The following material is excerpted from Sensor Business Digest
edited by Peter Adrian and published monthly by
Vital Information Publications
754 Caravel Lane, Foster City, CA 94404
Tel/Fax

Microchip Helps Drive Smart Sensing

Microcontrollers and other embedded control products play a vital role in enhancing the intelligence capabilities of the sensor and sensor-based applications. Having smart functionality--such as signal conditioning, amplification, A/D (analog-to-digital) conversion capability, memory, and control logic--in the sensor package or sensing system can render the sensor's output signal more useful for a particular application and enable the sensor to cost-effectively enhance the performance of the product, control system, or sub-system in which it is incorporated.
sen

Microchip Technology, Inc. (Chandler, AZ, ) (www.microchip.com)(Nasdaq:MCHP)--formed in 1989 via the purchase of a nonvolatile memory chip operation owned at that time by General Instrument Corp.--is finding expanding sensor applications for their microcontrollers, stand-alone analog devices and other embedded control products.

David Zehrbach, manager, strategic marketing, noted that Microchip's microcontrollers are used in conjunction with a wide range of sensors (such as those that measure pressure, acceleration, force, and temperature). Microchip typically sells its microcontrollers (MCUs) to users of sensors for a wide variety of sensor-related applications, such as automotive (e.g., air bag sensors), industrial (e.g., motor controls, compressors, thermostats, utility metering, process control, robotics, smoke detectors, pumps), office automation (e.g., PC servers), and consumer (e.g., remote controls, carbon monoxide detectors, consumer electronics and home appliances).

A key area for Microchip's 8-bit MCUs consists of CO detectors, where the company's microcontrollers can work with various gas sensing technologies (including tin oxide semiconductor and colorimetric/biomimetic methods). Their microcontrollers are typically not used when a large amount of digital signal processing is required at the sensor (such as in medical ultrasound sensors).

Damon Chu, strategic marketing manager PICmicro, explained that the largest segment of the overall microcontroller market consists of 8-bit MCUs, adding that more 8-bit MCUs are used in sensor applications than 16- or 32-bit MCUs. He noted that 4-bit MCUs tend to be widely used in kitchen appliances, such as rice cookers that incorporate temperature sensors. Most of the MCUs Microchip provides for sensing applications are 8-bit microcontrollers. The company is finding opportunities in sensor applications for devices with greater than 10-bit A/D resolution.

Chu added that Microchip's customers, many of whom develop their own protocols, are looking at wireless data transmission. In sensor applications, Microchip is beginning to see customers who would like to send wireless information at lower frequencies. For such wireless transmission to proliferate, the per node cost of protocols would have to substantially decrease.

Microchip's 8-bit microcontrollers allow for flexible programmability and, therefore, could provide an economical design for building the MCU into the sensor to achieve distributed intelligence (where the sensor would serve as a network node). However, there are key environmental and packaging issues to be addressed in order to integrate the MCU into the sensor.

Paul Pickering, product marketing manager, Microperipherals Products Division, noted that Microchip's stand-alone analog-to-digital converters are used in a wide range of industrial applications; and their microcontrollers with an integrated analog-to-digital converter are used in a broad range of standard industrial sensors, such as those that measure temperature, pressure, force, or current. A microcontroller with a built-in A/D converter is especially suitable for sensors that provide a signal conditioned 0-5V analog output. A sensor that provides a digital output would not require a microcontroller with an on-board A/D converter.

Pickering explained that the typical stand-alone analog-to-digital converter can provide better performance with respect to non-linearity and total harmonic distortion than a microcontroller with an integrated A/D converter of comparable resolution. He added that customers in some cases desire stand-alone analog-to-digital converters that provide a more cost-effective solution and offer lower power consumption, increased resolution, and compact size.

Opportunities for Microchip's microcontrollers and assorted embedded control products in sensor-based and other applications are proliferating, as manufacturers increasingly use integrated circuit-based embedded (buried) control systems to achieve an integrated solution for application-specific control requirements. Embedded control systems allow manufacturers to differentiate their products, replace less efficient electromechanical control devices, enhance product functionality, and reduce product cost.

