What Finally Came of the Organ Pipe Project
Steven Ngai, December 2008
I had thought of constructing an organ since the summer of 2008, but because of academic factors I was able only to tinker a little here and there. To learn about some of the experimentation that led to this point, see Pipes (manufacture since revised) and Windchest/Blower.
All of the final building was therefore compressed into a few days during Christmas break. The following pictures show the organ in various stages until the point when it first became remotely playable on Christmas morning, 2008. The organ formally debuted that night when it was used to accompany Christmas carols.
Built with an evolving plan and fervent wishes that it would work after everything came together, the whole organ is still to some extent a prototype. The earliest "permanent" part was the windchest, fed by a Fasco blower I obtained from Craigslist.
Because of the current economic clime, and because the organ was just a prototype that I wanted to get working, I strove to reduce costs as much as possible. Thus the wood for the organ cost less than $20 total (!) and came from two sources: stained wood boards intended for exterior picket fencing (about $1 per 3/4"x6"x6' nominal board) and a nice particleboard I happened to have lying around. Because it would become the lid, the width of the board dictated the corresponding dimension of the windchest.
Beneath each toehole is an simple pallet valve. The valve hinge, inspired by the leather used in old organ pallets, is paper. Yes, paper. In testing this proved remarkably durable even with considerable travel; fortunately, the end of the pallet travels not even 5mm in practical use. Equally shocking is that the valve returns via a rubber band that is threaded through the windchest lid. This too may seem inelegant, but the traditional spring would not only require a guide to ensure straight return but be considerably more expensive.
Even though time concerns forced me to scale back my original plans to demonstrate stop action with a second rank, I never really liked the slider stop action. Inspired by, and yet much simpler than, the Austin universal windchest stop action is the dual-hinge construction demonstrated here. This allows the implementation of stops; even if the short arm is pulled down by a key or pedal, the rank may be stopped by a bar that blocks the entire row of long arms.
About half of the boards I sliced lengthwise with two cuts into three pieces, forming two thicker beams and one thinner beam between them. These would become the naturals and sharps of the pedalboard. These needed to be long, measuring four of the six feet available from each board, in order to convey the pedal motion underneath the bench and to the pulldowns. Because the sharps were cut on both sides, their surface was unstained, and I got old-fashioned reversed key colors for free. The frame and front assembly were also constructed from similar pedal-width beams.
The front assembly that keeps each pedal in place is somewhat self-explanatory. I would have put in similar spacers at the back end, but what I had worked well enough, and the cut boards were slightly warped on the other end. A wooden dowel passes through all the pedals near their back end, serving as the hinge; it, and therefore the pedals, are removable. The pedals return via, you guessed it, more thick rubber bands draped across the hidden backside of the assembly. You may laugh, but the feel is just about right, and I wouldn't have it any other way: there are few organs whose replacement parts you can find wrapped around your broccoli at the grocery store.
(In the picture to the right, two nails for the rubber bands are visible. Also visible is the beveled edge of the natural-key pedals; think about a picket fence, and you'll understand the shape.)
As part of my plan-as-you-go construction, I had the windchest on three folding chairs for several days before finally constructing a rack for the pipes and a frame for the organ.
And as the pipes went in, the organ began to look a little more like an instrument. While I was unable to find a substitute for the foot construction detailed in Pipes (and those unfamiliar with this earlier experimentation will be surprised to find out what the foot is constructed from!), I did make a satisfactory advance in the construction of the upper lip by cutting directly into the pipe using a miter box-like fixture of my own construction.
Here I'll digress briefly to explain the physics of the open flue pipe.
When the pallet leading to a pipe is first opened, a thin sheet of air (the "windsheet") exits the flue, traversing the mouth to strike the edge of the upper lip. The portion of the wind that first enters the pipe sets up a sound wave containing a multitude of jumbled wavelengths; this propagates at the speed of sound toward the pipe's open upper end. When the wave exits the confined pipe and encounters the open atmosphere, the abrupt impedance change causes a matching wave to be reflected back down the pipe toward the mouth. As this wave bounces repeatedly between the two ends of the pipe, destructive interference causes the cancelling of all waves except those whose whole number of half-wavelengths matches the resonating length of the pipe; in this way the length of the pipe dictates the fundamental pitch and its harmonics. As the wave becomes well established, it pulls and pushes on the delicately balanced windsheet, sending the wind alternately into and out of the pipe in step with the oscillation within. This positive feedback thereby maintains the vibration that has been established.
If we assume the pipe's sound contains minimal upper harmonics, the following is what happens in the steady state: as the windsheet is being blown away from the mouth of the pipe, air is also exiting the pipe through the open top, and pressure drops to a minimum at a point halfway along the pipe's length. Subsequently, in response to the partial vacuum, the cycle begins to reverse. As the windsheet is being sucked into the pipe, air is also rushing into the pipe through the top; pressure rises to a maximum at the halfway point. The rapid alternation between these two states constitutes the sound emitted by the pipe.
If you look under the windchest, you will see how the pedals transmit their motion to the pallets: strings. (Actually thin yarn from Michael's, about $2 for hundreds of yards.) This was the part I had been most unsure about. Fortunately, it turned out I did not need to construct any complicated lever action: instead, I simply dropped a string from each pallet, passing it through a hole on the bottom of the windchest and terminating it in a paper clip. I then passed another string through each pedal near its back end, tying it to the paper clip in a noose. In this manner the string could be tensioned to the point that a light depression (5mm) toward the front of the pedal--about 1mm at the string attachment point--was enough to open the valve and allow the pipe to speak clearly. Because the spacing between the pipes and pedals differs, the strings formed a nice fan shape. Despite the de facto pulley action that each hole serves, the strings perform admirably.
Video and audio will be forthcoming, but for now this will have to do. I am glad this project demonstrated that a perfectly functional pipe organ can be built by a hobbyist inexpensively and without specialized supplies; I hope it inspires some of you out there. Questions? ssngai att gmaill dott comm.