Sandia National Laboratories’ Primary Standards Laboratory is pursuing new sensing technology to address a long-standing accuracy bottleneck in non-DC high-voltage measurement. The work aims to develop an alternative to the resistive divider that currently serves as the reference standard, potentially reducing uncertainty in measurements above 10 kV. Researchers say the effort could also strengthen international standards for high-voltage reference measurement systems.
High-voltage calibration currently relies heavily on the NIST-N1 resistive divider developed by the National Institute of Standards and Technology. As the principal reference for calibration transfer standards, NIST-N1 has an expanded measurement uncertainty of 0.5% at a 95% confidence level.
For pulse voltage measurement, that figure has become the dominant source of overall uncertainty. Because NIST-N1 is the only currently available device for high-voltage pulse measurement, an alternative system capable of measuring voltage directly is significant for strengthening International Electrotechnical Commission standards in high-voltage reference measurement systems.
The research team compared several sensing mechanisms, including free-space optical systems using the Kerr or Pockels effect and fiber-optic sensors. Fiber-optic sensors reduce alignment complexity but remain affected by temperature effects and measurement errors caused by path-length differences. Surface acoustic wave sensors offer high sensitivity, but because their substrates are typically less than 1 mm thick, they are difficult to scale for applications above 10 kV.
Bulk acoustic wave sensors use thicker piezoelectric crystals, making them better suited to high-voltage signals. The Sandia work extended BAW sensor applications from a previous 1100 V DC range to DC, AC and pulsed voltage signals. By comparing two lithium niobate crystal orientations, researchers analyzed how acoustic mode propagation affects sensor performance.
The expanded testing moves the technology beyond steady direct-current measurement and toward the signal types encountered in high-voltage pulse and alternating-current environments. That shift is central to the laboratory’s effort to develop a practical alternative to existing reference standards.
The study used a one-dimensional impedance matrix method to model the crystal, incorporating two acoustic ports and one electrical port. Using piezoelectric constitutive relations, researchers converted voltage-induced strain into an incremental change in crystal length, establishing a direct link between voltage and strain.
The approach provided a theoretical basis for measuring DC voltages from 256 V to 1100 V and supported development of a complete uncertainty budget for the measurement system. The modeling framework is intended to help assess both the performance and the limits of the BAW sensing approach as work advances toward higher-voltage applications.
| Parameter | Detail |
|---|---|
| Current high-voltage reference | NIST-N1 resistive divider |
| Measurement uncertainty | 0.5% at 95% confidence |
| Target application range | Above 10 kV |
| BAW prior demonstrated range | 1100 V DC |
| BAW expanded signal types | DC, AC and pulsed voltage |
| Modeled DC range | 256 V to 1100 V |
| SAW substrate thickness | Typically less than 1 mm |
