MEMS accelerometers have finally reached a point where they are able to measure vibration on a broad set of machine platforms. Recent advances in their capability, along with the many advantages that MEMS accelerometers already had over more traditional vibration sensors (size, weight, cost, shock immunity, ease of use), are motivating the use of MEMS accelerometers in an emerging class of condition-based monitoring (CBM) systems.
As a result, many CBM system architects, developers, and even their customers are giving consideration to these types of sensors for the first time. Quite often, they are faced with the problem of quickly learning how to evaluate the capability of MEMS accelerometers to measure the most important vibration attributes on their machine platforms.
This might seem difficult at first, as MEMS accelerometer data sheets often express the most important performance attributes in terms that these developers may not familiar with. For example, many are familiar with quantifying vibration in terms of linear velocity (mm/s), while most MEMS accelerometer data sheets express their performance metrics in terms of gravity-referenced acceleration (g).
Fortunately, there are some simple techniques for making this translation from acceleration to velocity and for estimating the influence that key accelerometer behaviors (frequency response, measurement range, noise density) will have on important system-level criteria (bandwidth, flatness, peak vibration, resolution).