I have made a new frame for a set of 1″ chimes, so it made sense to extend the playing range up to the top G at this juncture. The frame has already been delivered, so now I have the time to make the bells.
The first thing I need is an example bell, in this case I asked for the F. Not only does this enable me to take all the measurements that I need in order to make any parts, but it also helps during the tuning process so that the new bells fit within the existing set.
It is now virtually impossible to get brass tube in the size that I want, let alone the correct material. Brass is an alloy, so there are lots recipes to get the required properties for the desired application. Over time, brass is being replaced (presumably by plastic) so the commercial requirement for these mixes has largely disappeared. Combined with the mills being bought out during global monopolization, this has resulted in higher prices and less choice. Of course, if I buy sufficient quantity (a metric tonne) I can get whatever I want, but I am instrument maker – I make musical instruments for a living therefore I am poor; spending thousands of pounds sterling on lengths of brass tube is just not going to happen. Additionally I would need four external diameters, and three wall thicknesses, that is twelve tons of brass tube! If I had that kind of money, I would retire to the Caribbean.
Fortunately I do have some stock, which still equates to well over a thousand pounds just sitting on a shelf! In amongst those tubes I did indeed have a length of the correct material, which is the major hurdle negotiated. So the first job is to chop it into lengths longer than the bells I need to make. There is nothing worse than making a bell only to find that it is too sharp right at the end of the process when the tuning happens. With the two new tubes cut, I drill the holes for the string to match the existing bell and stamp the tubes.
Using the existing bell as a reference, I measure the cap dimensions so that I can form the inside of the cap from solid bar. This form is then cut off the bar giving me two crude caps. I use a donor bit of tube offcut as a temporary bell and spot solder the caps in place. This enables me to hold the offcut tube in the lathe to form the external shape. My lathe doesn’t have a large enough bore to pass the tube through its headstock.
With the caps now made, it is just a matter of removing them from the donor tubes and soldering them in place on the actual tubes. After they have cooled, I hand polish the whole bell, then tune it and send it off to be chrome plated if required.
Moving parts by their very nature will always cause problems, especially so if they are not regularly serviced or designed and made badly. This is the case with these Premier tubular bells. In defence of the customer, there isn’t actually anything they could have done in particular to have prevented the noise produced from the damping system.
As seen clearly the noise problem in this tubular bell damping system arose mainly from the choice of materials; the wooden dowel. In defence of Premier they have economic constraints; everyone wants to spend as little as possible on musical instruments, so for Premier, and indeed any manufacturer, they have to shave off costs at every opportunity. Wooden dowels are cheaper than ptfe rod, so wooden dowels are used; spending time to minutely check every component takes time which in turn increases production costs. There is a solution however; the customer has to pay more – simple. Whether initially they pay more to have a proper instrument made, or they pay more to have a cheaper instrument re-engineered, either way the only answer I see is the musician paying more money. Ultimately you get what you pay for.
It could be worse however, and I have seen worse systems, at least I could work with what I had to silence the problems. Other than the damping system, there were creaks that originated from the frame in general, these were removed by re-assembling the frame with a care and the usual attention to details.
Premier tubular bells, in good condition now sold.
The problem with the brass finished tubes is that they discolour over time, but they can be linished with fine wet and dry paper (I use 400 grit) and they will come up like new.
The frame is in good condition on the whole. There are a few minor splits in the oak panels, but they are very minor and don’t affect the instrument structurally.
The damper vane on this range of Premier bells is a return to the original design of a sliding mid gate, as opposed to a rotating vane. The latter is a board that rotates between the two rows of bells. The sliding mid gate is simply another board with holes in it that moves from side to side in the instrument and pinches the bells. It is more reliable that the vane, as long as the holes are in the right place – one of which is not, so the low F is not damping. Again this is something that can be easily corrected.
The low F# was split at the cap. These tubes are no longer made due to changes in EU legislation, so the only option I had was to invert the tube, repair the split and retune it.
The pitches of the bells were all over the place, now they have all been tuned. So the set complete with a cover is ready for use.
As usual, any questions feel free to ask.
The beginning of this post is 1241: Bell Frame (pt 1) which covers most of the work that was done. Whilst the metal work was away being powder coated, the board was finished and varnished. What I end up with, is a whole pile of bits to be reassembled on various instruments.
I was asked about the strength of the hinge and the boards flexing under body weight in a comment on (part one). The main bar around which the whole step rotates supports the front edge of the forward board, and there is a steel brace to support the outer edge of that forward step.
