company news about How to Minimize the Preheating Drift of Pressure Sensors

The temperature drift phenomenon of pressure sensors can cause reading fluctuations until the system reaches the working temperature. This situation usually has little impact. However, in medical equipment such as hospital ventilators, lung function testing devices, and neonatal monitors that require continuous high precision, this temperature drift is unacceptable. Checking the basic piezoresistive pressure sensor helps to understand the impact of preheating drift.

This sensor consists of a main body (i.e., the "chip") and a thin silicon diaphragm with four piezoresistive torsion structures on its surface. The piezoresistive elements change their resistance values with stress changes, and they are usually arranged in a bridge structure and precisely installed on the diaphragm surface to enhance the response to diaphragm deformation. This design can effectively improve the response sensitivity when the pressure difference on both sides of the diaphragm changes.

There are two main sources of preheating drift in basic pressure sensors. One is the preheating offset of the sensing element. When the system reaches the operating temperature, the tube, surface temperature, and the resulting hot spots (surface contribution) cause an imbalance in the resistance bridge on the chip and diaphragm surface. The temperature rise of the resistance sensing element is proportional to the dissipated power and thus proportional to the square of the sensor excitation voltage (ΔTαV2).

Therefore, when the excitation voltage is halved, the temperature rise of the sensing element will be reduced by a quarter, thereby reducing the preheating surface condition by four times. Since the sensor signal level will also be reduced by a quarter in both cases (with the reduced supply voltage), the overall effect is to reduce the preheating error due to the surface contribution by half. However, reducing the sensor power supply will have an adverse effect on the system electronic noise level.

Another preferred solution is to adjust the sensor supply voltage according to the system bandwidth requirements. Specifically, the sensor is powered only when needed. This design adjusts the sensor's power-on time to the average duty cycle (i.e., the working cycle), effectively suppressing the thermal startup drift phenomenon. Although the implementation mechanism of this method is slightly more complex, it offers excellent performance without affecting the system noise level.

Here, the period p between power pulses of the application refers to the time when the power is off plus the time when the power is on. This is the time required for all signals to stabilize and for the sensor to take readings.

For example, consider a device that needs to take readings every 500 ms, with a stabilization time of 4 ms and a signal acquisition time of 1 ms. Compared to a non-modulated system, the average power of the sensor is only 1% of the applied power ((1 ms + 4 ms) / 500 ms). Of course, this time period depends on the sampling requirements of the application. Due to the influence of surface charges, the constancy of p and time t is very important. However, considering the benefits of regulating the sensor power supply, this is a secondary limitation.

Temperature compensation technology

Another root cause of preheating drift is actually more related to the sensing characteristics, which is closely related to the temperature compensation technology of the system. Such systems are usually equipped with external temperature sensors to calibrate the pressure sensor to eliminate the influence of temperature. In a dual-sensor system, a temperature gradient will be generated between the external device and the diaphragm surface. The time required for this temperature gradient to stabilize will be perceived as the preheating drift phenomenon.

By using the sensor resistance (the bridge resistance that varies with temperature) as the temperature sensing element, this influence can be minimized. Here, the pressure sensor bridge replaces the thermistor (a resistor used to measure temperature changes) typically used in the circuit, effectively forming a Wheatstone bridge. The sensor bridge has a relatively high positive temperature coefficient (TCR), so an increase in temperature will gradually cause the signal output voltage (Vt) of the temperature monitoring part of the circuit to show a negative change. The change of Vt relative to the reference voltage (Vref) is actually an effective measurement of the sensor temperature itself. The system electronics use this measurement as the calibration temperature reference for the pressure sensor. Since there is no need to rely on an external temperature sensor, there is no temperature gradient in the system, thus eliminating the so-called preheating drift phenomenon. Even more pleasingly, by combining power regulation and temperature compensation techniques, the influence of preheating drift can be almost completely eliminated.