Newnes Sensor Technology Handbook 2005 Yyepg Lotb Part 12 pps

Chapter 16 430 2. With semi-automatic command, the user initiates the action. After it is trig- gered, the system sequences through multiple functions controlled by timers or shift registers. This could include solenoid actuation to switch from measurement to reference pressure, followed by the auto-reference function, then a return to the measurement mode. Figure 16.1.17 illustrates a basic semi-automatic circuit. 3. Using automatic command, the system steps through multiple functions similar to the semi-automatic command. However, on returning to the measurement mode, additional timing circuitry triggers and, after a set measurement time, the sequence is restarted. Depending upon the degree of complexity desired, a small microprocessor-based system and its related software could consolidate the auto-reference circuitry, timing and control logic all into one unit. Auto-referencing requires an established system reference point. Batch processing and continuous processing are the two main categories of measurement cycle. In a batch process, a reference condition exists at some time, usually at system power-up. For example, a toilet tank has a high water level prior to flushing, corresponding to some reference pressure. When the flushing cycle is complete, the tank is filled to the previous level. The obvious point for auto-referencing is just prior to flushing when the water is at a known level. In a continuous process, there is no easily accessed reference condition. For example, the volume of fluid in a water tower is being monitored. This is a function of the depth of the water and can be sensed with a pressure sensor. Unlike the toilet, without actually taking a pressure measurement, there is no point in time at which the depth will be known. The sensor/auto-reference/enable system can be used for a simple case when a known reference exists periodically. Also, a reference condition actuator such as a solenoid valve can be used. It can switch the sensor input from the measured pressure to some other reference pressure. The solenoid can be activated by the user, by some condition such as power-up, or by a timer activated circuit. (See Figure 16.1.18.) The valve must be activated long enough for the pressure to have a chance to sta- bilize so a valid reading may be taken. For instance, consider the water tower. A gauge pressure sensor near the bottom senses the water depth. A vent tube to Figure 16.1.18: Auto-reference with reference condition actuators. Pressure Sensors 431 Figure 16.1.19: Timer-actuated circuit, single-port sensor. Figure 16.1.20: Timer-activated circuit, dual-port sensor the surface serves as a pressure reference. A three-way solenoid valve is the actuator, connecting the water and the vent to the sensor input port. A timer circuit is the enabler. (See Figure 16.1.19.) Next, suppose the water exits through a single pipe of constant diameter. The velocity can be measured with a differen- tial pressure sensor. A two-way solenoid connected between the two inlet ports serves as the reference actuator as shown in Figure 16.1.20. Latest and Future Developments Many of the latest developments associated with pressure sensing involve sensors that provide more than pressure measurements. As more and more control systems become CAN- based networks and the systems themselves get “smarter,” sensor manufacturers are seizing the opportunity to provide more “information” as opposed to simple output signals. CAN technology has led to smart sensors that offer a variety of self- and process-related diagnostic functions. They are showing up in “intelligent” appliances, on plant floors, and on-board aircraft and vehicles. Another trend involves the bundling of different types of sensing elements within the confines of a single chip. Honeywell used this approach when it used a mass airflow sensor to create a microsensor that simultaneously measures ambient temperature, pressure, thermal conductivity and the specific heat of a fluid. Such technology can be applied in chemical plants, chemical storage facilities and on automobiles where, for instance, it enables the engine to adjust itself to changes in fuel properties each time the fuel tank is refilled, resulting in improved gas mileage and cleaner exhaust. This same approach is being used to develop a high-sensitivity, high-temperature Silicon-On-Insulator (SOI) piezoresistive technology that provides pressure sensing, temperature sensing, and feedback and bias resistor networks all integrated on a single miniature 90-mil-square chip. This technology was designed for applications with both high temperature and high pressure. Its high gage factor is very important in applications where a