GMSN! Pure Digit

28th July 2018

cc-by 4.0

Rob Spencer

gmsn.co.uk

forum.gmsn.co.uk

Open Source Synth Designs.

Intro

This repo contains both the Library and Sketches for the GMSN! Pure Digit.

A small programmable utility module which accepts +/- 10V in, does some digital processing, and outputs a +/-10V signal.

It has a 2 Channel 12bit ADC In, 12bit DAC Out, a 24 position rotary encoder with switch and a 7 segment LED display.

To install the library

1. Download the zip

2. Follow the Importing a .zip library how to.

Sketches

• VCO
• Noise
• EmptySketch
• RandomSequencer
• Clock Divider
• MicroTuner
• BitCrusher
• Drums
• RhythmPlayer
• SwingMachine
• LFO
• StepLFO
• DCLevel
• Module Testing Functions
• FrequencyCounter
• DCMeter

Methods

encodeVal(int value) - Call this to increment or decrement a value with the rotary encoder.

displayLED(int digit, byte mode, bool dp) - The display has two modes: 1) Standard 0 - 9 digital digit. 2) Positional representation with 12 values. Where the knob is pointing! This uses the result of a modulo calculation so the value can be greater than 12 and it will keep rotating. It can't go less than zero though.

displayOff() - Turns the display off.

dacWrite(int value) - Write a value to the DAC. Yes, it's got a 20V P2P voltage!! :D 0 = +10V 2048 = 0V 4095 = -10V

dacWriteCal(int value) - Write a value to the DAC, but using the calibration values. Use this when you need to ensure the output matches the module input, especially when processing audio or pitch CV.

adcRead(byte channel) - Read a 20V P2P from the DAC, channel 1 or 2. Voltages as above.

dontCalibrate() - If you don't actually need to use dacWriteCal(), then call dontCalibrate() before begin().

calibrate() - Some applications need the output to follow the output as closely as possible. If you have an audio signal coming in, you want the output to be the same. Due small differences in components, there will be both Gain Error and Offset Error that needs to be accounted for. This is a very simple routine which runs once everytime the module is flashed. The calibration values are then stored in EEPROM and used whenever dacWriteCal() is called.

getSwitchState() - This method returns the state of the switch, 1 for pressed, 0 for not pressed.

How to calibrate

If the module hasn't been calibrated, it will run through a self calibration routine. You just need a single patch cable, no other modules or test kit required!

First of all the module will flash "8" a few times, before entering the calibration routine.

1. When it displays "1" patch from the output back into input 1.
2. Long Press the encoder.
3. Patch from output into input 2.
4. Long Press the encoder.

It will then count down to zero and flash. The module is now calibrated.

N.B. The calibration settings are stored in EEPROM, which is overwritten every time the module is flashed, therefor the calibration routing will run everytime the module is flashed.

Example Sketch

// A random sequencer with variable delay

#include <Digit.h>                    //Include the Digit library

Digit digit;                          //Create a Digit called digit. This object now has the methods above available to it.

//Setup variables
int encPos = 1;                       //Just some variables for us to use in the code

void setup() {
 digit.begin();                       //Tell digit to configure its self.
}

void loop() {
   encPos = digit.encodeVal(encPos);  //Increment or decrement the varible encPos
   digit.displayLED(encPos, 2, 0);    //Display the value of encPos as a Positional Representation with no decimal point
   digit.dacWrite(random(4095));      //Output a random value between 0 and 4095, which is analogue +/- 10V
   delay(encPos);                     //Delay for encPos milliseconds.
}