For most engineering applications, choosing the right testing tools will significantly impact the test results. This article will help readers select the correct accelerometer sensor. Let's begin with the classification and principles of sensors.
Basic types of accelerometers
Generally speaking, there are two types of accelerometers: AC response accelerometers and DC response accelerometers.
As an AC-response accelerometer, as its name suggests, its output is AC-coupled. Such sensors cannot be used to test static accelerations, such as gravitational acceleration and centrifugal acceleration. They are only suitable for measuring dynamic events. DC-response accelerometers, on the other hand, have a DC-coupled output and can respond to acceleration signals as low as 0 Hz. Therefore, DC-response accelerometers are suitable for simultaneously testing static and dynamic accelerations. It's not that DC-response accelerometers are only chosen when testing static acceleration.
Acceleration, velocity, displacement
Many vibration studies require information on acceleration, velocity, and displacement, which are crucial for engineers designing and validating structures. Generally, acceleration provides a good reference, while velocity and displacement are variables needed for calculations. To calculate velocity and displacement from acceleration, the acceleration signal output from the sensor is integrated digitally and analogically, first and second times, respectively. This can cause problems with AC-coupled sensors. To demonstrate this problem, consider using an AC sensor to measure a wide-pulse half-sine wave signal. Due to the inherent limitations of the AC RC time constant, the sensor output does not match the input pulse well. For the same reason, at the end of the pulse, the sensor output will exhibit a negative zero-point offset. The figure below shows the relationship between the sensor output (red curve) and the wide-pulse half-sine acceleration input (blue curve).
This seemingly small difference in amplitude will result in a significant error after integration. DC-response accelerometers do not have this problem because their output accurately follows slowly changing inputs. In practical everyday applications, the input signal may not be a simple half-sine pulse, but this problem will always exist when testing any slowly changing signal with an AC-coupled sensor.
Now let's take a look at various commonly used accelerometer technologies.
AC response accelerometer
The most commonly used AC response accelerometers employ piezoelectric elements as their sensing units. When acceleration is input, the sensing mass in the sensor "moves," causing the piezoelectric element to generate a charge signal proportional to the input acceleration. From an electrical perspective, the piezoelectric element acts like an active capacitor, with an internal resistance on the order of 10 x 9 ohms. The internal resistance and capacitance determine the RC time constant, which in turn determines the sensor's high-frequency throughput characteristics. For this reason, piezoelectric accelerometers cannot be used to measure static events. Piezoelectric elements can be derived from nature or man-made. They have varying signal conversion efficiencies and linearities. Two types of piezoelectric accelerometers are available on the market: charge output and voltage output.
Charge output type accelerometer
The main piezoelectric accelerometers utilize zirconium titanate ceramics, offering a wide operating temperature range, a wide dynamic range, and a wide frequency range (usable frequencies >10kHz). Charge-output accelerometers encapsulate the piezoelectric ceramic in a hermetically sealed metal housing. Due to their ability to withstand harsh environments, they exhibit excellent durability. Because of their high impedance, these sensors require the use of a charge amplifier and a low-noise shielded cable, preferably a coaxial cable. Low-noise cables refer to cables with low triboelectric noise², a noise generated by motion originating from the cable itself. Many sensor manufacturers also supply such low-noise cables. Connecting the charge amplifier to the charge-output accelerometer eliminates the effects of parallel cable and sensor capacitance. With an advanced charge amplifier, charge-output accelerometers easily achieve a wide dynamic response (>120dB). Due to the wide operating temperature range of piezoelectric ceramics, some sensors can be used in environments ranging from -200°C to +400°C, or even wider. They are particularly suitable for vibration testing at extreme temperatures, such as monitoring turbine engines.
Voltage output type accelerometer
Another type of piezoelectric accelerometer outputs a voltage signal instead of a charge signal. This type of sensor contains an internal charge amplifier. Voltage-mode sensors are available in 3-wire (signal, ground, power) and 2-wire (signal/power, ground) configurations. The 2-wire type is also known as an integrated circuit piezoelectric sensor (IEPE). IEPE is very popular due to its ease of connection using coaxial cables (2 wires, core wire, and shield wire). In this mode, the AC signal is superimposed on the DC power supply. Adding a coupling capacitor in series at the output removes the DC bias voltage of the sensor, thus obtaining only the sensor signal output. Many modern instruments provide an IEPE/ICP3 input interface, allowing direct connection to the IEPE sensor. If the IEPE power supply interface is unavailable, a signal amplifier with a constant current source is required to power the IEPE sensor. 3-wire sensors require a separate DC power supply line.
Unlike charge-output accelerometers, voltage-output accelerometers, in addition to a piezoelectric ceramic element, contain a miniature circuit. The operating temperature range of this circuit limits the overall operating temperature range of the sensor, typically not exceeding 125°C. Some designs have increased this to 175°C, but this results in a decrease in other performance aspects.
Available Dynamic Range – Due to the extremely wide dynamic range of piezoelectric ceramic elements, charge-output accelerometers offer great flexibility in range definition, as their full-scale range can be freely adjusted by the user via a remote charge amplifier. Voltage-output accelerometers, on the other hand, have a fixed full-scale range determined by their internal charge amplifier, which cannot be changed once manufactured in the factory.
Piezoelectric accelerometers can be made into very small packages, making them suitable for dynamic testing of lightweight structures.
DC response accelerometer
Two techniques are frequently used to fabricate DC response accelerometers: capacitive and piezoresistive.
