Inexpensive Electronic Instantaneous PVC Piping Air Valve

[mfetch22]

My boss has a boat load of extra PVC piping laying around the shop and I wanted to use it to make a pipe organ. I need a way to electronically open and close the connections from each individual pipe to the main air supply. I'm on a tight budget and I'll need to make 88 valves in total. I have found a few valves online but they are far to expensive. I was hoping I could find a way to make a cheap electronic control that could achieve this purpose. I am inexperienced with electronics and am in need of as many ideas as I can get. Here are a few points about the design that might be helpful:

PIPE LAYOUT
[1.] I plan to have each pipe sealed at the bottom with a cap (all though, buying 88 caps may be too expensive so other options are welcome)
[2.] I plan to drill a hole in each cap and attach surgical tubing or something similar.
[3.] Pipes will be grouped into sections (I'm thinking four groups: bass octave, low range, mid-range, high range)
[4.] Each pipe in a group will be connected to a "parent pipe" which will be sealed at one end with a cap and the other end with an electric fan.
[5.] The four parent pipes will produce differing air pressures to accommodate the change in pipe sizes.

CONTROL SYSTEM
[1.] I plan to recreate a simple keyboard with 76 wooden keys that 'see-saw' over a wooden dowel rod that will run through each key.
[2.] I plan also to do something similar to create 12 foot pedals (for the bass octave).
[3.] As each key is depressed, an electric switch is actuated (still not sure exactly how I am going to set that up)
[4.] Each switch will correspond to an electric valve that is located on the parent pipe of the note depressed, and will allow the air to flow from the parent pipe to the tubing, and thus sounding the flue pipe itself.

This design is in a very early stage and I am certain changes will be made as issues arise or are brought to my attention. For now, I need to figure out the cheapest way to control the air flow electronically.

 

[Baluncore]

Consider using a butterfly valve like the choke or throttle valve in a carburettor.
http://en.wikipedia.org/wiki/Butterfly_valve
You could use a small electric motor to rotate the spindle 90° between open and closed. The motor would be low voltage so torque opens the valve, a spring can hold it closed because the valve is balanced.
The seal and stop could be a half cylinder section of tube inserted in the valve tube that the butterfly disk seats against when open or closed. One half upstream, the other half on the other side, and downstream.

 

[mfetch22]

Intuitively I feel like I see what you are describing. A disc on a rotating motor which rests at "closed" and is twisted "open" when the motor is actuated. I am having trouble understanding a few things. When the motor is actuated, what will stop it from continuously rotating the disc? How do I get it to stop at 90 degrees? Do they sell motors that only rotate a quarter turn when actuated? Where would the spring be attached inside the pipe? Where would the spring be attached to the disc? What material would be best to make the discs out of?

 

[Averagesupernova]

I would look into making the valves actuated by air pressure. If you can fashion a valve on each key then have that signal open the valve on the main pipe. This sounds like a very interesting project. Going about it electrically certainly would have advantages in that you could sample all the signals from the keys and record them and play it back.

 

[Baluncore]

Any small DC motor for model will do the job. Make one valve and test it.
Stop is a step inside the tube that stops disk rotation at 90°. On one side it is upstream, on the other downstream, with a corner that stops valve when in open position. It could be cut from half a PVC tube, maybe the same but spring inside the main tube.

Use a motor for torque, not continuous rotation. Torque is proportional to DC current.
You can power open with a pulse, then hold open against the gentle return spring with a fixed current.
To close you can let spring operate alone, but better to give motor a reverse pulse.

Because thermal expansion is large with PVC, disk would be made from flat PVC, same as tube material. It can be cut from flat PVC sheet with a hole saw in a drill press, no centre drill = Trepanning.
Disk is mounted on a shaft or two pins through two holes in pipe wall.

The spring could be wound as a flat spiral. I would use piano wire, maybe 5 turns so little change, 5%, in force over 90°.
Spring is mounted outside the tube on one end of disk pin, maybe the same end as motor.

 

[AlephZero]

Before you get too far into designing the complete instrument, make sure your ideas about a single pipe will actually work and produce sound. From your OP, it's not obvious whether

1. You already know how to do this, but didn't bother to mention it, or
2. You are assuming that all you need to do is blow air into the closed end of a tube. That won't work!

