CMS390 series orthogonal combi-sensor

The CMS390 is designed for orthogonal side mounting rather than flat placement. In this orientation, its angular-rate Z axis and acceleration X axis are parallel to the PCBA, while the acceleration Y axis is perpendicular to the board. This configuration can simplify an all-axis sensor arrangement when the device is combined with a flat CMS300 on the same assembly. The maximum stated rate- and acceleration-sensor misalignment is ±3% relative to the package or mounting face. Designs requiring tighter alignment should use controlled placement fixtures, assembly inspection and processor-based correction of residual cross-axis error.

The appropriate settings depend on the maximum expected motion and the required signal resolution. The ±150°/s rate range provides 0.005°/s typical resolution, while the ±300°/s setting provides 0.01°/s resolution and more headroom for faster rotation. Acceleration can be set to ±2.5g with 0.079 mg typical resolution or ±10g with 0.313 mg typical resolution. Nominal gyroscope bandwidths are 45, 55, 90 and 110 Hz, while the acceleration channels provide 45, 62, 95 and 190 Hz settings. Each channel can be configured independently, allowing bandwidth to be limited where lower noise is more important than fast transient response.

The sensor operates from 2.7 to 3.6 V, although the complete performance specification applies within 3.15 to 3.45 V. Both start-up and steady-state current are specified at a maximum of 8 mA, excluding current used to charge external decoupling components. Full performance is reached after 150 ms typically and 300 ms at maximum. The SPI clock can run from 100 kHz to 7 MHz, with message rates from 1 Hz to 10 kHz. The host processor should allow the start-up interval to complete before relying on calibrated data and should select an update rate compatible with the configured sensor bandwidth.

The CMS390 operates from −40°C to +125°C, but some detailed accelerometer bias limits are stated over the narrower −40°C to +85°C interval. Gyroscope bias variation with temperature is ±1.0°/s typical and ±1.75°/s maximum, while scale-factor variation is below ±1% typical and ±2.0% maximum. The integral temperature channel has an 11 LSB/°C typical scale factor, with offset and repeatability limits of ±20°C and ±5°C respectively. These values make the temperature output suitable for compensation rather than precision ambient-temperature measurement. Applications with strict attitude or navigation accuracy should characterise the complete mounted assembly and apply host-level correction across the expected thermal profile.

The gyroscope has 0.06°/s typical in-band RMS noise measured to 45 Hz, with a 0.1°/s maximum and an angular random walk of 0.40°/√hr. Its stated bias instability is 10°/hr, so uncompensated integration will accumulate angular error over time. The acceleration channels have 1 mg typical and 2 mg maximum RMS noise in a 45 Hz bandwidth, together with velocity random walk below 0.05 m/s/√hr. These figures suit motion measurement, stabilisation, tilt sensing and lower-cost attitude systems, but filtering and calibration are still required. Final system error will also depend on bandwidth, sample rate, mounting alignment, vibration and the fusion algorithm used by the host processor.

The package measures 10.4 × 6.7 × 2.7 mm and has a mass of approximately 0.6 g. Because the device is mounted on its side, placement control is more critical than for a conventional flat LCC component, particularly where axis perpendicularity influences system accuracy. The operating range is −40°C to +125°C, storage is permitted from −55°C to +150°C and humidity is limited to 90% RH non-condensing. The sensor is rated for 95g operating acceleration over a 6 ms half-sine pulse and 1,000g powered survival over 1 ms. PCB layout, soldering processes, mechanical support and post-assembly alignment checks should therefore be included in the integration plan.

The commanded built-in test applies known responses to the gyroscope and accelerometer channels so that the host can check whether each sensing path is operating within its expected window. At +25°C, the typical gyroscope response is +64°/s in the ±150°/s range and +107°/s in the ±300°/s range. The accelerometer response is typically +1.25g in the ±2.5g range and +6.2g in the ±10g range. Defined response variation is also provided across −40°C to +125°C, allowing temperature-aware pass and fail thresholds to be implemented. Built-in test is useful at start-up and during maintenance, but it does not replace application-level calibration or verification of mechanical alignment.