Activating this link will open 2 new browser tabs.
Products Mentioned
ADXRS610Obsolete
The ADXRS610 is a complete angular rate sensor (gyroscope) that uses the Analog Devices, Inc. surface-micromachining process to create a functionally complete...
Datasheet ADXRS610 on Analog.com
The user inquired about the cross-axis sensitivity of the ADXRS610 gyroscope. ADI does not specify this value explicitly in the datasheet but maintains tight process controls to ensure it is better than ±5% max, typically 1% or less. Additional factors like package tilt can contribute to cross-axis sensitivity, and a 3-axis tumble test can help calibrate the sensor for high accuracy. For IMU products like the ADIS16334 and ADIS16488, the orthogonal alignment is specified to 0.05 degrees, equating to a cross-axis sensitivity of ~0.09%.
AI Generated Content
Does ADI spec cross-axis sensitvity for their Gyroscopes? The Gyro I'm interested in is the ADXRS610.
ADI does not explicitly specify cross-axis sensitivity for the ADXRS6xx family of gyroscopes. if we did, then this value would be in the specifications table :)
However, ADI has a significant amount of process controls in place which to ensure that the cross-axis sensitivity for the device should be better than ±5% max. most gyroscopes will of course exhibit much better performance than this, typically 1% or less. the gyroscope structure itself contributes very little in the way of cross axis sensitivity (the coriolis force is very picky in how it response to rotation). the primary concern for guaranteeing good performance in this regard is tight control over the assembly of the die in the package. any tilt of the die will instigate cross-axis rate sensitivity.
ADI maintains tight control over cross axis sensitivity by implementing assembly process controls and process monitors to ensure that the tilt of the die in the package is within acceptable limits.
Just to add to what Matt said - Any tilt of the package versus the reference plane will also contribute to the cross axis sensitivity. If you add 1° of package tilt in your assembly process you add SIN(1°)=1.7% cross axis. Of course, the package tilt error is uncorrelated to the internal die tilt error, so you would not just add them up. Rather, you take the root sum of squares.
Actually when users are really looking for high cross axis accuracy they normally perform a 3 axis tumble. That is. a rotation of the device in the three reference planes. You measure the response of the gyro to these three rotations and you can back out what the actual position of the sensor die is. Normally you then process this (through matrix operations) to mathematically rotate the sensor data to line up with reference planes.
Just a quick note. When considering IMU products, such as the ADIS16334 or ADIS16488, the orthogonal aligment specification provides insights into this behavior. In the case of our IMU products, this internal frame is calculated to an alignment of 0.05 degrees, which equates to a cross-axis sensitivityof ~0.09%. This is 20-40x better than most assembly proceses can provide, without correction.