large “signal-to-noise” ratio is essential to achieving high ac- Chapter 16 432 curacy performance, particularly over wide temperature ranges such as that required by the turbine engine and down hole oil industries. Sample test results show that the sensor design has successfully eliminated and/or minimized mechanical and thermal hysteresis error sources to levels that challenge the measurement capability of present state-of-the-art instrumentation. References and Resources 1. “Piezoresistive Technology and Pressure Measurement Types,” Honeywell, Inc. http://content.honeywell. com/sensing/prodinfo/pressure/technical/c15_101. pdf 2. “Pressure Sensors Conversion Factors and Chart,” Honeywell, Inc. http://content.honeywell.com/sensing/prodinfo/pressure/technical/c15_125.pdf 3. “Pressure Sensors Plumbing and Mounting Considerations,” Honeywell, Inc. http://content.honeywell.com/sensing/prodinfo/pressure/technical/c15_121.pdf 4. “Pressure or Force Sensor Switch Circuits,” Honeywell, Inc. http://content.honeywell.com/sensing/prodinfo/force/technical/c15_119.pdf 5. “Protecting Pressure Sensor Diaphragm From Rupture Due To Water Hammer,” Honeywell, Inc.http://content.honeywell.com/sensing/prodinfo/pressure/technical/c15_120.pdf Pressure Sensors 433 16.2 Piezoelectric Pressure Sensors Roland Sommer and Paul Engeler, Kistler Instrumente AG The brothers Pierre and Jacques Curie discovered the piezoelectric effect in 1880. They found that some crystalline materials were generating an electrical polariza- tion when subjected to a mechanical load along some crystal directions. Among the materials they investigated were quartz and tourmaline, two crystals which are today still often used in piezoelectric sensors. The first piezoelectric pressure sensor was reported around 1920, but commercial sensors were not available until the 1950s, when electrometer tubes of sufficient quality became available. Today, piezoelectric pressure sensors are widely used in laboratories and in production. The main applications are found in combustion engines, injection molding and ballistics, but they can be used in any field requiring accurate measurements or monitoring of pressure varia- tions. The main advantages of piezoelectric sensors are: ■ wide measuring range (span to threshold ratio up to 10 8 ) ■ high rigidity (high natural frequency) ■ high linearity between output signal and applied load ■ high reproducibility and stability of the properties (when single crystals are used) ■ wide operating temperature range ■ insensitive to electric and magnetic fields It is often stated that piezoelectric transducers based on the direct piezoelectric effect can only be used for dynamic measurements. This is partly true, as they react only to a change in the load and hence cannot perform true static measurements. However, a good sensor with a sensing element made of single crystal material, in conjunction with adequate electronics, can be used for accurate measurements down to 0.1 mHz. In other words, quasistatic measurements lasting up to a few hours are possible. This chapter will give an insight about the design, properties and applications of piezoelectric pressure sensors based on the direct piezoelectric effect (charge genera- tion under mechanical load). These sensors are called active sensors, as they do not need any external power supply. They have a charge output which requires an external charge to voltage converter. Essentially, there are two types of converters, the electrometer and the charge amplifier. The charge amplifier was invented by W.P. Kistler in 1950 and gradually replaced electrometers during the 1960’s. The introduction of MOSFET or JFET circuitry and the development of high insulating materials such as Teflon™ and Kapton™ greatly improved performance and propelled the field of piezoelectric measurements into all areas of modern technology. Chapter 16 434 Technology Fundamentals Piezoelectricity Piezoelectricity is basically defined as a linear electromechanical interaction in a material having no center of symmetry. One distinguishes the direct and the converse piezoelectric effect. In the first case, a mechanical load or deformation of the crystal induces a proportional charge or electrical potential. In the second case, an electric field applied to the crystal induces a mechanical deformation or a load proportional to the field. In this paper, we will focus on pressure sensors based on the direct piezoelectric effect only, which can be described in the following way: D = ε·E + d·X where D is the induced electric displacement, E the applied electrical field and X the mechanical stress applied to the material. The dielectric constant ε and the piezoelectric coefficient d are describing the materials properties. D