(G A #A B C D E F)
(G .. F+)
||open flue cello
|25 * Counter
(C .. C)
||open flue violin|
(G .. C)
||open flue violin|
|20 * Chimes (C-G)||-||101-120||72-91||tuning forks w. resonators|
set of stoppered basses turned out a
little too meek.
finalized the height of the cabinet to be more than planned when I
made those pipes long ago, I made a new set, this time
open pipes with the same diameters (51..71 mm, 2..3 in). Only the two
ones, for G (1.75 m long) and A had to be mitered.
The body of them is made from 7 mm pine plywood except the front that is oak. In respect of their size I made good use of a table saw and a shoulder plane to make the long joints with grooves and tongues. Some extra work that pays off when it comes to gluing. First back and sides, with the front loosely in place to keep the sides parallel. Next round, after fine trimming the two bottom blocks are glued and, having clamped those, the front.
|The mouth design like shown, gives a fairly large degree of freedom in voicing. A number of loose pieces, in particular a little bar f, resting in notches in the ears e, to hold down the adjustable upper lip a with an elastic leather gasket in between. Nothing is glued here, all is held together by the slim screws through the ears into the sides of the body, plus a crosswise threaded bolt with nuts through the intonation roll d. The roll has a big hole in it to allow for fine adjustment of its position. So far I have always found the lowest point of the roll should be on the plane of the languid, the adjustment to be done is the distance to the flue jet. Perhaps the least elegant adjustment is for how the jet hits the labium - can be modified with shims between a and b unless you do some carving in b or remake a. The round hole in it is a pull-out handle, decoration, and trigger for people to ask what for.|
These pipes have tuning sliders
at their top end, somewhat
elaborate than with the smaller pipes. They are made from brass sheet.
Inside the pipe, each side of the tuning slot, is an oak stick with a
threaded hole, held by a locking screw. Tightening these screws presses
slider firmly against the inside of the pipe front.
Four of these eight pipes are equipped with detuners, such that the rank implements a fully
the lowest bass pipes,
cylindrical 'bass clarinets', one quarter
wavelength. Also this rank replaces an initial one, namely conical
trombones. These clarinets are somewhat
than the trombones, but they speak one octave lower. Since the boot has
a rather small volume it is fitted with a spring loaded diaphragm to
a 'schwimmer'. This acts to reduce the AC pressure variations inside
and is good to ease voicing and to stabilize reed oscillation.
The boots of these pipes have little flags, strips of motor cloth, fixed with screws at the top ends, and marked with white dots at the low free ends. Behind each flag is a small hole such that when the pipe is blown the flag swings out to visualize that the pipe is on.
The shallots are made from copper pipe, slotted and with a brass cover plate soldered on. This plate has a hole extending under part of the tongue and is covered with glued on pouch leather. The tongues are brass,1.2 to 1.5 mm thick, held by a pair of screws threaded into the shallot plate.
This is a set of
violin-type pipes that are a continuation downward from the melody
violins. The design is shown at the section of the countermelody
violins, register 6.
|These pipes form a link between
the bass (reg 0) and melody (reg 10) flutes having the same scale.
Bridges have been added later to allow for wider flues and more power without overblowing. On the chest this rank is placed between the cellos (reg 2) and the reeds (reg 4) such that nobody can see those ugly bridges.
resonators of these are a continuation upward of the bass reed pipes,
speaking one octave below the other ranks on the accompaniment chest.
The shallots and boots are inherited from an earlier, but discarded
rank of conical pipes.
The resonator front plate is bent slightly inwards in the lowest part, in order to give room for the tuning wire.
The jack mechanism to adjust the tuning wires with a screwdriver is no good invention. It is rather easier to tune them the conventional way knocking the wires.
oak slabs, recovered
from an old parquet floor.
They are deeper than wide by a ratio of 4:3. This is the only rank (except the trumpets) where the cross section deviates from square. Bridges cover the entire mouth area. These features allows the airband to be quite thick without overblowing, so these pipes are comparatively loud.
These very skinny pipes are made from recycled old fir. The resonators of the dozen smallest pipes are formed as a milled grooves in integral back pieces, covered by front plates, chamfered for the labium. The bigger pipes are assembled conventionally from four plates plus a bottom block.
The cut up is fairly low and they are vigorously blown. To prevent overblowing the height of the bridges was carefully adjusted in the same time as the flue width was made as large as possible.
