PVT Guide: Testing a Quiet Turntable Indicator
A production-validation framework for a moving recording indicator, covering tolerances, vibration, acoustic fixtures, end-of-line tests, and traceability.

PVT must prove that a production line can repeatedly build and detect a quiet, visible mechanical recording indicator—not merely that one tuned prototype works. The acceptance system needs controlled limits, capable fixtures, traceable measurements, and a reaction plan for failures.
What makes a moving indicator a production risk?
The C1 turntable performs two jobs: it communicates recording state and it rotates near two sensitive microphones. Variation in molded parts, shafts, bearings, gears, fasteners, adhesives, motor drive, and final assembly can change speed, runout, contact, current, and vibration. A unit can look correct while producing a tonal artifact in recorded speech.
PVT therefore needs a chain from design characteristic to measurement:
| Characteristic | Possible defect | Production evidence |
|---|---|---|
| Axial/radial clearance | Contact, wobble, intermittent stall | Gauge or displacement measurement |
| Rotation speed | State cue too fast, slow, or unstable | Optical timing over a defined interval |
| Motor current | Binding, damaged gear, wrong preload | Current profile during start and steady state |
| Structure-borne noise | Gear or imbalance tone enters microphones | Fixture spectrum and reference recording |
| Visual state | No motion, wrong direction, occlusion | Camera or operator functional check |
| Cover clearance | Scrape, optical distortion, cosmetic rub | Go/no-go and visual standard |
How should the tolerance stack be built?
Start at the functional gap, not at individual drawing dimensions. Identify every contributor between the rotating part and fixed cover: datum selection, housing flatness, boss height, bearing or shaft fit, gear stack, adhesive thickness, and assembly force. Evaluate worst-case limits and statistical variation separately. Worst-case analysis protects interchangeability; statistical analysis predicts typical yield only when the process distribution is known.
Measurement systems need their own validation. If a gauge’s repeatability and reproducibility consume a large part of the tolerance, production data will classify good and bad units inconsistently. PVT should include fixture correlation, calibration, golden units, and a method for checking drift at the start of a shift.
How do you build an acoustic end-of-line test?
A useful fixture controls source signal, device location, room or enclosure response, microphone path, and analysis window. Run a known speech or broadband stimulus with the mechanism off and active. Compare defined spectral bands and the resulting recorded file against approved limits and reference units. The goal is not to make a factory room acoustically perfect; it is to create a repeatable discriminator.
Do not let software conceal a mechanical defect. If aggressive noise suppression removes a gear tone in the final transcript, also inspect a pre-suppression or diagnostic capture. Conversely, do not reject a unit solely because a generic sound-level reading includes harmless room noise. The test must target the defect mechanism.
Which stations should verify the complete product?
- Incoming and subassembly checks: critical part identity, dimensions, motor/gear behavior, battery and board traceability.
- In-process checks: connector alignment, microphone seals, fastener torque or controlled fastening result, moving clearance.
- End-of-line functional test: boot, controls, LED/state feedback, turntable motion, microphones, storage, Bluetooth, USB-C, charging, and identifiers.
- System transaction test: record a known clip, transfer it, verify the computer receives a decodable file, commit the note path, and confirm the intended device state.
- Audit test: periodic deeper acoustic, battery, radio, and mechanical sampling that would be too slow for every unit.
End-of-line tests should save a compact result keyed to the unit identifier: station, fixture, software version, measurements, pass/fail, and failure code. Avoid saving customer-like voice content; use a controlled fixture signal.
How do you establish process capability?
Capability statistics are meaningful only after the process is stable and the specification limits are justified. Plot distributions by cavity, line, supplier lot, fixture, and shift. Investigate multimodal data instead of compressing it into one average. A high overall pass rate can hide one cavity that repeatedly produces marginal clearance or one fixture that drifts.
Run production-intent samples through temperature, handling, charging, and transport conditions, then repeat the functional and acoustic checks. The mechanism should not pass only before settling, wear-in, or shipping vibration.
What is the reaction plan when a unit fails?
A failure code must point to a containment action. Stop or quarantine the affected scope, confirm the fixture, preserve the unit, identify the last known good point, and trace common lots or stations. Rework needs an approved instruction and a full retest of functions that the rework could disturb.
Never “tune until pass” without recording the adjustment. That practice erases process information and can produce units with no known margin. PVT closes when the line, fixtures, limits, work instructions, traceability, and reaction plan all operate together.
How does PVT connect to user-visible claims?
Public claims such as up to 30 hours of recording, approximately 3 m pickup range, two microphones, Bluetooth 5.4, and USB-C need controlled verification outside the cosmetic inspection flow. Manufacturing documents can contain tighter internal limits, but public content should use only approved, repeatable product specifications.
For the earlier root-cause method, see the EVT acoustics and PCB guide. For the full capture-to-delete state matrix, see the DVT local transcription guide.
Sources and further reading
- Microsoft audio-device design guidance — official capture-quality and microphone integration considerations useful for fixture design.
- IETF RFC 6716: Opus — the codec definition used to verify that production test files are structurally decodable.
- OneMira C1 specifications — approved public dimensions, mass, materials, battery, microphones, connections, and processing boundary.