Low-pressure silicon sensors can respond to a tenth of a pound / in2 of stress, and there are currently production and applications of such products.
In a very small range of changes, silicon sensors have been able to measure small changes in liquid level pressure or the pressure exerted by exhalation. These devices use a new technology of pressure sensing, which can replace mechanical pressure devices for many applications.
But how can you make the best choice in the application? A good way to judge is to test the differences between silicon sensors and traditional sensors, market advantages and the technical specifications of the silicon sensors you use.
What can correctly compose a low-voltage sensor system? In general, the typical pressure range in silicon sensors is between 0–0.15psIF.S., And can often rise to 0–10,000psi. Manufacturers define different low voltages based on the design of their sensors and their production process. Sensors that respond to less than 5psi generally require different die topologies and process technologies. According to the discussion of this issue, that is, the pressure less than 5psi is defined as the low pressure range. Differences in silicon technology form low-pressure sensing silicon micromechanical components manufactured by similar processes to form standard pressure range components, but there are key differences. Standard range components include resistance bridges etched on a thin diaphragm (Wheatstone method). When the pressure excited by the voltage source forms a diaphragm offset, it will change the resistance value and cause the output voltage to change. Low-pressure silicon sensors work similarly, but there are also obvious features, including a protruding diaphragm structure and about 50% of the diaphragm area used for stress concentration. The precise etching film thickness and strategic placement resistance can also greatly expand the capacity of the sensor. In addition, silicon sensors using traditional semiconductor batch processing technology will also have its cost advantages of mass production.
Low-voltage silicon sensors are commonly used in the following three major markets:
1. HVAC. Low-voltage sensors are indispensable components in heating, ventilation and air-conditioning systems. In other applications, they can monitor ventilation and air flow, determine the volume of airflow, detect load problems caused by dirty filter changes, and control the pressure of the entire system. These applications require the product to be able to detect pressure differences during pressure changes of 0.015 psi.
2. Medicine. Without small-sized sensors, many medical applications are unlikely to detect environmental changes and conditions. Therefore, the design must meet the following requirements: a relatively small sensor (1-2 mm) that can measure the fluid pressure in the human body. These devices are generally urinary catheter devices that can be inserted into the skull. Heart cavity or uterus, real-time monitoring of difficult surgical operations, this situation is absolutely a one-time low-cost product. Low-pressure products used in respirometers require sensor ranges below 0.5 psi. These operations were previously performed by mechanical pressure switches that changed over time.
3. Cars, low cost and high reliability are the basic elements that constitute the sensing of external car conditions, including injection. Gasoline steam and exhaust. Recent cars use silicon sensors to monitor tires. Line and brake pressure. The application range of silicon semiconductor sensors can also be extended to manufacturing packaging technology to protect the use of automotive components under harsh environmental conditions.
Application requirements and sensor types Because each application has its own characteristics, you must fully consider all aspects of the entire system during the selection process. Be sure to determine the pressure source input, desired output, and various related operating conditions. The installation position and direction, the distance from the pressure source. Wire stress. Interface and pressure connection. Determine the impact performance and accuracy range of the package. For example, in medical applications, membrane elements that sense pressure must be packaged without packaging. In the measurement process affected by temperature fluctuations, a more advanced sensor must be used to compensate and calibrate the offset, so as to eliminate the temperature coefficient generated by the material. The surrounding environment, such as humidity and pollution, can also have an impact on medical protection applications.
The tradeoff in accuracy is to control certain characteristics through the microprocessor and eliminate undesirable results. The compatibility of the system components (such as power supply, amplifier, A / D converter, control circuit) and the analog output signal of the sensor must reach the desired solution and the accuracy of the entire design.
Each type of pressure sensor has material properties that can form measurable output data based on changes in applied pressure. When you are considering the technical data of the universal detector, you must keep in mind your technical requirements. Silicon micromechanical materials. These products consist of a micromechanical silicon diaphragm injected with electrical resistance, which can produce piezoelectric changes under pressure. Applications include medical ventilation equipment, body pressure detection, automotive ventilation, vacuum cleaning, and steam pressure.
Electronic materials: These products include stress gauges and variable capacitance sensors (not made of micro-mechanical silicon). The stress gauge uses deflection thick / thin film. Coupling effect of metal foil or bonded metal foil. The variable capacitance element combines the change in capacitance and converts it into a measurable signal. Stress gauges are used in contact and mechanical pressure applications.
Vacuum gauge: The vacuum sensor uses the heating wire as a leg of the balance bridge, and detects the pressure by measuring the change in its resistance. They are mainly used to detect the vacuum value reached by the vacuum pump chamber, such as electron microscopes and other process equipment.
Piezoelectric materials: These sensors are made of polymers.
Crystals: made of ceramics and thin films, which are polarized when they are mechanically applied. This type of product is very suitable for dynamic pressure sources but not for continuous static conditions like atmospheric pressure. They can work well in harsh environments.
Variable magnetoresistive material: The variable magnetoresistive sensor combines an inductive half-bridge configuration between the deflection sensing diaphragms of two metal films. The result of the change in induction / impedance can produce a changing AC signal. Typical applications include low-voltage HVAC. Understanding the usage specifications Once you have determined the design requirements, you can determine the most important selection criteria according to priority and determine which technology is most suitable for your application. How do you determine which standard is the most important? System specifications usually require that certain parameters must be guaranteed, mainly the pressure operating range, and its general positioning has the highest priority. For low-voltage micromechanical devices used in volume, you can consider a wider response range on the premise of price.