The inner surface of the bridges is intuitively streamline shaped. The chamfering of the covers is essential to let air be entrained with the jet. The sound from these pipes is amazingly loud and penetrating, regarding the minute size of the openings in the mouth region.
head piece, made from oak, has upper and lower cylindrical surfaces
axes. This offset gives room for the tuning wire running parallel to
tubes. The resonator tube and the boot are both
cylindrical. Nominal 1/4 wavelengths are quite close to
actual lengths of the resonator tubes. Diameters range from 29 to 17
halving number about 24.
The tubes are made from standard half hard 0.2 mm brass sheet. This is much springier than pipe tin alloy, makes it difficult to shape them, but has a good tensile strength such that I can have such a thin wall and save a lot of weight. The tools to shape them was a set of cylindrical oak dowels, diameters 29, 25, 22, 19, 17, 15, 13 mm. I used only these listed diameters, about 5 pipes each size, no need to make individually interpolated diameters.
was built with the same basic design and measures as the open flutes
below (reg 10). Because they are stoppered they however speak one
octave lower, and this calls for a higher cut up, other parameters
For the other pipes lengths are proportional to the inverse of frequency, and diameters to the square root of that.
shape of these resonators I arbitrarily decided on the area function A=Ao*exp(3*x2),
where Ao is the
throat area and x is the relative length coordinate along
the horn, from 0 to 1. This is not a common exponential horn because
argument is squared. This means the flare is moderate in the throat and
much more drastic toward the mouth, an expedient to make it a little
more trumpet-like. There is a factor 3 for the argument,
meaning that the mouth area will be exp(3)=20 times larger than the
throat area. On top of this I prescribed that both the throat and the
mouth are to be squares. Mouth diameter is then about sqrt(20)=4.5
times throat diameter.
This defines the relative shape of the resonators. After making a prototype I found the length of the resonators should be 0.35 wavelength. Skinniness was arbitrarily selected to L/W(69)=10 (half wavelength/mouth diameter for reference note a') and the halving number M=24. This results in horn length Lh and mouth diameter Wm for the a' pipe as shown beside.
|To make the curved side pieces I started from this drawing (blue contours), as computed from the formula above, respecting contribution from the trapezoidal shape of the companion sides (red contours). The cross section of the horn at half its length is a quite oblong rectangle. Next step was to make copies of the drawing, one for each pipe, appropriately scaled lengthwise and cross wise. These were glued to the plywood stock, laid in two layers, and served as templates for cutting out the blanks.|
highest pipes most are equipped with intonation rollers between the
ears to prevent overblowing. The precise position of a roller has to be
tried out during voicing operations.
After cutting a small recess in the cover, the inside of the cover and the outside of the body are ground flat. One idea with this design was that it might make it easy to control the flue slit by selecting and perhaps grinding down the thickness of the gasket. Later experiences tended to show it is easier to file down the languid, so that is what I did for most of the other flue ranks.
For reference, here is an expanded
table, also linking the MIDI note
(0<n<127) to the
conventional note names (A .. #G),
and the fundamental frequencies f
(pitches) of these notes in Hz. Only
the span covered in the organ is shown. The
underlying basis is the equally tempered scale and that standard pitch
is a'=440 Hz, MIDI note
number 69. The table thus shows the fundamental
frequency as defined by
the formula f = 440 * 2
The table also highlights the continuity of scaling across the different chests. For instance that the flutes of the bass, accompaniment, and melody chests essentially form a coherent principal rank.
|The length dimensions stated are
nominal fractions of fundamental wavelength. Because of the end
correction the physical length of the tuned resonators are shorter than
this by about 2W for open pipes, by W for stoppered and reed pipes. The
pipe external length is greater, to include pipe foot/boot and material
adjacent to the top tuning device.
|Note #1: Four of the bass flue
pipes have detuners to enable them to generate either of two adjacent
|Note #2: The melody chest
additionally holds a rank of stoppered pipes with the same key
dimensions as its open flute rank, but thus speaking one octave lower.
|Note #3: The counter melody
flutes are 1.3 times deeper than the stated width W
|Note #4: The counter melody
violins are identically the same design as the melody violins.
|Note #5: The melody trumpets are
flared horns with a mouth area about 20 times larger
than the throat area. The stated width W is that of the square mouth