Experiments with Flue Pipes

Organ pipes come in two main variants - reed and flue. The reed pipes are similar to clarinet, oboe and Saxophone instruments, flue pipes are very similar to recorders and whistles. The thought of making a reed pipe worries me. Reeds might not be as hard as I fear but I think they are going to cause problems. Recorders don't look hard to make and that thought spurred me on to my first investigation of home-made flue pipes.

The basics of flue pipes is simple. A stream of air is forced over a sharp edge and a length of pipe behind the edge forms a resonator for the sound waves to build up in. All that's needed behind the airstream and the edge is a place where the sound waves can live (in the jargon a 'cavity') and with the right circumstances of air pressure and mechanical alignment it's hard to prevent the sound from forming. Referee's whistles and the whistles on steam engines work on the same principle. Typical organ flue pipes are either closed at the end or left open (closed pipes are called 'stopped' pipes) and perhaps surprisingly the closed pipes make a note an octave lower than the open ones. There is also a noticeable difference in the sound of the two types of pipe because the harmonic content is dramatically different. A lot of what I've learned about this comes from the links listed below.

My first experiments in producing flue pipes couldn't afford the luxury of spending the time to reproduce anything that looked like a traditional church organ pipe: because I was in a hurry I took a look around my local hardware store to see what I could find in modern materials that might be useful. I soon found a length of plastic bathroom overflow pipe and some wooden dowelling that was a push fit inside it - these were exactly what was wanted.

I cut a length of pipe that looked similar in proportion to one of the children's recorders that they play at school, maybe about a foot long. I cut an inch or so of dowel and filed a flat on it so that when it was pushed into the pipe, there was a small but visible gap where the air could flow between the flat and the curve of the pipe. That would give me my airstream, next I would need a sharp edge that stuck into the airflow. I wasn't sure how to do that but thought that as a first attempt perhaps I could just file a hole in the pipe above the dowel. I took a half-round file and with the rounded side pointing away from the dowel, up the pipe, I filed a hole about a quarter of the way through the pipe. When the proportions looked vaguely like the children's recorder I stopped and tried blowing through the pipe with the dowel inside it. The dowel wasn't a tight fit and lots of air leaked everywhere so I wrapped some Gaffer tape around it (except over the flat part) and banged it back into the pipe with a hammer since it was very tight after that.

The first attempt of blowing it produced a clear and loud musical tone! I was astonished. I put my finger over the open end of the pipe and it dropped about an octave. The pipe did seem to be rather sensitive to the air pressure so I put two strips of blu-tack parallel to the sides of the pipe to prevent the airstream from spilling sideways around the edges of the hole and that noticeably stabilised the pipe.

A second pipe twice the length also worked first time, but that one is much less happy about being closed at the end, having a tendency to squeak up a couple of octaves rather than providing a strong note at the expected pitch. This is almost certainly because the 'scale' (ratio of width to length) of the pipe is wrong. With some more trial-and-error experimentation I am sure that I can produce a useful range of pipes from this simple and quick construction technique. The local plumbing shop has a range of pipe diameters available so presumably there will not be any difficulty in getting the length and widths that will be needed.

Further experiments on flue pipes have been delayed until the xylophone is fully working. When they re-start I'll document them here. The task of producing enough air at the 3" water pressure (approx 700 Pascals) that the pipes need, the bellows and the electromechanical air valves is undoubtedly going to be a challenge when I come to working on a full rank of pipes.

Further Reading