for purpose, and will make the rest of the frame easier to strengthen; solution = square tube

20121031-112443-AM.jpg

Adams Xylo Tuning (Job No: 1071)

Another xylo in for tuning.
20120913-080443-PM.jpg

There is some edge damage on the accidentals in the middle of the instrument. Close up below:
20120913-080623-PM.jpg

The white residue is superglue – precisely the wrong thing to use. Although it is available in different viscosities, the most readily available brand is way too thin, and penetrates too readily then dries and crumbles internally. Furthermore, nothing else sticks to it including finishes.

As each note arrives at its first tune, it is inspected for damage.

20120913-095545-PM.jpg

Difficult to see, but at the right of the arch, the edge is splitting. This needs to be broken out – glueing is just not going to fix the problem, the crack could be deeper than can be seen from the surface.
20120913-100629-PM.jpg

As soon as the investigation starts, it can be seen that the splits are deep.

http://www.youtube.com/watch?v=W5WGjVi5cwY&feature=youtube_gdata_player

First I break out fibres with a knife, then I often cut them away on the sander. Next I rasp the surface using a random cut rasp; this picks up any residual loose fibres. Then file and sand it smooth ready to be tuned and finished.

However, this note was still dead, a sure sign that there is a crack. A face crack was found:
20120913-101554-PM.jpg

Again, this has to be picked out to see how deep the crack is:

20120913-101743-PM.jpg

Looks drastic, but the note is dead, so there is nothing to lose. This just lifted out, but all around the wood is just falling apart. In for a penny, in for a pound…

20120913-102010-PM.jpg

This demonstrates exactly what I mean about internal damage. This is a very badly damaged note – I suspect that it could have been caused by bad seasoning of the wood.

20120913-102416-PM.jpg
It is clearly deep, but the note is now resonant, and therefore can be tuned and played until a replacement is obtained. Also observe the direction of the grain on the end of the bar, not ideal especially on xylophones.

Bergerault Xylophone Tuning (Job No: 1059)

Due to the use of hard beaters, xylophones are prone to face and edge damage.
Whilst it is easy to see the wood splintering at the edges, damage to the face is harder to see, but both will reduce resonance, and of course affect the tuning.
Vibrations don’t only travel along the length of a note bar, but also across the width and through the depth. The grain of the wood runs up the length of the bar. Below is an example of the cellular structure of hardwood:
20120907-115219-AM.jpg

In general the cells are lying like long strings of sausages alongside each other, stuck together with cellulite. Vibrations go along the length very easily, but meet more resistance going across the strings of cells. A crack creates a void that the vibrations cannot cross. So depending on where the crack is will determine whether anything can be done.
Glueing the pieces back together is not practical, but also will not solve the problem. A crack through the entire depth of the note means a new note, any repair will be temporary – the wood will always be trying to get to a relaxed state, glueing and clamping is stressing it.

20120907-120359-PM.jpg

So the only option is to remove any loose wood that is dampening the sound. As it is removed, the improvement in resonance can be clearly heard. It looks drastic, but the note is useless as it was, and repair is cheaper than making a new note.
The photo shows that the left hand edge has been cleaned up, but the note still didn’t sound, so the face was examined, and a crack discovered. That groove in the face is about 3mm deep.