The picture above shows the prototype I completed in preparation for the 2018 One-Handed Musical Instrument Trust (OHMI) conference. If you look closely, a few features have been added that were not previously discussed. For instance, there are 3.6mm jack connectors which provide a means for connecting standard accessibility switches to the device, controlling playback, record, loop and playback/record speed. This is a fairly straightforward feature to add, but I’m aware that the overall cost of the bill of materials will increase with the addition of this hardware. I’ve found a source of low-cost 3.5mm jack sockets from eBay (link) although there’s no mention of RoHS compliance which, to be frank, does make me feel a little uneasy about the toxicity substances and materials used in the fabrication of this part. This is is an area I have little knowledge of, but nonetheless, I’ll drop the seller an email to see if they can provide any further information about the part.
As you can see from the picture, the wiring has become quite complex, which resulted in it being quite difficult to debug the circuit. I’m pleased to say that the prototype functions well aside from an issue with audio quality. Essentially the circuit is suffering from high-frequency interference again. I believe this is emanating from the ATMega368p which is being used to provide the touch sense capability. I’m hoping that this can be rectified strategic placement of decoupling capacitors but this is difficult to test with the current spaghetti junction of wires on the board.
I’m pleased to report that the DIY vactrol works as expected, providing control over playback and record speed. It was fashioned from a small piece of PVC pipe which was eventually covered in heat-shrink tubing. The LED and photo-resistor were simply glued in place using a glue gun.
I had hoped that I would by now be in a position to create a series of prototype boards to test in a workshop environment. I have been working on a board layout in Eagle, but have found the number of connections (or traces) required of the circuit to be challenging to fit on a single sided board. I’ve discovered a capacitive touch sense library for Arduino which allows a single Analogue pin to be used for each sensor rather than two pins, plus additional passive components. Use of this library would reduce the number of traces required by the circuit. Additionally, I would like to explore removing the ISD1820 chip entirely. It’s possible to record audio directly to the ATMega368p (as demonstrated by Bastl), at better quality than the ISD1820. There would then be options for adding additional audio effects which could prove to expand the musical possibilities of the device. Of course, I need to be mindful of not overcomplicating the device.