ENGINEERING
New crystal oscillator vendor short-listed
Three TCXO candidates tested at the ±0.0005Hz line. Vendor B is the front-runner — sub-ppb drift over a 4-hour soak. Cost is 2.3× current part; evaluating volume pricing.
Context
Carrier-frequency precision is the single biggest factor in long-session reproducibility. The current TCXO meets spec at room temperature but drifts under thermal load — the same drift we suspect is behind the HRV finding logged 2026.06.10.
Method
- 01Three TCXO candidates (vendors A, B, C) soldered into a reference jig.
- 02Each soaked at 35°C for 4 hours while frequency was logged against a GPSDO reference.
- 03Cost compared on a 1k-unit quote from each vendor.
Findings
- ▸Vendor B drift was sub-ppb over the full 4-hour soak.
- ▸Vendor A drifted 3.8ppb; Vendor C drifted 11ppb.
- ▸Vendor B per-unit cost is 2.3× the current part.
Next steps
Push Vendor B for a volume quote at 5k units. If pricing closes to 1.6×, we move it into the Mk III BOM and re-run the HRV cohort against the new oscillator.