Microchip notes that the microcontroller, which is typically the principal active component in an embedded control system, is a self-contained computer-on-a-chip that includes a central processing unit, non-volatile program memory, random access memory for data storage, and various input/output capabilities. In addition to the microcontroller, a complete embedded control system incorporates application-specific software and may include specialized peripheral device controllers and external non-volatile memory components (e.g., EEPROMs for storing additional program software).

Microchip Technology's embedded control products (which include microcontrollers, application-specific standard products (ASSPs), related mixed-signal and memory products, and application development systems) are used in a plethora of applications in the consumer, automotive, office automation, telecom, and industrial markets. The company specializes in field programmable and mixed-signal solutions, and also offers complementary microperipheral products, including stand-alone analog, interface, and microID (TM) RFID devices; serial EEPROMS; and KEELOQ® security devices.

Microchip's PICmicro® microcontrollers have a proprietary architecture, which features dual data instruction pathways (Harvard dual-bus architecture), RISC (reduced instruction set computing), and variable-length instructions. The company's microcontrollers, priced from about $0.49 to $10.00 per unit, provide speed advantages over microcontrollers based on single-bus, CISC architectures. Since their introduction in 1990, over one billion PICmicro® microcontrollers have been shipped.

Microchip initially focused on the low-cost segment of the 8-bit microcontroller market. With their baseline products, they have become a leading supplier of field programmable microcontrollers. Over the past five years, Microchip has introduced over 120 microcontrollers aimed at the baseline, mid-range, and high-end segments of the traditional 8-bit microcontroller marketplace. Leveraging their scalable product architecture, Microchip has expanded into the lower end of the 16-bit microcontroller market, as well as the higher end of the vast 4-bit microcontroller arena.

Microchip focuses on enhancing the design, development, and production of their PIC microcontroller products in accordance with the types of systems their customers are now building and will build in the future. They have developed their advanced, low-cost user programmability feature by incorporating non-volatile memory, such as EPROM, EEPROM (electrically erasable programmable read only memory), and flash memory, into the microcontroller, in addition to masked ROM program memory being included in the microcontroller.

In late September, Microchip introduced the PIC16F7X mid-range family of low-cost, high-performance flash microcontrollers with an integrated 8-bit A/D converter and only 35 single-cycle instructions (each 14 bits wide). Implemented on Microchip's new 0.5 micron process technology, the PIC16F73/74 and PIC16F76/77 allow OTP customers to migrate to a more flexible flash-based platform. The size reduction from the single transistor flash cell allows these new flash devices to be priced lower than Microchip's OTP counterpart devices.

In quantities of 1,000, the PIC16F73 is priced at $3.43 per unit; the PIC16F74 is priced at $4.15, the PIC16F76 is priced at $4.86; and the PIC16F77 is priced at $5.22 each. Engineering samples are targeted for December, and release to production is slated for Q1 2001.

Featuring a patented integral multi-channel A/D converter, the PIC16F7X flash microcontrollers offer high precision and accuracy without significant power consumption, rendering them highly suitable for battery-powered applications. Moreover, the A/D converter streamlines the overall embedded system design by providing a direct analog interface for sensors that measure such parameters as temperature, pressure, and motion. This capability is beneficial when implementing cost-effective embedded solutions for industrial controls, instrumentation, high-performance PC peripherals, appliances, office automation, automotive controls, and vehicle and home security applications. Such applications require multiple I/O sensing (multiple inputs from sensors), high processing speed, and extensive communication with other processors or resources.

Other on-chip peripherals include a timer subsystem, providing a real-time clock, or two 8-bit and one 16-bit counter/timer modules. The low-power PIC16F7X devices typically draw 20 µAmps from a 3 V supply while operating at 38 KHz. Integral "sleep" mode instruction and the ability to switch off the A/D circuitry when it is not being used typically decreases current draw to less than 1 µAmp.

In May, Microchip announced a road map of 37 new PIC16FXXX and PIC18FXXX flash microcontrollers to facilitate field upgrades and Internet connectivity across numerous embedded applications.