and E are vectors, ε is a 2 nd rank tensor, d a 3 rd rank tensor and X a 4 th rank tensor. This means that the piezoelectric properties are anisotropic, the active coefficients for d and ε being determined by the crystal symmetry. The orientation of the crystalline measuring element is therefore critical and determines its properties. Longitudinal and transverse cuts, volume effect In a longitudinal cut, the surface A Q on which the charge Q is induced is the same as the surface A F on which the load (force) F is applied. The piezoelectric sensitivity depends solely on the longitudinal piezoelectric coefficient d L . Conversely, in a transverse cut, the induced charge and applied load do not share the same surface. The sensitivity depends both on the transverse piezoelectric coefficient d T and the surface ratio A Q /A F (Figure 16.2.1). longitudinal cut: Q = d L · F transverse cut: Q = d T · F · A Q /A F = d T · F · l / t where l and t are the length and the thickness of the slab. The advantage of the transverse cut is that the sensitivity can be increased by the geometrical factor l/t, provided the longitudinal and transverse piezoelectric coefficients are the same. This is always the case for quartz and all crystals belonging to the same crystal symmetry as quartz. For tourmaline, lithium niobate, lithium tantalate and piezoceramics, the transverse coefficient is only 1/10 th to 1/3 rd of the longitudinal coefficient; hence, transverse cuts are seldom used. Pressure Sensors 435 Figure 16.2.1: Longitudinal cut (left): induced charge and applied force share the same surface. Transverse cut (middle): charge is not induced on the same surface on which the force is applied. Volume effect (right): a hydrostatic pressure is applied to the sample, the charge being generated on two opposite surfaces. Both the longitudinal and the transverse effects are basically uniaxial. The pressure to be measured is converted into an uniaxial force by a diaphragm. The volume effect differs from the uniaxial effect in that the force is applied on all surfaces of the sensing element. The sensitivity is the sum of the longitudinal and the transverse contributions. Volume effect: Q = (d L · p + 2 · d T · p)·A Q = d h · p · A Q where d h is the hydrostatic piezoelectric coefficient and p the applied pressure. Note that the hydrostatic piezoelectric coefficient is always zero for quartz and all crystals having the same symmetry as quartz. The volume effect is mainly used in shock wave sensors or in hydrophone applications, where the direction of propagation of the pressure wave is not known. For the longitudinal, transverse and volume effects, the load is applied perpendicular to the piezoelectrically active surface. Shear cuts, where the load is applied parallel to the surface, could in principle be used, but would not significantly improve the properties. Piezoelectric materials The sensing element is the heart of the transducer and should hence be selected care- fully. Electronic or software compensation for “bad” crystal properties are hardly possible. High sensitivity, high insulation resistance, high mechanical strength, high rigidity, low temperature dependence of the properties over a wide temperature range, Chapter 16 436 low anisotropy, linear relationship between charge and mechanical stress, no aging, no pyroelectricity (insensitive to temperature changes), good machinability, low manufacturing costs—these are a few requirements for a good piezoelectric material. Of course, the ideal material fulfilling all the criteria mentioned above does not exist. Quartz has been used extensively in the past years and is still the material of choice for most pressure sensors. It is very stable, has a very high mechanical strength, can be used up to 400°C, has outstanding electrical insulation properties, has minimal sensitivity deviation up to 350°C (with special crystal cuts), is not pyroelectric, and is available at low cost. The only disadvantage of quartz is its relatively low sensitivity and its tendency to twin under very high loads. Tourmaline has a lower sensitivity than quartz, but its temperature range extends to at least 600°C. However, it is pyroelectric and it is only available as natural crystals. Lithium niobate and lithium tantalate have a higher sensitivity, but they are strongly pyroelectric and their insulation resistance is quite low, limiting their practical use to purely dynamic applications. Crystals of the CGG group (typical crystals are Ca 3 Ga 2 Ge 4 O 14 and langasite, for instance) have been intensively studied over the last 20 years. Belonging to the same crystal symmetry as quartz, they are not pyroelectric and possess a higher sensitivity. Unlike quartz or gallium orthophosphate, they have