Perry Rylance

Personal & Professional blog

Building the frame

My train horn piano needs a frame on which the horns and other components can sit. I’ve grabbed some lengths of slotted angle iron and given them a lick of yellow Hammerite.

Giving the angle iron a coat of yellow

I have a 12v 110Ah lead acid deep cycle battery, with a tray, and have got the compressor out. The arrangement of these two components as well as the brain box will determine how much width I need at the rear of the instrument.

To determine the length, I’ll need to allow some room for the solenoids and choose how to hold the rear of the assembly together – I’m thinking I’ll probably end up using fascia board for this as it’s not going to be water damaged (as opposed to wood), and it’s fairly lightweight.

In order to do this I’m going to need to know how long the rear section is, and how long the horn section is, and then add space for the pneumatic hoses and solenoids.

I took the audio from the first successful sound test, and ran frequency analysis on the signal with Audacity. The results are fairly clear.

Frequency analysis of the test horn sounding

I used a tone generator to double check this and have adjusted slightly. I don’t think the algorithm has enough data, in terms of quality or length. I’m very confident tuning almost any instrument by ear, as far as I can tell, the fundamental frequency of this horn is 211Hz.

Knowing the frequency of the test horn, and the length of the test horn, I can then choose the low note I want (E3 in this case) and then use a spreadsheet I made earlier to find out the lengths of the respective horn pipes.

Now I not only know roughly what the overall length of the build will be, I know to the millimetre what I need to cut each respective horn pipe to in order to get chromatic tuning.

This spreadsheet uses the 12th root of 2 to calculate frequencies and tube lengths, which is the standard for twelve-tone equal temperament instruments – by far the most common Western system in use.

In other news, the brain box has received a 16-way blade fusebox, which has been wired into the common terminals on the relay board. A 30A breaker has been fitted for the entire brain box and solenoid bank. Later today I’ll get the components arranged in the manner I would like them on the rear of the frame, and hopefully follow up with more updates soon!

I’ve sent the filament off to Sarah Tamsin for printing, some small adjustments have been made to the flange on the inside of the components. The fit was very, very snug with 34mm drainpipe, I’m wondering if forcing the pipe into these components cracked them, if the forces involved with the sound pressure created the cracks, or a combination of both. In any case, I opened up the STL templates for the horn, identified a flange there, and did a boolean subtraction using a cylinder exactly 34mm in diameter. This should give us a very snug fit without any cracks or air leaks.

More to follow!