This is a good site on "traditional" pipe organ construction: http://www.fonema.se/

I expect your biggest practical problem here will be that the wind pressures are low (only a few inches water gauge) but paradoxically avoiding air leaks is critical, since pipes can produce unwanted noises from very small air flow leaks. Trying to make airtight seals between two pieces of PVC doesn't sound very practical.

 

[Baluncore]

Trying to make airtight seals between two pieces of PVC doesn't sound very practical.

The seals are usually made by gluing a strip of felt to one of the surfaces. This seals the low pressure air at the same time as dulling the sound of a valve when it closes against the seat, or hits the stop when fully opened.

 

[AlephZero]

The basic idea of butterfly values is much more complicated than traditional pipe organ design. The simplest version that would match the OP's requirements is called a cone valve windchest. The valve gear is all inside the OP's "parent pipe". Each pipe has a linear-acting solenoid, and the valve is a conical shaped plunger that blocks the end OP's "surgical tubing". Add whatever sealing material you want - the traditional material was soft leather, which is more hard wearing than felt.

The advantages of conical valves are that they are self-centering, and so long as the valve and pipe both stay circular, geometrical tolerances and changes of dimension are compensated for automatically.

Some pictures and dimensioned drawings here: http://logosfoundation.org/instrum_gwr/Bomi.html

Note, some of his parts are commercial organ building parts from Laukhuff, and he gives the part numbers. Their catalogs are a great source of technical information. http://www.en.laukhuff.de/catalog.html

 

[mfetch22]

I like both the valve ideas. Its just a matter of which I find to be less expensive to produce. I do plan to test each pipe as its created, rather than cut them all first. As well I was thinking of doing a mock up of both the valves to ensure that I can get it all to work. Any recommendations on the specific type of electric fans I should use?

 

[AlephZero]

Any recommendations on the specific type of electric fans I should use?

That depends on the details of your pipes, and the wind pressure. See http://www.rwgiangiulio.com/math/flowrate.htm for the basic calculations.

Note, if you are planning for your 88 note keyboard to match a piano keyboard, an open pipe for the bottom note will be about 20 feet long! The example on that page at 16 foot long pipe, and that one pipe needs about 11 CFM of air. As a rule of thumb, the bottom octave of pipes need as much air flow as all the rest put together.

The good news is that long pipes don't have to be straight. It's quite common in traditional organ building to use miter joints to bend pipes through 90 or even 180 degrees. It makes no difference to the tone, and it helps to keep out inquisitive mice etc looking for a new home.

 

[mfetch22]

The majority of pipe I have to work with is 16 feet long. I measured the diameters and they range from 1.5 inches to 9.5 inches, both approximately measured with a ruler. I was planning on having stops on the pipes for the bass octave so that they would be only half the length. So I just need to ensure that I have enough CFM of air present in the "parent pipe" such that (if all the keys were activated for that section) every pipe still sounds? Does CFM of air required for the parent pipe come from just adding the CFM of air required for each connected pipe together?

 

[AlephZero]

The majority of pipe I have to work with is 16 feet long. I measured the diameters and they range from 1.5 inches to 9.5 inches, both approximately measured with a ruler.

9.5" is in the right ballpark for a 16' open or 8' stopped pipe, but a bit small to get a useful volume of sound from it. In the 19th century somebody "proved" theoretically that the diameter should halve every 16 pipes (i.e. short pipes should have a smaller length/diameter ratio than long ones). In practice 16 is a bit to low. For large diameter pipes halving the diameter every 17 or 18 pipes is about right. For smaller diameter pipes, halving every 20 or even up to 24 pipes may be better.

But whatever number you use, 1.5 inches is much too wide for the smallest pipes in your 88 note range, as you will see if you play with the numbers in the previous paragraph. The length halves every 12 pipes, of course. For example if the first pipe had diameter 10 inches and you halve every 22 pipes (just to make the math easy), the 88th pipe diameter would be 0.625 inches. Even that is too wide to be practical, considering the acoustic length of the shortest pipe will only be about 1.5 inches.

I was planning on having stops on the pipes for the bass octave so that they would be only half the length.

Nothing wrong with that!

So I just need to ensure that I have enough CFM of air present in the "parent pipe" such that (if all the keys were activated for that section) every pipe still sounds? Does CFM of air required for the parent pipe come from just adding the CFM of air required for each connected pipe together?

Yes in principle. In practice, the air pressure will decrrease as the CFM increases, unless you have a pressure regulator in the air supply system.