2019/04/24

Another roadblock...

I was doing some experimentation over the past couple of weeks, and I was able to make some progress on the largest sizes of horns. What I was attempting to do was drive a large diameter of pipe using a smaller diameter diaphragm, and I was able to drive a 3-inch 10 ft pipe using a 1-1/2-inch diaphragm. The problem arose when I realized that even the full 10 ft length of pipe was not actually going to be low enough for my target lowest pitch. This is a problem for a number of reasons, but the main issue is that, to put it briefly, I probably won't be able to use air horns for the lowest pipes. If I did, they would need to be close to 20 feet in length, and there would be a dozen of them that were longer than 10 ft and larger than 3 inches in diameter. The main reason for this is that an air horn, essentially acting as an open pipe, needs to be a half wavelength at the pitch it plays. Since my target lowest note is somewhere in the vicinity of 32 ft, that makes the whole thing even more inconvenient to build and move then I originally expected, not to mention the expense in parts and the ridiculous airflow requirements. On the other hand, if I move away from air horns, there are options available to make quarter wave pipes instead of half wave, which means that the lowest pipes are considerably shorter in length and therefore much easier to build and arrange.

My plan at this point is to try to build something resembling a more standard style of closed-end organ pipe, which basically functions like a giant whistle. I'm hoping I'll be able to make the fipple (the part that creates the tone) as a fixed design that works for multiple pitches, so that I can just stick one onto the end of each pipe of the largest size without needing to design each one individually. If this works, it will simplify things considerably, because making 3"-diameter air horns work correctly without shrieking wildly on the attack (the beginning of the note) is quite difficult, and it takes a not-insignificant time for a horn that size to start resonating.

2019/01/11

Keyboard synth!

This Monday I made some progress on the keyboard wiring, and realized that if I finished it in time, I could throw together a quick demo for Tuesday's monthly maker meetup in my area. The idea of the meetup is that people can give informal 5-minute presentations to the group about projects they're working on, ask for help with designs or repairs, talk about interesting goings-on, or just come and see what other people are up to. I've given a couple of presentations there in the past, but this time I had the time and found the motivation to scramble together a bit of a demonstration.

2019/01/04

This post might not make sense yet

None of this is likely to make any sense right now, but it hopefully will soon. I'm just trying to get some thoughts published so I can come back to them later; I might actually edit this post into something coherent as well.

Large diaphragm diameters pose some problems both for fabrication and for operation. Flow rates, diaphragm materials, and attack quality are all problematic. Being able to use a smaller diameter of diaphragm while keeping a large horn diameter could solve some of the problems.

An air horn has a pressure peak in the middle, which is also a flow node (point of minimal air movement). This may be relevant to changing horn diameters; since PVC couplings create a discontinuity in the diameter rather than a smooth taper, it might be necessary to ensure that a change in diameter is made where the flow is at a minimum, to avoid losing amplitude as compared with a horn that's full diameter for its entire length.

The discontinuity could cause the horn's attack to be weird or unpleasant, and large enough changes in diameter could actually cause the horn to not work properly at all. I need to test a couple of things:

  1. horn with significant diameter change at half-length (ex. 1" to 2")
  2. horn with same diameter change but near the diaphragm end (much less than half-length)
  3. horn with extreme diameter change at half-length (e.g. 1/2" to 2")
  4. if 1 and 2 have similar tone and volume, try extreme diameter change near diaphragm end

2018/11/24

Progress update!

The current state of the organ:

Electronics
  • The PCBs have been redesigned to use connectors that are easier to assemble and work with, as well as surface-mount components to reduce the footprint and add mounting solutions.
  • The input boards have been fully tested, and the output board design has been tested for basic function.
  • The wiring harnesses for the keyboard are currently being assembled; completion is about 50% as of this post. It's a meticulous process that takes some attention to detail.
  • The interface between the control boards and the main controller needs attention; live input latency may be an issue with the current design due to frequency limitations of the isolation board.
Software
  • The software for MIDI playback is functional, but improvements are planned.
  • Live input from the keyboard has been tested, but live output to the organ may have latency issues. I'll need to put together a test system, probably with just one set of horns, before I can evaluate latency.
Hardware
  • Manifold configuration is basically set: I'm just gonna drill and tap some 2" PVC pipe for 1/2" NPT and screw in the valves directly. This will probably require removing the solenoids from the valves while assembling, which will be time-consuming, but it's got a very low part count; I only need to buy enough pipe and fittings to have room for all of the valves to be attached.
  • I'm going with bigger, more expensive valves than I initially planned, because I just can't get the throughput I need with the small ones unless I want to design and build a bunch of pneumatically-controlled valves and use the small ones as controls, which adds a ton of work to the build. I've talked about this before, and I had a valve design that seemed more likely to work, but the complexity is too high.
  • Horn designs are pretty much set, although I need to spec the sizes for wavelength vs. pipe diameter. Being fully chromatic, I'll have a difference in tone when the size changes, but hopefully it won't be too dramatic a change.
  • Once the horn sizes are determined, I'll need to figure out how to package everything. I'd like to keep the thing as compact as possible, and that probably means folding the largest horns at least twice.
  • If possible, I'd like to include a sort of damper box with louvers that can be opened or closed to somewhat reduce the sound output when necessary. This may prove difficult, impossible, or ineffective.

2018/11/02

About air horns and their construction

Air horns were one of the first puzzle pieces that gave me the idea for this project, and while I don't remember what originally got me interested in them, I remember finally figuring out how they work and being surprised by how simple the design actually is.

2018/10/27

I'm not dead yet!

The blog has sat idle for more than a year, and I'm sure that any visitors between then and now have been disappointed by the lack of content (and there's been quite a lot of content planned that's yet unwritten), but progress has been made in the interim. I've redesigned the PCBs completely (they now use surface-mount components, RJ12 connectors instead of pin headers for communication, and proper IDC connectors for the main I/O interfaces), I've decided on a more expensive but much (much much much) less labor-intensive solution for valves, and I now have an oscilloscope that will be quite helpful for troubleshooting if necessary (like, for example, the I/O isolation board with the seemingly glacial-speed optoisolators). I'd like to write posts about the SPI interface, about some of the component choices, about how the code works, and a lot of other things that deserve at least a few paragraphs worth of detailed explanation. I'd also like to get a diagnosis and start some sort of treatment for ADHD, but that requires me to stop procrastinating and actually make some progress.

I've learned a lot since I started this blog, and I'm sure I've plenty more to learn before the project is finished (and who knows how long that will be yet). I've learned a lot about python, about electronics, about pneumatics, and about my variety of personal challenges. I don't want to get too existential, but I'm looking forward to moving forward after all this time.

2016/04/11

I goofed (again)

I built a version 1 input board and it didn't work.

As it turns out, it didn't work for several reasons. One of those reasons can be found upon inspection of the back of the board; one of the traces on the output connector goes right across another pad, connecting them. Another problem was that I probably connected the board wrong. The main issue was related to communication, and specifically the polarity of the latch or chip select signal. I'll post soon with some details about how communication actually works with the chips I'm using. The important thing at the moment is that the problem seems to be mostly solved. I'm not entirely sure, though, because I think I fried a shift register.

2016/03/30

Input PCBs

I put together a preliminary layout for the input PCBs and ordered them today. I'm using SN74HC32 4-channel OR-gates and SN74HC165 parallel-in serial-out shift registers to read sets of 8 notes from the keyboard. Here's the layout, as rendered by OSHPark:
board top
board bottom

2016/03/09

Electronics are hard.

Sometimes, the easy way turns out to be a little bit harder than you thought.

As I will explain as soon as I get to the keyboard layout explanation post, there are some pairs of keys on the keyboard which play the same note. I had initially planned to connect the common pins of each pair of keys and have one connection on the circuit board, but I realized that I had missed something painfully obvious about that approach.

[more inside...]

2016/02/26

Not so fast...

The valves don't actually work. Once I hooked them up completely, they either didn't open when they should, or they never closed after opening. Some leaking in the valve body was expected, but the input side leaked too much and broke things. A new design is in progress, and I'm hoping to get to testing soon. I'll do a full write-up soon.