When the MeeBlip synthesizer was first introduced, I was excited to recieve the micro version of the device, hoping to build a cool new device to jam, but the sound didin't inspire me a lot and my knowledge of synthesis and asm was too little to change this.
Now that the MeeBlip Anode has been around for quite some time, still marketed as hackable and "covered by a permissive Creative Commons and GPLv3 license. ", I feel it's time to actually hack it and see what happens if you swap the AtMEGA32 with a TI MSP432 LaunchPad.
I believe that using a RTOS could allow cleaner code for examle by dividing between UI and Synthesis in different threads. My plan is to design a BoosterPack including the Twin-T filter, MIDI Interface and Audio Output while allowing the use of other BoosterPacks to create the UI.
In addition to the original schematics, there will be an option for MIDI output, as well as digital control of the filters resonance and the neccessary 3.3V to 5V le
I decided to split the Project into three seperate Boosterpacks, called "synth", "knobs" and "midi" to provide higher granularity. You can find the schematics in my github, also I uploaded the board files. Since I am just begining to work with eagle, my board design is far from perfect, and this is to be considered WIP. If you want to help with the layout, please send me a pull request.
Seperate BoosterPack details:
Synth
12-bit DAC
Twin-T Filter
PWM Cutoff control
Resonance control via digital potentiometer
Knobs
8 Potentiometers
4 Switches
ToDo: Add multiplexer to save pins, maybe a tiny display
MIDI
MIDI In/Out
Gate/Trigger In/Out
PWM CV Out
Everything provided as 3.5 Audio Jacks
ToDo: Add ADC for Audio or 1 V/Oct input
If all three Boosters are finished, it should be possible to stack them to get a complete Synthesizer.
As there is already a SPI device in the original design (MCP4922 DAC) I want to add a 10k digital potentiometer(MCP41010) to replace the resonance poti. This would allow MIDI automation, patch memory or different response (lin/log) while only using one additional pin for the slave select.
Speaking of pins: Since the TI Launpad is running at 3.3V, but the audio circuit is powered by 5V, some sort of level conversion should be done. Luckily, this is not only easy, but also well documented by - for example - sparkfun.
SDA and SCL need to be bi-directional, SCK and the chip select lines on the other hand can be protected by a simple voltage divider.
I will try to finalize the schematics tonight and provide a git repo, so you can see what I am talking about.