Hexarhythm

synaesthesia · Last edited by synaesthesia on Sun Aug 17, 2014 2:10 am; edited 1 time in total

It was a Friday evening and nothing else to do, so I took the breadboard out and plugged a circuit together that I had on my mind since a few days. It uses only two chips and but can generate thousands of different melodies (I call them patterns in the remaining description).

It is not really a sequencer, although you can select the base pattern played by it. Two independent counters generate four bits, of which one to four are OR-ed together to enable one of the four oscillators. The oscillator frequencies are pre-set in ratios 1:2 and 2:3. This is an octave and a perfect fifth. The pattern repeats after 16 clocks. By pressing button SW1 to reset just one of the counters, they can be brought out of synchronization, which varies the rhythmic pattern again. Finally, by pressing modifier buttons HIGH and LOW (SW4..SW7), two of the oscillators can be tuned up or down temporarily. All of this together makes a simple device that generates thousands of possible tone patterns.
If my math is right (it was wrong, edited), there are 16*16^4=1048416 possible patterns, not including using the modifier buttons.

I have used DIGI switches in the schematic, but of course this can also be patched on the breadboard, or any other switches may be used. The small blue markers are at the connections that I used for the recordings. The condensators and resistors may be varied as long as the ratios are kept the same. The 22K and 33K resistor values are important for the 2:3 ratio. I ran it at 5V, but as it is all CMOS, this can be varied.

To give you an idea of the variations I made four recordings. The first one simply varies the speed of the counter clock. The second modifies the pattern with the HIGH and LOW buttons. The third varies the pattern by changing the counter bits that drive the oscillators. I changed only one bit at a time and always kept two bits per oscillators selected. Many more combinations are possible. The fourth recording keeps the selected pattern stable, but brings the two counters out of sync by holding one of them in reset for a moment, which results in a new rhythm. I edited this recording to capture each pattern four times. And I finally built a cable to record directly to my laptop, which gives much better sound quality.

There is still one inverter left, which calls for an extension. I tried inserting it in the clock line for the second counter, but the generated patterns then sound too arrhythmic. Maybe someone has an idea how to use it to increase the possibilities even more...

Hexarhythm.PNG

Description:

Filesize:

38.88 KB

Viewed:

652 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

Hexarhythm-Speed.mp3

Description:

Download (listen)

Filename:

Hexarhythm-Speed.mp3

Filesize:

420.81 KB

Downloaded:

1038 Time(s)

Hexarhythm-Modifier.mp3

Description:

Download (listen)

Filename:

Hexarhythm-Modifier.mp3

Filesize:

595.92 KB

Downloaded:

976 Time(s)

Hexarhythm-Sync.mp3

Description:

Download (listen)

Filename:

Hexarhythm-Sync.mp3

Filesize:

1.03 MB

Downloaded:

960 Time(s)

Hexarhythm-Pattern.mp3

Description:

Download (listen)

Filename:

Hexarhythm-Pattern.mp3

Filesize:

1.21 MB

Downloaded:

984 Time(s)

commathe · Posted: Sat Aug 16, 2014 7:23 pm Post subject:

Very interesting! Would it be possible to get away with 4 diodes instead of 16 though?

synaesthesia · Posted: Sun Aug 17, 2014 12:53 am Post subject:

Each oscillator has four diodes for the 4-input OR-gate. You could reduce the number of inputs and have only 2 per oscillator. Saves you 8 diodes and 1 DIP-Switch, but brings the number of possible combinations down to 4096.
If you want to try the circuit with only four diodes, you could try to connect one oscillator per pair to Q3 and one to Q4 of the counter via a resistor (try 100K) and patch one of the other counter outputs Q1 to Q4 via a diode.

commathe · Posted: Sun Aug 17, 2014 1:49 am Post subject:

looking at it again. I now realise that the oscillators will only sound when all four diodes are back biased. Didn't really get why there needed to be separate diodes but I can see it now.

RingMad · Posted: Mon Aug 18, 2014 5:29 am Post subject:

Thanks for sharing this... it looks pretty cool! I hope to try it out soon.

-- James.

Tomoroh Hidari · Posted: Mon Aug 18, 2014 8:42 am Post subject:

wow, this one looks and sounds really interesting. Thanks for sharing.