I performed some rudimentary vibration tests on Razor 6 DOF board with Arduino
- which has same gyros and accel as ArduIMU (LPR530AL and LY530ALH, ADXL335). In the above link you'll find complete description of test setup and bunch of images and graphs.
- Razor / Arduino, which is bulkier then ArduIMU, was placed inside HawkSky ( a la EasyStar) and tests were performed on the bench. I tried, as much it's possible on the bench, to simulate flight condition (i.e. no rigid connections to bench)
- there were three groups of tests - no padding, "IMU" padding, and "IMU" + motor mount padding
-Accelerometer data, gyro non-amplified signals, and gyro 4x amplified signals were recorded for part throttle and full throttle
- High Pass filter should probably be changed / eliminated - as noted in other threads by Andrus and Doug W.
- I wouldn't mind if bigger capacitors for low pass filters (accels and gyros) were put in place
- Padding matters. Location of padding + board placement matters.
(some previous data on Razor at http://diydrones.com/profiles/blogs/razor-6dof-imu-arduino-dcm