With the confirmation of the pressure range and sensitivity, the standard order generally ranked by level is physical size, cost, accuracy, temperature performance, reliability, long-term stability, and media compatibility. Practical proof accuracy and temperature performance are often the most complex processing factors. Every manufacturer will describe their accuracy specifications well, but these parameters must be converted into comparable units to increase their comparability (don't be fooled by these misleading specifications!). To help you understand the performance specifications, the following lists the most general interpretations of the definition of accuracy and the most noteworthy changes to each content. The first group assumes a reference temperature of 25 ° C. Zero point / offset. The value of the output voltage excited in the textbook is usually considered to be the zero voltage applied at 25 ° C. Generally expressed as 0 ± mV. Due to the relative change, more offset will be generated when a higher voltage or current is applied.
Hysteresis of zero-point pressure: When the sensor is used for the cyclic measurement of one or more full-scale pressures, the problem of repeated measurement of zero point will occur. The unit is expressed as a percentage of the full-scale output. When calculating the hysteresis performance of the zero-point pressure, different sensor manufacturers will use different pressure cycle times and different full-scale ranges.
Hysteresis of pressure range: When the sensor is used for one or more pressure cycles, the repeated measurement value of the range is output. This value is usually expressed as a percentage of the full-scale output in the worst case.
Temperature hysteresis performance of the range: when the sensor is cycled at the same temperature, the repeatability of the range is judged. In other words, the sensor is subject to minimum and maximum temperature limitations. The difference in the range reading after recycling is the temperature hysteresis of the range. This parameter is expressed as a percentage of the full-scale output. Therefore, the full-scale pressure reading determines the worst-case change in range when cycling at various temperatures.
Sensitivity: The ratio of the output signal change to the pressure change. The actual situation of this unit depends on the manufacturer, but the value is generally expressed in mV / V or I / psi. Sensitivity is a key performance indicator for determining system solutions.
Long-term stability: a measure of range and zero-point performance changes over a period of time. Generally expressed in mV. Figure 1. In order to calculate linearity, the best straight line (BFSL) error method can provide an average error measurement. There is a certain error between the BFSL straight line and the actual detected pressure (see Figure 1). The final basic linearity is determined by the worst case measurement of pressure from zero to the midpoint of the full scale.
Non-linearity of the range: When the pressure increases, the proportional output produced by the silicon micromechanical sensor is generally low. So for high pressure the pressure transducer will show a lower output, and the actual linearity can only be speculatively indicated. There are currently two basic methods for calculating linearity. The best straight line (BFSL) error method can provide average error measurement. There is a certain error between the BFSL straight line and the actual detection pressure (see Figure 1). The final basic linearity is determined by the worst case measurement of pressure from zero to the midpoint of the full scale. The nonlinearity of the terminal is generally 2xBFSL nonlinearity. The second category of specifications describes temperature-related accuracy. When the temperature changes, all sensors have non-repetitive zero drift. Typical changes in bridge resistance at room temperature are approximately + 3000ppm / ° C (sometimes expressed in mV / V / ° C). However, below room temperature (usually –20 ° C), the impedance will exhibit a zero slope and convert. Under this condition, it is assumed that the linear relationship between resistance and temperature is invalid, so the correction coefficient term must be incorporated into the sensor module. The best thing that can be done is to minimize the error. In practice, it is related to the tolerance standards used by individual manufacturers. In order to optimize their performance standards, manufacturers will propose different performance indicators.
Temperature coefficient of zero point (three ways): This value is difficult to define accurately, and the precise regulations are affected by changes in manufacturers. Some manufacturers use simple linearity, that is, a straight line connecting two end points at a reference temperature (25 ° C), and the special error should be more than the ideal range of the butterfly valve. The error should be lower than the given percentage of full scale. This is the minimum measurement standard for this performance, but it is not widely adopted. The third method is to calculate BFSL with three data points. Figure 2. There are three techniques that can be used to determine the temperature coefficient of the zero point. A simple straight-line method with two endpoint temperatures as reference points can provide the most information. The full-scale error method provides the least information on device performance, but it is widely used. Among the three technologies, the first one provides the most data information, and the third one (BFSL error) provides users with limited accuracy. The second method provides the least device performance information.
Temperature lag at zero point: When the sensor is used for one or more temperature cycle cycles, the reading reflects the repeatability of the zero point measurement. After a certain cycle, the difference in the zero reading reflects the hysteresis of the temperature, which is generally defined as the percentage of the full-scale output. Therefore, the full-scale pressure reading determines the worst-case temperature deviation over the entire temperature cycle.
Temperature compensation: This is a method to eliminate the influence of temperature changes, which uses a complex mathematical model to determine the correction value of the thick film resistance. Operating systems running at various changing temperatures require both compensation through electronic systems and prior adjustment by sensor manufacturers (Shanghai Automation Instrumentation Co., Ltd.). The combination of various error elements can be used to describe the effect of sensor errors on system performance. The following two definitions are widely used to describe the accuracy of the entire sensor: worst-case error. The sum of various related single errors, here: error (worst case) = E1 + E2 + E3 ... En the most likely error. Defined as the square root of the sum of squares of individual errors, where: error (most likely) = (E12 + E22 + E32 + ... En2) RMS is not affected by the direction of the error. Product selection is now that you can choose. You have determined the selection criteria. You have a list of design constraints that you can use to make decisions.