Pressure sensor with high precision, reasonable request error, error compensation of pressure sensor is the key to its application. Pressure sensors are mainly offset error, sensitivity error, linear error and hysteresis error, this article will introduce the four errors of the mechanism and the influence of the result of the test, at the same time will be introduced to improve the measurement accuracy of pressure calibration method and application example.

Sensors on the market at present species diversity, which makes the design engineer can choose the required by the system pressure sensor. These sensors include the basic converter already, also include more complex with high integration sensor circuit on a chip. Because of these differences, the design engineer must be enough to compensate the pressure sensor of measuring error as far as possible, this is to ensure that the sensor could satisfy the requirement of design and application of important steps. In some cases, the compensation can improve the overall performance of sensor in the application.

Offset, scope, calibration and temperature compensation can be realized through the thin film resistor network, this kind of thin film resistor network in the process of encapsulation with laser correction.

The sensor is usually combined with micro controller, and micro controller embedded software itself sensor mathematical model is established. Microcontroller reads the output voltage, through the transformation of adc, the model can converts voltage stress measurements.

The mathematical model of sensor is the most simple is the transfer function. The model can be in the entire calibration process is optimized, and the maturity of the model will increase with the increase of the asm.

From the perspective of metrology and measurement error is quite strict definition: it is characterized by measuring the pressure and the difference between the actual pressure. Often unable to get the actual pressure directly, but can through the adoption of the standard to estimate the appropriate pressure, metrologists usually adopt the accuracy of at least 10 times higher than the device under test standard as measuring instrument.

Because of without calibration system can only use the typical sensitivity and offset value converts the output voltage to pressure, measured pressure will produce the error as shown in figure 1.

The initial error without calibration consists of the following parts:

A. the offset error. Due to the constant vertical migration in the whole pressure range, so the converter diffusion and amendments to the laser control changes will produce offset error.

B. error sensitivity, produce the error size is proportional to the pressure. If the device sensitivity is higher than the typical values, error sensitivity will be pressure increasing function. If the sensitivity is lower than the typical values, error sensitivity will be pressure decreasing function. The error reason lies in the diffusion process of change.

C. linear error. This is a small effect on initial error factors, the causes of the error is physical nonlinear of silicon, but for sensor band amplifier, should also include the nonlinear amplifier. Linear error curve can be concave curve, can also be a convex curve.

D. lag error: in most cases, completely negligible hysteresis error, because silicon has high mechanical stiffness. Generally only in case of pressure change a lot considering lag error.

Calibration can eliminate or greatly reduce these errors, and compensation technology usually requirements determine the parameters of system transfer function is actual, rather than simply using typical values. Potentiometer, adjustable resistance, and other hardware can be used in the process of compensation, while the software can work more flexibly to realize the error compensation.

Zero point calibration method can be resolved through the transfer function of the drift to compensate for the offset error of this kind of automatic zero calibration method is called.

Offset calibration is usually under the zero pressure, especially in the differential sensor, because under the condition of nominal differential pressure usually is 0. For pure sensor, the offset calibration is more difficult, because it is either need a pressure load system, used to measure the pressure in the environment under the condition of atmospheric pressure calibration value, or need to get expected pressure pressure controller.

Changes of zero pressure sensor calibration is very accurate, because the set pressure strictly to 0. On the other hand, the pressure is not 0 means that the calibration accuracy depends on the controller or the performance of the measurement system.

Select set pressure

Selection of set pressure is very important, because it determines the best precision pressure range. In fact, after calibrating the offset error between actual asm minimum and has kept a smaller value. Asm, therefore, must choose according to the target pressure range, and pressure range can not consistent with the scope of work.

In order to convert the output voltage to pressure value, due to the actual sensitivity are often unknown, so the mathematical model of the sensitivity of typical for a single point calibration is normally used.

Red curve (PCAL = 0) offset calibration error curve, can be found that the error curve relative to the said before the calibration error of the black curve has a vertical migration.

This calibration method is compared with a calibration method is more strict, make the cost is also higher. However, compared with point calibration method, this method can significantly improve the precision of the system, because the method not only calibration offset, and calibrate the sensitivity of the sensor. Because this is in error sensitivity can be used in the calculation of the actual value, rather than the typical values.

Green said curve accuracy improved. Here, the calibration is from 0 to 500 million bar (full scale) conditions. Due to error close to zero in the asm, so in order to get the smallest within the desired pressure measurement error, these correctly set point is particularly important. In some application requirements within the scope of the pressure maintaining high accuracy. In these applications, the multi-point calibration method can be used to obtain the ideal results. In multi-point calibration method, not only consider the sensitivity and offset error, linear error, considered the most purple curve shown below. Here with a mathematical model to each calibration interval (in) between two fixed point of two level calibration are exactly the same.

Three calibration

As mentioned above, with a consistent form of linear error, and error curve accords with quadratic equation of the curve, with predictable size and shape. This is especially true for not using sensors amplifier, because in essence is based on the mechanical reasons of the sensor (which is caused by silicon thin film pressure).

Features a description of the linear error can be calculated the linear error of typical examples, determine the polynomial function (a * 2 + bx + c) the parameters of. After they identified a, b and c model for the same type of sensor is effective. This method can without a third point effectively under the condition of linear error compensation.

The error compensation method by two-point calibration the low-cost sensors can be improved for high performance devices (error less than 0.05% of full scale).