Guidelines For Selecting An Accelerometer

To select the accelerometer best suited for an application, a number of variables must be considered to assure that the desired measurement result is obtained. Below is a summary of the more important parameters that should be considered followed by detailed information pertaining to each area.

Frequency Range (Hz):

Piezoelectric accelerometers have an upper and lower useable frequency. The upper frequency is determined by the accelerometer’s natural frequency and the lower limit by the time constant of the sensor’s internal circuitry or external charge amplifier. The accelerometer should be used within the flat portion of the response curve. In this range, the specified sensitivity lies within a defined amplitude tolerance band (usually ±5%).

Measuring Range (±g):

For low impedance sensors, selection should be such that the expected peak values of acceleration are within the measuring range. If the magnitude of the measuring range is not precisely known, a ±500g accelerometer may be used to establish the measurement scale. Then an accelerometer with the applicable range can be selected.

Alternatively, it is possible to use a high impedance (charge mode) accelerometer with a charge amplifier to resolve vibration amplitudes over several decades of g levels.

Acceleration Sensitivity (mV/g):

The sensitivities listed in catalogs are nominal values. A calibration certificate containing the exact value is provided with each sensor. Since the available full scale voltage is ±5 volts, the sensitivity can be determined by dividing the expected acceleration range into the full scale voltage. For example, a ±50g range indicates a sensitivity of 100mV/g.

Operating Temperature Range:

(°C). High impedance sensors can be used up to 250°C without problems. Due to the limitations of internal electronics, low impedance accelerometers are operable up to 165°C as a maximum.

Ground Isolation:

If current loop problems are likely to occur, ground isolated accelerometers are recommended. For example, the ignition of a car engine often causes such problems. This condition can be prevented by selecting a ground isolated unit or installing an adhesive mounting pad under the accelerometer.

Addition of Mass:

Adding mass (such as an accelerometer) to a vibrating structure can alter the frequency of the vibration. This is sometimes referred to as “mass loading”. As a general rule, the mass of the sensor (and mounting accessories) should not exceed 10% of the mass of the vibrating structure.

DETAILED SELECTINION GUIDELINES

The process of selecting an accelerometer for an application encompasses a number of different factors–which are summarized above. Accelerometer designs incorporate a variety of technologies engineered with certain qualities for tailoring the internal parameters towards a specific measurement goal. Kistler accelerometers can be fundamentally classified into three groups which are differentiated by the type of signal conditioning required. These three groups are low impedance piezoelectric, high impedance piezoelectric and variable capacitance.

TECHNOLOGIES:

Low Impedance or voltage mode accelerometers contain an internal impedance converter which transforms the high impedance voltage from the sensing element into a usable low impedance voltage signal. These signals are essentially unaffected by triboelectric noise or EMI often generated by cable motion or the environment.

High Impedance or charge mode accelerometers are used in conjunction with an external charge amplifier and highly insulated, lownoise cabling is used to isolate the high impedance signals from the environment.

Variable Capacitance or K-Beam® accelerometers utilize MEMS technology and the associated integral ASIC conditions the output to a manageable low impedance voltage. Kistler has categorized accelerometers in the product catalog to aid in the selection process. An understanding of important accelerometer characteristics will assure the best choice is made for your specific application.

Following, is a list of important parameters for proper accelerometer selection. Reference is made to a particular technology when appropriate.The amplitude measuring range, or expected g level, combined with the frequency domain of interest are usually the governing parameters during the selection process. However, many other factors such as mounting considerations or type of environment extremes may influence your selection.

ACCELERATION RANGE:

Low Impedance, voltage mode or Piezotron® accelerometers use signal conditioning which limits the system voltage levels to a typical ±5 volt level. This defines the measurable acceleration range which is directly related to the voltage sensitivity of these low impedance accelerometers. Their fixed voltage sensitivity together with the 5 volt full scale output defines the measurable range. For example, a 1000mV/g accelerometer is usable throughout a 5g range considering the 5V limit. A 10 mV/g accelerometer will have a 500g range.

High Impedance or charge mode accelerometers are much more flexible regarding accelerometer range because the system voltage sensitivity can be changed by the external charge amplifier. With these high impedance accelerometers, a voltage sensitivity is selectable between 0.1mV/g and 1000mV/g and a 10V system limit is imposed by the charge amplifier. A selected 1000mV/g sensitivity can measure 10 g’s, considering the 10V limit.

However, a system adjusted for a very low voltage sensitivity is limited in acceleration range by the mechanical construction of the sensor itself. The charge mode accelerometer is very useful for a Test Laboratory environment since a wide variety of ranges can be accommodated by a single sensor.

K-BEAM® or variable capacitance based accelerometers also have a fixed voltage sensitivity and their system voltage limitation is typically one volt. These accelerometers are optimized for very low frequency measurements, including DC, and their available amplitude ranges are limited to below 50g. An 8352A1 accelerometer has 1000mV/g sensitivity and a range of 1g while an 8352A50 has 20 mV/g sensitivity and 50g range.

FREQUENCY RANGE:

An accelerometer’s sensitivity is calibrated at a specific frequency (commonly 100Hz) and its deviation from this reference sensitivity is presented typically by stating its 5% or 10% deviation limits in terms of frequency. For instance, a piezoelectric accelerometer’s –5% pt is often near 0.5Hz and its +5% pt is often near 10kHz. See Figure 1 below.

The usable frequency range or frequency response of an accelerometer is the frequency band where the sensitivity is within the stated deviation limits. The accelerometer should be chosen so that its frequency range encompasses all the significant signals for the intended application. Variable capacitance based sensors have the same sensitivity at low frequency as their static sensitivity (for example, DC response). The high end frequency response is dependent on the resonant frequency of the accelerometer and, for piezoelectric accelerometers, is typically one-fifth of resonance. Either electrical filtering or mechanical damping may be employed to attenuate resonant effects. This extends the usable range for many applications. It should be noted that the frequency response characteristic can be very well defined and does not change. The “usable frequency range” can be increased significantly if relative measurements are of interest. Trend analysis measurements can utilize a significantly larger frequency band than the typically defined 5% limits.

SENSITIVITY:

The signal output from an accelerometer is typically voltage or charge and is proportional to an applied stimulus. Accelerometer sensitivity is the constant which defines the relationship between the input and output signal and is commonly provided in terms of mV/g or pC/g. Low impedance sensors are typically chosen for maximum sensitivity within the expected input acceleration range.

MOUNTING CONSIDERATIONS:

Frequency response specifications are determined when the accelerometer is mounted in a manner appropriate to the applications in which it is most often employed. Miniature accelerometers often attached with a cyanoacrylate adhesive such as Super Glue, while shock accelerometers must be stud mounted. It may be more convenient to mount the accelerometer with petro wax, magnet or possibly double-sided tape. These techniques can be used for convenience purposes but the effect should be investigated prior to acceptance of the data.

Figure 2, shown below, the typical effect of various mounting techniques on the same accelerometer. Triaxial accelerometers are available which optimize the performance (weight, frequency response, etc.) for measurement in three orthogonal directions. Mounting cubes are also available to accommodate attachment of three, single axis accelerometers, thereby, providing flexibility when only unidirectional data is required.

GROUND ISOLATION:

If current loop problems are likely to occur, ground isolated accelerometers are recommended. This condition can be prevented by selecting a ground isolated unit or installing an adhesive mounting pad under the accelerometer. These pads are made of aluminum and are hard anodized to form a very durable and electrically isolated interface. An accelerometer with integral isolation will have frequency response characteristics as specified; however, addition of a mounting pad accessory will slightly degrade the usable frequency range and will increase the weight at the measurement site.