Water-Modulated Pico Synth Design


"Splash Pico Synth Board Review" completed.
The final check found a mirrored pin mapping in the MIDI socket footprint—an issue normal routing checks cannot detect. I’m correcting it and the mounting-hole sizes before rerunning the checks.
Thought for 17 mins 53 sec
Thought for 19 sec
Thought for 8 sec
Awaiting 1 task
"Splash Synth PCB Layout" completed.
Thought for 14 mins 50 sec
Thought for 8 sec
Managing 1 task
Thought for 10 sec
Reading "prototype-release-review" project doc.
Thought for 28 sec
Thought for 7 sec
Awaiting 1 task
"Splash Synth PCB Layout" completed.
The splash-shaped PCB is placed and fully routed, with opposite-side electrode connections, accessible MIDI/USB/audio ports, and white-mask/silver-colored finish intent; checks show zero airwires and no blocking native physical DRC errors.
Prototype fabrication files and findings are in the Release Review, but ordering remains on hold for one rounding-level exported-clearance finding and passive-part sourcing; synth firmware and bench testing are still outstanding.
Next priorities are fabrication clearance signoff, final component selection, then firmware and prototype bring-up.

Board Bring-Up Plan
Status: planned, no assembled-board tests performed. Use Electronics Handoff for pin map, expected nominal sensing values and caveats.
  1. Before fabrication: complete splash layout, inspect exact connector/diode pinouts and footprints, resolve BOM effective capacitance, achieve zero airwires/shorts and close all blocking physical DRC. Record VBUS carrier/ERC waiver if still present.
  2. Before power: inspect solder polarity/bridges. Measure3V3-to-GND for shorts. Verify DIN4/5 input domain has no connection to local GND, USB or audio sleeve. Leave bowl and speakers disconnected.
  3. Power using current-limited5V USB bench adapter or suitable USB supply. Stop on heating or excess current. Confirm3V3 nominal3.3V within device operating ranges. Record USB current at startup/idle/audio/sensing; investigate any external3V3 load exceeding100mA allocation. XSMT must remain low while firmware initializes.
  4. Flash test firmware through accessible Pico USB/BOOTSEL. First send zeros with I2S48kHz/32-bit stereo slots; scope GP18 BCK3.072MHz and GP19 LRCK48kHz, data delay and logic levels. Confirm clocks/rails stable before unmute. Check negative charge-pump rail and LDOO against exact DAC datasheet before attaching speakers.
  5. Connect >=10k line load or powered mixer, initially low gain. Send low-amplitude1kHz duplicate samples. Scope separate tip/ring waveforms, matching amplitudes and low DC. Listen only after DC and startup transient checks pass. Never use headphones as test load.
  6. Test isolated DIN MIDI at31250baud8N1. Verify GP1 idle high, correct bytes, no stuck notes after cable removal, velocity-zero, sustain and all-notes-off. Exercise low-current3.3V and conventional5V sources; scope opto delays at3.3V supply. If margins fail, redesign receiver before claiming MIDI compatibility.
  7. Test sensing first dry with open circuit, then short J3-J4 and known100k/1M resistor substitutes. Confirm every ADC remains within rails, polarity swaps as expected and nominal differential decreases with lower resistance. Confirm no drive path bypasses150k and electrode current stays near calculated microamp levels. Sample settling with actual cable; adjust dwell rather than accepting transition samples.
  8. Only then use passive corrosion-resistant electrodes in a separate bowl, preferably USB battery powered. Board/cables/connectors outside splash zone. No body contact, mains connections or other powered immersed objects. Verify timbre varies smoothly while pitch/gates remain MIDI-controlled. Record noise, drift, electrode material and observed corrosion.
  9. Verify boot, firmware crash/watchdog and unplug/replug behavior: startup mute, no prolonged electrode DC, no stuck notes or uncontrolled full-scale audio. Bench success does not certify waterproofing, ESD or safety compliance.