Capacitive
Capacitive accelerometers (where capacitance changes with acceleration due to a sensing mass) are the most common type of accelerometer today. They are irreplaceable in certain fields, such as airbags and mobile devices. High production volumes make these sensors inexpensive. However, this low-cost sensor is limited by a lower signal-to-noise ratio and a limited dynamic range. All capacitive accelerometers have an internal clock (~500kHz), an essential part of the detection circuit, which often interferes with the output signal due to leakage. This noise frequency is much higher than the measurement signal frequency and generally does not affect the measurement results, but it is always superimposed on the test signal. Because of the built-in amplifier chip, they typically have a 3-wire (or 4-wire differential output) interface. They can operate with a DC power supply.
The operating bandwidth of capacitive accelerometers is typically limited to a few hundred Hz, partly due to their large internal structure and heavy air damping. Capacitive accelerometers are suitable for measuring low-range accelerations, generally up to 100g. Aside from these limitations, modern capacitive accelerometers, especially instrument-grade devices, exhibit excellent linearity and high stability.
Capacitive accelerometers are typically suitable for on-board testing, partly due to their low cost. They are also an ideal choice for low-frequency motion testing, where acceleration is generally low. For example, they are used in vibration testing in civil engineering.
Piezoresistive type
Piezoresistive accelerometers are another widely used type of DC-response accelerometer. Unlike capacitive accelerometers, which measure acceleration through changes in capacitance, piezoresistive accelerometers output acceleration signals through changes in the value of a strain gauge, which is part of the sensor's inertial sensing system. Many engineers are familiar with strain gauges and know how to measure their output. Most piezoresistive sensors are sensitive to temperature changes, thus requiring temperature compensation for their output signal, either internally or externally. Modern piezoresistive accelerometers incorporate a dedicated integrated circuit for on-board signal processing, including temperature compensation.
Piezoresistive accelerometers can operate at frequencies up to 5000 Hz. Many piezoresistive accelerometers employ either air damping (MEMS type) or liquid damping (bonded strain gauge type). Damping characteristics are an important factor in sensor selection. In some applications, the input mechanical vibration contains high-frequency components (or excites a high-frequency response). Damped sensors can prevent themselves from ringing (resonance), thus preserving or increasing the usable dynamic range. Because the output of a piezoresistive accelerometer is differential pure resistance information, the signal-to-noise ratio is generally excellent; its dynamic range is limited only by the quality of the subsequent DC amplifier. For high-acceleration impact testing, some piezoresistive accelerometers can measure accelerations exceeding 10,000 g.
Due to their wide frequency response, piezoresistive accelerometers are suitable for pulse and collision tests, where both frequency and acceleration are typically high. As sensors with DC response, their acceleration output provides users with velocity and displacement information without integration errors. Piezoresistive accelerometers are commonly used in automotive safety testing, weapons testing, and earthquake testing.
summary
Each accelerometer technology has its advantages and disadvantages. It is crucial to understand their differences and testing requirements before making a choice. First and foremost, for applications requiring the measurement of static acceleration or low-frequency acceleration (<1Hz), or applications that need to calculate velocity and displacement using acceleration, an accelerometer with DC response is necessary. Both DC and AC response accelerometers can measure dynamic signals. When only dynamic signals need to be measured, users can choose according to their preference. Some users dislike handling the zero-point bias of DC response accelerometers and prefer AC-coupled, single-ended piezoelectric accelerometers. Other users do not care about handling zero-point bias, prefer 3-wire or 4-wire interfaces, and like load resistance self-test (shunt) and gravity acceleration self-test (2g flip-flop) functions. They will choose a DC response accelerometer.
In summary, charge output mode piezoelectric accelerometers are the most durable design, primarily due to their simple structure and robust, reliable materials. For dynamic testing at high temperatures (>125°C), charge output mode piezoelectric accelerometers are the undisputed, and indeed the only, choice. For charge output mode piezoelectric accelerometers, a low-noise coaxial cable and a charge amplifier (or in-line charge converter) are essential.
Voltage-output piezoelectric accelerometers are the most commonly used for dynamic testing. They feature a small size, wide bandwidth, and a built-in charge amplifier, allowing direct connection to existing instruments or data acquisition units (with IEPE/ICP interfaces). Voltage-output piezoelectric accelerometers are generally used in applications below 125°C, but due to their low output impedance, low-noise coaxial cables are not required.
Capacitive accelerometers are typically designed with zero-crossing damping or overdamping, making them suitable for low-frequency testing. Their low cost and SMD packaged devices make them suitable for high-volume applications in automotive and consumer goods, where accuracy requirements are often not high. More expensive instrument-grade MEMS capacitive accelerometers offer good zero-point stability and low noise. Capacitive accelerometers generally have low output impedance, a 2–5V output swing, and require a stable DC power supply.
Piezoresistive accelerometers come in many types, with varying frequency response and dynamic ranges. As sensors with DC response, they can measure static acceleration, thereby accurately calculating velocity and displacement. The frequency response bandwidth of piezoresistive accelerometers is sufficient for most dynamic tests. Their damping can be designed with different values (ξ = 0.1–0.8), making them adaptable to various test conditions, including impact tests. Pure piezoresistive accelerometers (without signal conditioning circuitry) can be made very small and lightweight, with moderate output impedance (<5000Ω) and a full-scale output of 100–200mV. Piezoresistive accelerometers with amplification (integrated signal conditioning chip) have lower output impedance (<100Ω) and a full-scale output of 2–5V.
Notes and References
AGPiersol,TLPaez,Harris'ShockandVibrationHandbook6thEd.,p.10.9,McGraw-Hill,2010
AGPiersol,TLPaez,Harris'ShockandVibrationHandbook6thEd.,p.15.19,McGraw-Hill,2010
ICP is a registered trademark of the PCB company. There are many different companies in the market that use the same trademark for this type of sensor.