no phase transition up to their melting point (above 1300°C), so that their properties remain very stable up to very high temperature and no twinning occurs. However the growth of large crystals is more difficult than for quartz, although langasite crystals up to 4 inches diameter have been grown. Gallium orthophosphate has the same crystalline structure as quartz. Its sensitivity is twice that of quartz and is practically constant up to 500°C. Its phase transition lies around 970°C (573°C for quartz), extending the useful temperature range to at least 600°C. It is however very difficult to grow (the growth lasts a few months to one year) and is not available as large crystals. PZT-based piezoceramics and Lead-Metaniobate have a very large piezoelectric sensitivity (up to 100 times that of quartz), but aging (time dependent depolarization), poor linearity and huge pyroelectric effect limit its use to applications for which accuracy is not critical. High temperature piezoceramics (bismuth titanate based materials) can be used up to 500 or 600°C, but suffer similar (although not as serious) problems as PZT. Their sensitivity is about 5 to 10 times higher than that of quartz. They are used in high temperature applications (e.g., accelerometers operating up to 600°C). Pressure Sensors 437 Electronics As already mentioned in the introduction, piezoelectric transducers are active systems (they do not require any power supply) which have a charge output (high-impedance output). For data acquisition and signal analysis, the charge output must be converted to a voltage, for instance by means of an electrometer or a charge amplifier. Charge amplifiers A charge amplifier is basically a high-gain inverting voltage amplifier with a very high input insulation resistance configured as an integrator. A typical measuring chain with a charge amplifier is shown in Figure 16.2.2; examples of laboratory and indus- trial charge amplifiers are shown in Figure 16.2.3. The most common output voltage U out is ±10 V. The range capacitors C r can be switched usually between 10 pF to 100 nF, allowing for measurements over a wide charge range. A time constant can be switched on with the resistor R t (usually 1 GΩ, 100 GΩ). The Reset/Operate (R/O) switch allows setting the zero point of the charge amplifier. Figure 16.2.2: Measuring chain with charge amplifier. The sensor is modeled as a current source (charge source) in parallel with the sensor capacitance C s and resistance R s . The properties of the cable are described by a capacitance C c and an insulation resistance R c . The charge amplifier consists of a high- gain amplifier with a range capacitor C r , a time constant resistor R t and a reset/operate (R/O) switch. The output voltage of the ideal charge amplifier (infinite open loop gain A, no leakage current, no offset voltage at the input) depends only on the induced charge and the range capacitor, where the input impedance has no influence. Chapter 16 438 Output signal U out = –Q / C r Lower frequency limit (–3dB) f l = 1 / (2π · R t · C r ) If no resistor R t is selected, the charge amplifier operates in DC mode and the steady- state behavior is governed by drift. These relations are sufficient for most applications. In some extreme cases, the properties of the real charge amplifier must be taken into account: Upper frequency limit f u = 200 … 500 kHz When operating at frequencies above 100 kHz, the input impedance can no longer be neglected, as the open loop gain of the amplifier depends on frequency. Drift due to leakage current I L < 10 fA (MOS-FET), I L < 100 fA (J-FET) Leakage currents cause a drift in the output voltage, which eventually bring the amplifier to saturation. The time dependent charge Q L = I L · t generates a time dependent output voltage U out (t) = –I L · t / C r . Drift due to offset voltage and low input resistance U off ≈ a few mV The offset voltage at the amplifier input induces a current I d = U off / (R s // R c ). As for leakage currents, this current may bring the amplifier into saturation. Some charge amplifiers have a built-in zero adjustment to keep the drift to a very low level. Should the input resistance be very low (for instance when measuring at high temperature), switching on a time constant resistor R t or adding a coupling capacitor in series between sensor and amplifier might solve the drift problem. In both cases, the lower limit frequency f l increases. High input capacitance (> 1 µF) U Q C A C C A C out r s c r = − + + + ⋅       1 1 In applications for which very long cables are needed, the cable capacitance cannot be neglected, in particular if the open loop gain of the amplifier is not very high. This results in a decrease of the output signal. [...]