Log In

Log In

Need an Account? Start Here


Introduction to ICP® Accelerometers

An ICP® accelerometer is a sensor that generates an electrical output proportional to applied acceleration. ICP® accelerometers are designed to measure vibration and shock for a wide variety of applications. They are simple to use and accurate over a wide frequency range which makes them the recommended choice for many testing situations.

ICP® is a PCB® registered trademark that stands for "Integrated Circuit Piezoelectric" and identifies sensors that incorporate built-in microelectronics. The electronics convert a high-impedance charge signal generated by a piezoelectric sensing element into a usable low-impedance voltage signal that can be readily transmitted, over ordinary two-wire or coaxial cables to any data acquisition system or readout device.

A variety of mechanical designs are used to perform the transduction required for ICP® accelerometers. The designs consist of sensing crystals that are attached to a seismic mass. A preload ring or stud applies a force to the sensing element assembly to make a rigid structure and insure linear behavior. Under acceleration, the seismic mass causes stress on the sensing crystals which results in a proportional electrical output. The output is collected on electrodes and transmitted by wires connected to the microelectronic circuitry in ICP® accelerometers.

All PCB® ICP® accelerometers require power from a constant current DC voltage source. PCB® offers different types of ICP® signal conditioners that provide 2 to 20 mA of current at a DC voltage level of +18 to +30 volts. Do not attempt to power ICP® sensors with commercially available power supplies. The unregulated current will damage and destroy internal electronics.

Some data acquisition systems include ICP® power. In this case a separate signal conditioner is not required.

The DC bias level (turn-on voltage) of the accelerometer will typically fall in the +8 to +12 volt range. The measured DC bias voltage is checked during calibration and listed on the calibration sheet. More information on powering ICP® accelerometers can be found here.

Most ICP® accelerometers have a full scale output voltage of ±5 volts. The output is directly related to the measurement range and sensitivity of the accelerometer. Examples 1 and 2 illustrate this:


352C33 has a sensitivity of 100 mV/g and a range of ±50 g’s peak.
50 g’s X 100 mV/g = 5000 mV = 5 volts


353B04 has a sensitivity of 10 mV/g and a range of ±500 g’s peak.
500 g’s X 10 mV/g = 5000 mV = 5 volts

The sensitivity of the accelerometer remains linear from small scale inputs up to full scale. This is illustrated in Figure 4.

Figure 4: Amplitude linearity of model 353B03 (up to 500 g full scale range)

ICP® accelerometers are AC coupled devices and will not measure or respond to uniform acceleration (also known as static or DC acceleration). Capacitance and resistance values internal to the accelerometer set the discharge time constant (DTC) and low frequency response. If uniform acceleration is applied, the output signal will decay according to the DTC. The output signal will completely discharge after five discharge time constants. More information on the low frequency response of ICP® accelerometers can be found here.

Every ICP® accelerometer has a natural frequency that will restrict the measurement frequency range to some upper limit. The natural frequency (resonance) is a mechanical characteristic imposed on the accelerometer by its physical design characteristics. Sensitivity rises rapidly as the natural frequency is approached which can often result in an overload of signal output. An example of resonance is show in Figure 5.

Figure 5: ICP® accelerometer at resonance

It’s important to note that mounting plays a role in obtaining accurate high frequency measurements. Consult installation drawings and product manuals for proper mounting techniques of specific models. Additional information on accelerometer high frequency response and mounting can be found here. More information on the high frequency response of ICP® accelerometers can be found here.

PCB® includes a calibration certificate with every ICP® accelerometer. This certificate documents the characteristics of each accelerometer and provides exact values for several key specifications. A sample calibration certificate is shown in Figure 6.

Figure 6: ICP® accelerometer calibration certificate

Back-to back calibration is performed with the test accelerometer mounted onto a reference accelerometer. This technique provides a quick and easy method for determining the sensitivity of an accelerometer over a wide frequency range.
The reference accelerometer is an extremely accurate device with specifications traceable to a recognized standards laboratory. It is possible to vibrate both accelerometers and compare output data by securely mounting the test accelerometer to the reference standard accelerometer.

The ratio of the output voltages is also the ratio of the accelerometers’ sensitivities because the acceleration applied to them is the same. The sensitivity of the reference accelerometer is known so the sensitivity of the test accelerometer can be calculated.
Recalibration services are offered for PCB® manufactured accelerometers as well as those produced by other manufacturers. Our internal metrology lab is certified to ISO 9001 and accredited by A2LA. The equipment used during calibration is directly traceable to NIST (National Institute of Standards and Technology).

The sinusoidal output of piezoelectric accelerometers includes positive and negative components resulting from bidirectional motion imparted into the sensor axis. For clear understanding, conventions must be set for the motion and for the electrical charge. Axial motion imparted from the base of the sensor, directed into the sensor is considered to be in the positive direction. Motion directed away from the sensor base is considered the negative direction. This motion is transferred to the sensor’s seismic mass which loads the piezoelectric sensing elements providing a high impedance electrical signal known as charge output.

Charge output accelerometers are typically wired with positive motion resulting in a negative output as they are typically converted to low impedance signals via an inverting op-amp. This convention makes charge sensors negative polarity unless specified otherwise.

ICP® accelerometers have a built in externally powered microelectronics to convert the high-impedance charge signal into a low-impedance voltage signal directly usable by most modern instrumentation without requiring external charge amplifiers. For this reason, ICP® sensors are typically positive polarity.

Note about charge amplifiers: different manufacturers have used both inverting and non-inverting op-amps within their amplifier electronics and it is important to understand which type of amplifier you are using with which type of sensor. PCB inline charge amplifiers are of a negative polarity, designed for use with standard charge sensors resulting in a positive signal to readout devices.

Charge sensors, in an inverting design, are labeled with a P-prefix to indicate the special positive polarity. Be aware that some custom ‘M’ models include the inverting functionality but do not have the P-prefix. Inverted ICP® sensors are extremely rare. Refer to the accelerometer datasheet to confirm if the specific model has P options available and confirm the P-prefix in the specific sensor’s model number.

Accelerometer output can be confirmed through use of an oscilloscope (possibly with use of a volt meter for sensors with relatively long time constants), hold the sensor in the palm of your hand, and tap gently on the accelerometer’s base (positive input). Similarly, polarity of ICP® sensors can be confirmed using a PCB battery operated signal conditioner where the needle will move to the right indicating positive polarity and move to the left indicating negative polarity.

All of this holds true for triaxial sensors that receive symbols etched into each sensor detailing which direction is positive for each axis (x, y, and z). There are specific applications where knowing the specific polarity has a greater impact – predominantly tests with multiple sensors or inputs at multiple locations such as modal analysis. Incorrect polarity, direction or inconsistent sign conventions will result in inaccurate results.