The Engineering Validation Test (EVT) is where sleek CAD models meet the laws of physics. For OneMira C1, EVT was a process of tracing noise. We had to pack high-sensitivity differential microphones, a lithium-ion battery charging circuit, and a mechanical motor into an enclosure smaller than a deck of cards. Here is the diary of the two acoustic failures that nearly forced us to scrap the turntable design.

Roadblock 1: The 8.2 kHz Power Switching Hum

When we assembled our first batch of EVT v1 PCBs, the microphone recordings were ruined by a sharp, high-pitched hum at exactly 8.2 kHz. Probing the circuit under the microscope, we found that we had routed the power traces of the boost converter parallel to the analog differential microphone lines. The high-frequency switching noise of the charger pump was coupling directly into the high-impedance analog inputs.

To fix this without spinning a new board, we did a brute-force workbench hack:

  • We used a scalpel to slice the parallel copper power traces on the PCB.
  • We ran 30 AWG enamel-insulated fly-wires directly from the regulator's output, routing them along the edge of the board far away from the audio traces.
  • We wrapped copper EMI shielding tape over the microphone preamp stage, grounding it directly to the battery's return pad.

This microscope patch lowered the noise floor by 14 dB, allowing us to continue testing. For the EVT v2 production PCB, we completely isolated the analog ground plane, splitting it from the digital plane with a 1.2 mm gap and connecting them at a single star-ground point.

Roadblock 2: The Motor Vibration Soundboard

Having a physical rotating disk is beautiful, but the motor is a physical noise source. In our first EVT chassis, we screwed the motor directly to the plastic housing. The case acted as a soundboard, amplifying the 120 Hz motor vibrations directly into the microphones. The hum was so loud (-32 dB) that our local transcription engine couldn't transcribe a single word.

We tested three isolation methods to decouple the motor:

  1. EVA Foam Mounts: We cut custom EVA foam washers. This reduced the noise to -42 dB, but the foam compressed unevenly over time, causing the turntable disk to tilt and scratch the outer aluminum trim.
  2. Rigid Dual-Chamber: We tried placing a solid partition wall, but the vibration still traveled through the plastic screws.
  3. Suspended Shore-30A Silicone Gasket: We redesigned the interior chassis to isolate the microphones in a sealed front acoustic cavity. The motor is mounted in the rear chamber, suspended inside a custom-molded Shore-30A silicone ring gasket. This soft gasket absorbs the motor's physical harmonics while maintaining alignment. The hum dropped to -54 dB—completely below room ambient noise.