How to install the accelerometer

    In this article, learn about different methods of mounting accelerometers, their effects on the accelerometer frequency response, and solutions for mounting MEMS accelerometers.

    In order to accurately measure acceleration, it is important to establish proper mechanical coupling between the accelerometer and the system being monitored.

    In this article, we will look at the different ways to mount an accelerometer to a unit under test (UUT).

    Accelerometer Mounting: Why Is It Important?

    As mentioned previously, it is important to consider the mechanical coupling between the accelerometer and the system.

    A common source of error in accelerometers is the resonance of the mounting fixture.

    For example, when using a microelectromechanical systems (MEMS) accelerometer, the PC board to which the accelerometer is mounted, as well as any other mechanical interfaces between the PCB and the object being monitored, can create a mechanical system with multiple resonances within the frequency range of interest.

    If the frequency of the acceleration signal is close to the resonant frequency of the mounting structure, the sensor will receive an amplified version of the original acceleration.

    On the other hand, if the mechanical coupling exhibits some attenuation due to damping, the sensor will measure a smaller signal than the actual signal.

    Proper mounting techniques should be used to fully utilize the bandwidth provided by the accelerometer. Mechanical mounting becomes particularly important when measuring acceleration signals above 1kHz.

    Accelerometer Mounting Methods: Stud, Adhesive, Magnetic

    Generally speaking, there are three main methods of mounting, stud mounting, adhesive attachment, and magnetic mounting, which we will briefly cover below.

    Stud Mounting

    Whenever possible, it is best to drill holes in the object and then use screws to secure the sensor to the device under test (DUT).

    The stud mounting provides a strong mechanical connection and is able to transfer high frequency vibrations of the object to the sensor.

    Figure 1 shows how to secure a piezoelectric accelerometer to a device under test using a bolt mount.

    When using stud mounting, the coupling surface should be as flat and clean as possible. It is recommended to use a thin layer of grease, oil or similar coupling fluid between the coupling surfaces, especially above 2kHz. Adding coupling fluid fills small gaps in the mounting surface and improves vibration transmissibility and mounting stiffness.

    In Figure 2, a graph shows the effect of using a grease film on the frequency response of a stud-mounted accelerometer.

    Figure 2. Graph showing the effect of grease on the frequency response of an accelerometer.

    As you can see, when no coupling fluid is used (yellow curve), the resonance frequency occurs at a relatively low frequency compared to the pink curve using a grease film.

    The graph also illustrates the impact of another important factor: tightening the studs to the recommended installation torque shown in the data sheet.

    The studs should be tightened to the manufacturer's specifications using a torque wrench. An under-torqued stud may not adequately couple the sensor to the object and further reduce the resonant frequency of the system (cyan curve in the graph). Over-tightening may also result in damage to the device.

    Adhesive Mounting

    Sometimes it is not possible to drill holes in the structure, or the design of some accelerometers does not allow us to use stud mounting.

    In these cases, we can use adhesives to secure the sensor to the object being monitored.

    The appropriate adhesive, such as epoxy, glue or wax, should be selected to meet the application requirements. Keep in mind that some adhesives are suitable for temporary installations, while others provide a more permanent installation.

    Another option is to use an adhesive mounting base or mounting pad where one side of the base is adhesive mounted to the test object and the other side provides a quality surface to stud mount the accelerometer.

    The solution is shown in Figure 3.

    Figure 3. Example of mounting an accelerometer using an adhesive mounting base.

    Because the mounting pad provides a smooth surface and vertical mounting holes for the accelerometer, it can improve the frequency response of the accelerometer.

    Additionally, adhesive mounting bases prevent adhesive from damaging expensive accelerometers by clogging the sensor’s mounting threads.

    Magnetic Mount

    The final third method of mounting an accelerometer is to use magnetism. Magnetic mounting can be used when the object being monitored has a ferromagnetic surface.

    In the case of non-magnetic or rough surfaces, we can weld or epoxy the steel pad to accept the magnetic base.

    Figure 4 shows the use of a two-legged magnet on a circular surface.

    Figure 4. Example of a magnetically mounted accelerometer. Image courtesy of Wilcoxon.

    Although they may imply ease of installation, magnetic mounts are often heavy, and the added mass can reduce the resonant frequency of the measurement system.

    Frequency Response of Accelerometer Mounting Methods

    Compared to the stud mounting method, adhesive and magnetic mounting provide a relatively loose connection and are therefore suitable for measuring low-frequency acceleration signals. These methods are typically used in applications involving acceleration signals below a few kilohertz.

    Figure 5 shows the frequency response of a given accelerometer for different mounting types.

    Figure 5. Frequency vs. amplitude plot of an accelerometer, depending on the type of mounting.

    With stud mounting (blue curve), the resonant frequency of the system occurs at a relatively high frequency.

    However, with adhesive (black curve) and magnetic mounting (green curve), the frequency response begins to peak at relatively low frequencies. The following table gives typical frequency limits when using different mounting methods.

    Solutions for mounting MEMS accelerometers

    Piezoelectric accelerometers are designed to support the proven mounting strategies described above. This makes installation of these devices relatively simple.

    There are also capacitive MEMS accelerometers, such as the AMA series from Jewell Instruments, which feature rugged designs that can operate in harsh environments and support stud/bolt mounting.

    However, most MEMS accelerometers should be mounted on a PCB. The PCB should then be connected to the object being monitored via sufficiently rigid mounting techniques.

    Some PCB design factors that may affect the overall bandwidth of the system include: the location of the accelerometer IC on the PCB, solder chemistry, and PCB size. In addition, any other mechanical interfaces between the PCB and the object should be checked to ensure that high-frequency vibrations are successfully transferred to the sensor.

    The accelerometer IC should be placed near a stable mechanical mounting location on the PCB. Multiple hard mounting points near the accelerometer IC are recommended.

    Figure 6 shows an example of a misplaced accelerometer.

    Figure 6. Example of incorrect accelerometer placement.

    In the above example, the accelerometer is placed in an unsupported PCB location and measurement errors may occur due to undamped vibrations of the PCB itself.

    Another factor that can affect system bandwidth is PCB thickness. Thicker (but more expensive) PCBs can reduce the impact of board component resonances on our measurements and improve measurement accuracy.

    Another solution to the MEMS accelerometer mounting problem comes from Analog Devices, which has created a mounting cube for securing the PCB to the asset.

    The cube has a central mounting hole that allows the module to be securely attached to the device under test using a #10 machine screw.

    Note that the sensor PCB is designed to be slightly thicker (3mm).