text.skipToContent text.skipToNavigation



TE Connectivity


By Bjorn Ryden, Global Product Manager, Vibration Sensors, TE Connectivity

In the past, accelerometers were primarily used for heavy, high-end machinery such as windmills, industrial pumps, compressors and Heating, Ventilation and Air-Conditioning (HVAC) systems.

Driven by increased automation, demand is rising for use in smaller systems produced in higher volume, such as machine spindles, conveyor belts, sorting tables and machine tools which require better predictive maintenance.

Machine downtime in these applications has an important effect on the customer experience and on profitability. Digital systems for condition monitoring can help extend operating lifetimes and eliminate unplanned downtime. An accelerometer is a key component of condition monitoring circuits. This Design Note compares the principal accelerometer technologies used today in industrial condition monitoring systems.

Key Performance Indicators of Vibration Monitors

For industrial condition monitoring and predictive maintenance applications, a small number of vibration measurement parameters are of critical importance.

Wide frequency response – to detect all possible failure modes of an electric motor, the frequency response of the accelerometer should be 40 to 50 times the motor speed expressed in revolutions per minute (rpm). For fans and gearboxes, the minimum upper limit of the accelerometer should be 4 to 5 times the blade passing frequency.

Resolution and dynamic range – the resolution of the vibration sensor is a function of the amplitude of the output signal to the broadband noise of the onboard electronics. An accelerometer with a superior signal output will allow the measurement of smaller vibrations. A sensor which can measure vibration of a lower amplitude enables the end user to predict a fault earlier than a sensor which has a lower dynamic range. As a general rule, reliable measurement calls for an output signal 10x stronger than the noise generated by the sensor.

Long-term stability – drift is a shift in the sensitivity and/or the measurement output when the input is zero. A shift in the sensitivity of the accelerometer could over time lead a monitoring application to issue a false alarm. A shift in the zero-output measurement will have the same effect. Piezoelectric sensors do not provide a DC response, so they are not susceptible to zero-output drift, only to sensitivity drift. A MEMS accelerometer can suffer from both zero-output drift and sensitivity drift over time.

The Two Principal Types of Accelerometer Technology

Piezoelectric accelerometers incorporate piezoelectric crystals which supply a signal when stressed by external excitation such as vibration. Most piezoelectric sensors are based on lead Zirconate Titanate (PZT) ceramics which are polarized to align the dipoles and make the crystals piezoelectric. PZT crystals are ideal for condition monitoring applications since they offer a wide operating-temperature range, broad dynamic range, and wide frequency bandwidth of >20kHz.

Variable Capacitance (VC) vibration sensors derive their acceleration measurement from a change in capacitance of a seismic mass moving between two parallel capacitor plates. The change in capacitance is directly proportional to the applied acceleration. VC accelerometers require an IC to be closely coupled to the sensing element to convert the very small capacitance changes into a voltage output. This conversion process can result in a poor signal-to-noise ratio and limited dynamic range.

VC sensors are typically manufactured from silicon wafers and are fabricated into miniature Micro-Electromechanical Systems (MEMS) chips.

Technology Comparison

Tests performed by TE Connectivity (TE) reveal the important differences in performance between the two types of accelerometer. The tests were conducted with a piezoelectric and a VC accelerometer which both had a full-scale range of ±50g.

Frequency Response

The frequency response of the two accelerometers was tested on a SPEKTRA CS18 HF high-frequency shaker with a range of 5Hz to 20kHz. The sensors were securely mounted to ensure accurate results over the full test range, as shown in Figure 1. Three sensors of each technology were tested.

A maximum ±1dB amplitude deviation is assumed as the usable bandwidth, although a tighter deviation of ±5% is often used for bandwidth tolerance.

The data indicate that the VC MEMS sensor has a usable bandwidth up to 3kHz, while the piezoelectric sensor has a usable bandwidth of >10kHz.

It is worth noting that the low frequency cut-off for the piezoelectric sensor was at 2Hz, while the MEMS sensor operates down to 0Hz since it is a DC-response device.

