If perfect sensors were available, this might be possible. Variation a sensor's output response for a given set of conditions creates a demand for design marging when it comes to "guarateed range," so classic A/D relationship is applicable in this manner. As an example, the ADIS163xx and ADIS164xxIMU products incorporate a mixed-signal processing system that translate uncalibrated sensors into a well-calibrated, frame-aligned sensor system. In this process, they apply correction formulas to each inertial sensor in their digital processing stage, which means that the A/D input range must accomodate all of the expected variation in each sensor's output signals. This can approach 25% in extreme cases, which means that an +/-300 deg/sec range on a gyroscope requires at least +/-375 deg/sec of equivalent input range on its A/D input. When evaluating the resolution step size of a sensor's output data, the most important thing to understand is its impact on critical performance criteria. In IMUs, the in-run bias stability on the gyroscopes is typically an important performance metric. Since these IMU products typically provide greater than 300 Hz of bandwidth, the total noise is normally higher than the resolution step size of each sensor's digital data. When the noise is higher than the resolution step size, digital post processing techniques can produce additional resolution. The Allan Variance behavior demonstrates this, as it requires sequential averaging of data to characterize noise and stability, with respect to integration time. The summation of data for these averages generates a bit growth, which can impact performance. In many cases, the minima of the Allan Varinace curve (also known as the "In Run Bias Stability) is below the resolution step size of the output register. For example, the ADIS16334 provides a resolution of step size of 0.05 deg/sec while supporting an in-run bias stability that is more than 7 times lower, at 0.007 deg/sec. See Table 10 in the ADIS16334datasheet for reference to the resoultion step size and Figure 7 the Allan Variance, which documents the in-run bias stability performance.

If perfect sensors were available, this might be possible. Variation a sensor's output response for a given set of conditions creates a demand for design marging when it comes to "guarateed range," so classic A/D relationship is applicable in this manner. As an example, the ADIS163xx and ADIS164xx IMU products incorporate a mixed-signal processing system that translate uncalibrated sensors into a well-calibrated, frame-aligned sensor system. In this process, they apply correction formulas to each inertial sensor in their digital processing stage, which means that the A/D input range must accomodate all of the expected variation in each sensor's output signals. This can approach 25% in extreme cases, which means that an +/-300 deg/sec range on a gyroscope requires at least +/-375 deg/sec of equivalent input range on its A/D input. When evaluating the resolution step size of a sensor's output data, the most important thing to understand is its impact on critical performance criteria. In IMUs, the in-run bias stability on the gyroscopes is typically an important performance metric. Since these IMU products typically provide greater than 300 Hz of bandwidth, the total noise is normally higher than the resolution step size of each sensor's digital data. When the noise is higher than the resolution step size, digital post processing techniques can produce additional resolution. The Allan Variance behavior demonstrates this, as it requires sequential averaging of data to characterize noise and stability, with respect to integration time. The summation of data for these averages generates a bit growth, which can impact performance. In many cases, the minima of the Allan Varinace curve (also known as the "In Run Bias Stability) is below the resolution step size of the output register. For example, the ADIS16334 provides a resolution of step size of

0.05 deg/secwhile supporting an in-run bias stability that ismore than 7 times lower, at0.007 deg/sec. See Table 10 in the ADIS16334 datasheet for reference to the resoultion step size and Figure 7 the Allan Variance, which documents the in-run bias stability performance.