The new extension board has braces running across the width of the board both fore and aft. These braces also provide the pivot points around which the legs rotate.
Below is a short video showing how it all works.
The first part of this post is 1206: 1.1/2″ Tubular bell (part 1)
In the first part of this post, which is the first day of work on the instrument, I made the cap, and the tube, and joined the parts together. On the second day, which gives everything 24 hours to settle down, I tune the bell.
Tuning is a simple job of gradually shortening the length of the tube to raise the tone to the desired pitch. When they are made in mass production, all the tubes are cut to standard lengths then chrome plated. This is why they are never in tune, even the fundamental (the overall length) is at the wrong pitch. Obviously my standards are a lot higher, and the fundamental is bang on the correct pitch, what I find out in the tuning process is whether my tweaks to the design have worked. In this case I was delighted.
There are two main tones in a tubular bell: the fundamental and the strike tone. The strike tone is the “clang” when the bell is hit, and it is this pitch that is hard to get correct. On this bell I was within 20 cents of a perfect octave, which means I might have cracked the problem. I have to replicate the bell now, and then work out the solution for the other 19 bells!
Once tuning is complete, I just clean the bell so it is shiny.
Tubular bells, like most percussion instruments are simplicity themselves; they are just a tube of brass normally reinforced at the top with a cap. However, like most percussion instruments, they are little understood, and the simplicity of design belies the challenges faced by an instrument maker to deliver a good sounding instrument. If it’s just a tube with a lid, how will I know whether it will sound good, and what can I do if it sounds rubbish?
I have yet to be convinced by any claims that manufacturers make in marketing about their ‘fantastic tubular bells. What they fail to publish, are reasoned arguments, supported by research and data. However they are not writing publications for me to read, they are writing to sell instruments to musicians who go on to accept their statements as fact [suckers!?]
So what is the solution? Research. Everything I do, I document; it is from this accumulation of data and observations that I gain a deeper understanding of all the instruments I work on. Furthermore, it provides the platform from which hypothetical improvements are made and subsequently tested. By improvements I mean manipulating an instrument to improve its musical timbre.
To make a set of tubular bells by mass production, a saw to cut the tube to a predetermined length is needed, and a drilling machine to drill the holes in a set position. There may or may not be a secondary tube for the cordage to pass through. To make the caps, a computer operated lathe will be used. The caps are then pushed into the tube using pneumatic rams. Polish and chrome; job done. One musical instrument made!
My method of making tubular bells is very similar. Starting with a suitable length of stock tube and referencing my accumulated data, the tube is cut a little longer than the longest bell of that pitch that I have ever seen. Why?
The first reason is that no two lengths of brass are identical. The assumption that materials are completely uniform is a fundamental flaw in the concept of mass produced musical instruments.
The second reason is to accumulate more data. Because I can only get two, possibly three bells out of a length of tube anyway, I cut them as long as possible. The result for me is two or three times more data is acquired as I raise the bell to pitch by shortening the tube. The extra length also provides opportunities to experiment and still be able to make a bell at the correct pitch.
Now I have an oversized tube I clean up one end clean up one end. This has to be done by hand, because I don’t have the room for a lathe big enough to take a tubular bell (unsurprisingly), so it takes a me a lot longer versus factory.
After I have the holes drilled, I remove all the sharp edges, making them rounded in cross section. This prevents the metal from slicing through the cord. It is this attention to detail that takes me time, but saves my customers time and frustration.
Before the cap goes in, I stamp the tube with the pitch of the bell rather than the cap which is the playing surface. In several instances, my customers accumulate a complete set of bells over many years, and I want to give them consistency of design, so I have a little pattern to get all the stamps in the right place.
What I never get is perfection. It’s bloody hard to get the stamp perfect, I used to get moaned at by my boss when I worked at a shop for wonky letters. Over a decade later, I still find it impossibly difficult and it still upsets me when they are not right, but I’m better at living with the disappointment and view it as a mark of humanity as opposed to the machine.
In the photo, I am checking the angle at which the chamfer is cut.
The cap is made to a sliding fit into the tube. The cap is solid brass, if it has to be forced into the tube it is the tube that will expand. This expansion of the tube creates internal stresses and makes the brass less ductile, and it is the reason why the tubes split at that point. In my view this is another indication of the embarrassing lack of knowledge displayed by the major manufacturers!
This blog post continues in 1206: 1.1/2″ Tubular bell (part 2)