... requires sensors with equally balanced capacitances on each signal line 440 Pressure Sensors Sensor Design and Applications Basic design connector sensor housing contact spring preloading sleeve sealing surface crystal element spacer ring diaphragm Figure 16.2.5: Basic design of a pressure sensor for general applications The description of each component is found in the text Sensor housing The sensor. .. very small sensors 443 Chapter 16 Standard sensors for general pressure applications Sensor with longitudinal crystal cuts Sensor with transverse crystal cuts Figure 16.2.7: Sensors for general pressure applications Left: Design with longitudinal cuts (crystal discs) Right: Design with transverse cuts (crystal slabs) Both sensors shown in Figure 16.2.7 have the same sensitivity While the sensor on the... uncooled pressure sensors are used These sensors are designed for continuous operation and feature a long lifetime (Figure 16.2.13b) Often, the measurement systems include an integrated charge amplifier customized for specific applications a b Figure 16.2.12a: Combustion pressure sensors with an M5 mounting thread (Courtesy of Kistler.) Figure 16.2.12b: Small pressure sensor integrated in a M12 spark plug... that would limit the life expectance of these pressure sensors Further, the front end of these sensors can be shaped to conform to the surface of the mold Most sensors are mounted directly into the mold without the addition of a protective sleeve This reduces the size of the sensor, making it suitable in molds used for very small parts Direct mount sensors have typically diameters from 4 to 8 mm (Figure... 448 Pressure Sensors crystal elements gap a b c Figure 16.2.11: Sensors for cavity pressure measurements a): front diameter 6 mm b) and c): front diameter 2.5 mm (courtesy of Kistler) Sensors for engines Piezoelectric pressure sensors were used early on to measure combustion pressure in engines Today, most piezoelectric pressure sensors are used for the development of automotive engines New sensors are... or glow plugs (Figure 16.2.12b) 449 Chapter 16 Water-cooled sensors improve zero stability (Figure 16.2.13a) Initially, water cooling of the sensors was required to protect the sensors from the high temperatures of the engines Modern sensors can be used without water-cooling, hence will not be damaged if for some reason the cooling is interrupted The smallest water-cooled sensor today has an M8 mounting... insulated pressure sensors ground insulation to ground crystal elements a: coaxial b: 2-wire Figure 16.2.9: Ground insulated sensors a): coaxial design b): 2-wire design 446 Pressure Sensors Ground isolated sensors reduce ground loop hum which arises when big potential differences build up between the sensor and the measuring chain (electronics) This may happen when electronics and sensors are connected... Chapter 16 Sensor Selection The sensor manufacturers can help you find the most appropriate sensor for your application Data sheets are nowadays available on the internet and together with search functions or application oriented tools greatly facilitate the search for the right sensor Hereafter is a list of the most relevant sensor and electronic properties you need to know before starting a search: Sensor. .. sensors Miniaturization Sensors have to be smaller This is a consequence of the trend observed in many application areas For instance, engines are always more compact, requiring sensors of diameters M5 or even smaller In injection molding of plastics, sensors with diameter 2.5 mm for pressure measurements up to 2,000 bar are now commonly used Reducing the size of sensors is a challenging process Apart... stopped at will to replace a defective sensor Some manufacturers offer sensors operating up to 600 or 700°C, based on tourmaline, lithium niobate or high-temperature piezoceramics 455 Chapter 16 Sensor identification Each sensor is calibrated by the manufacturer and comes with a so-called calibration sheet Before using the sensor, the operator has to enter the correct sensor sensitivity in the charge amplifier . signal. This requires sensors with equally balanced capacitances on each signal line. Pressure Sensors 441 Sensor Design and Applications Basic design Sensor housing The sensor housing protects. for very small sensors. Chapter 16 444 Standard sensors for general pressure applications Sensor with longitudinal crystal cuts Sensor with transverse crystal cuts Figure 16.2.7: Sensors for general. http://content.honeywell.com/sensing/prodinfo/pressure/technical/c15 _125 .pdf 3. “Pressure Sensors Plumbing and Mounting Considerations,” Honeywell, Inc. http://content.honeywell.com/sensing/prodinfo/pressure/technical/c15 _121 .pdf 4. “Pressure or Force Sensor

Định dạng
Số trang	40
Dung lượng	649,6 KB