Fig. 1: Comparison of the typical frequency response of piezoelectric and MEMS accelerometers

Measurement Resolution and Dynamic Range

To determine the measurement resolution and dynamic range of the piezoelectric and VC MEMS sensors, the samples were tested in a noiseisolated chamber with high-resolution measurement equipment. The units were installed in the same chamber and tested at the same time to eliminate any errors from outside environmental interference.

The measurements were conducted at four distinct bandwidth settings, and the residual noise was measured at each setting, as shown in Figure 2.

The measurement resolution and dynamic range were calculated based on a 0.03 to 10kHz bandwidth, as shown in Figure 3. The resolution of the piezoelectric sensors is around nine times better than that of the VC MEMS sensors. This results in a markedly better dynamic range, which enables the end user to detect potential problems much earlier.


Model 0.03 ~ 300Hz μVrms 0.03 ~ 1KHz μVrms 0.03 ~ 3KHz μVrms 0.03 ~ 10KHz μVrms
Piezoelectric #1 27.2 30.8 39.5 57.6
Piezoelectric #2 25.1 31.7 38.6 56.3
MEMS #1 377.6 405.2 412.7 498.2
MEMS #2 415.7 430.2 453.9 532.1

Fig. 2: Measurements of noise tests on piezoelectric and MEMS accelerometers



  Resolution Residual Noise Spectral Noise Dynamic Range Resolution
Model mgrms μVrms μgrms/√Hz dB Bit
Piezoelectric #1 1.4 57.6 14.4 88 14.6
Piezoelectric #2 1.4 56.3 14.1 88 14.6
MEMS #1 12.5 498.2 124.6 69 11.5
MEMS #2 13.3 532.1 133.0 68 11.4

Fig. 3: Comparison of the resolution of the piezoelectric and MEMS sensor types

Long-Term Stability

The long-term stability of piezoelectric sensors is well documented: these devices have been operating in the field for more than 30 years.

Piezoelectric crystals are inherently stable over time. The long-term drift parameters depend on the crystal formulation used, so an actual value is difficult to present. Quartz has the best long-term stability, but it is rarely used in condition monitoring applications because of its limited output and high cost.

PZT crystals are the most commonly used type in piezoelectric accelerometers, and are increasingly becoming the crystal of choice for most other applications.

VC MEMS accelerometers also have wide specification limits for long-term drift depending on the MEMS design structure. A bulk micromachined MEMS sensor will have the best long-term drift but will also be markedly more expensive, and typically only used in inertial applications.

For condition monitoring, MEMS vendors offer surface-micromachined VC MEMS sensors, which are much less expensive, but the end user will sacrifice measurement resolution and long-term stability. The MEMS structure of surface-micromachined designs is less stable than that of bulk-micromachined MEMS sensors.

Sensor Output Options

Depending on the installation and application, a choice of sensor output-signal options may be necessary. Most current predictive maintenance installations require an analog signal from the sensor, so the end user can decide on which parameters to monitor for a particular type of machinery.

Typically, the signal output is driven by the data acquisition device’s or programmable logic controller’s interface; an analog output of ±2V or ±5V is the most common choice. In installations requiring long cable lengths, however, loop-powered 4 to 20mA sensors are also common.

In the digital factory of tomorrow, digital output signals will become more widely required, as will smart sensors with onboard microprocessors which can make immediate maintenance decisions for the end user. Both these output-signal options are available in piezoelectric and VC MEMS sensors.

Summary of the Technology Comparison

All or some of the performance parameters discussed above will help the customer make an intelligent decision on the right technology for the condition monitoring installation. Table 1 provides an overview of the factors to consider.

In condition monitoring applications, products from TE Connectivity offer superior performance, high reliability and a long operating lifetime. Examples include:

  • 820M1, a single-axis, surface-mount piezoelectric accelerometer which has a bandwidth of >10kHz, and which offers a choice of dynamic ranges from ±25g to ±500g
  • 830M1, a triaxial, surface-mount piezoelectric accelerometer which has a bandwidth of >10kHz, and which offers a choice of dynamic ranges from ±25g to ±500g


Key Parameter Piezoelectric MEMS VC
Wide Frequency Response
Long-term Signal Stability
Dynamic Range
Operating Temperature Range
Packing Options
Ease of Installation
Sensor Output Options

Table 1: Summary of the benefits of the two vibration sensor