Self-programming capability enables remote upgrades to the flash program memory and end-equipment through various media, ranging from Internet and modem to RF and IRDA. Microchip's PICmicro flash and OTP (one time programmable) microcontrollers also feature In-Circuit Serial Programming (ICSP) (TM) technology, allowing the microcontroller to be programmed after being placed in a circuit board. The in-circuit debugging feature uses the ICSP interface, enabling the user to emulate and program the microcontroller through a single interface.

Microchip has noted that flash-based microcontrollers are finding increased usage in applications where there is an existing communication infrastructure, or a planned infrastructure, to support remote upgrades. Flash devices allow for self-programming from a remote location, and they accept data from an external source via a peripheral, such as an on-chip USART. The flash device then takes this data and writes into its own program memory over the device's entire operating and frequency range.

The new flash devices extend Microchip's Migratable Memory (TM) technology, which provides socket and software compatibility among all equivalent flash, OTP, and ROM memory controllers in the PICmicro family. Migratable Memory technology allows customers to select the microcontroller memory technology best suited to the product life cycle of their application and provides a smooth migration path to lower-cost solution when appropriate.

Demand for flash memory microcontrollers is being spearheaded by the expanding opportunities for Internet and networked connectivity in consumer appliances and a variety of other areas. By incorporating flash memory on the chip, the microcontroller is able to reprogram itself, allowing products and applications to be upgraded with enhanced functionality. This task can be accomplished by downloading software via the Internet into the appliance.

Products/applications where flash microcontrollers allow for enhanced features and functionality include: automotive susbsystems (e.g., upgrading the body control module); networked home appliances (e.g., refrigerators, dishwashers, air conditioners); home medical appliances (e.g., remote diagnostic capability for blood sugar analyzers, blood pressure monitors, and infant apnea monitors); remote controls (that are used, for example, with televisions, VCRs, or DVD players); parking meters (e.g., allowing status information, downloading new information to the meter, and recording the amount of money or coins deposited); and vending machines (e.g., enhanced inventory management, updating prices for products in the machine, machine repair).

Opportunities in the sensing arena for Microchip's flash microcontrollers are being propelled by the advent of smart sensor nodes whose calibration values must be regularly updated. Chu noted that customers are embracing OTP and flash microcontrollers, as the total cost of ownership for such devices is becoming lower than that for a raw microcontroller. Furthermore, the flash microcontroller allows the system designer to readily change the algorithms and calibration values of the microcontroller that is used with sensors.

In late September, it was reported that Realogy's (York, England) Real-Time Architect development environment and SSX5 Real-Time Operating System (RTOS) will support Microchip's PIC18CXXX architecture with a small footprint OSEK-enabled run-time kernel.

Launched in January, Realogy's Real-Time Architect is a new development and deployment solution for real-time system development. Based on recently developed real-time scheduling technology, it is designed to meet the requirements of developers using resource-constrained deeply embedded target hardware in such applications as automotive.

The time compiler component of Real-Time Architect is based on new extensions to Deadline Monotonic Analysis, a branch of real-time schedulability analysis. The time compiler uses this mathematical technique to calculate the worst-case response time for each task and interrupt service routine in the system. This analysis shows whether all user defined deadlines allocated to an application will be met in all circumstances. The time compiler takes into account all possible scenarios, including interference from higher priority tasks/interrupt handlers and blocking effects due to resources shared with low priority tasks. It also takes full account of operating system overheads.

In the automobile, embedded microcontrollers and on-chip peripherals are used to carry out a diverse range of sensing requirements, pertaining to such applications as climate control systems, stability control systems, alarm systems, engine management systems, and automatic windshield wipers. Since the auto industry tends to embrace simple and robust sensor technology, significant software processing of the sensor inputs is often required. Sensor processing is often part of the complex software system, which runs assorted control algorithms, communications, activities, and diagnostics at various data rates. The RTOS facilitates the development of such software and can enhance the software's efficiency.

Microchip's PIC18CXXX architecture, targeted at 8- and 16-bit microcontroller applications, is upward compatible with the company's mid-range PIC12C6XX core and PIC16CXXX core and high-end PIC17CXXX core. The PIC18CXXX architecture provides up to two million bytes of program memory address space, a C compiler friendly development environment, and 10 MIPS performance at 40 MHz.

The Real-Time Architect offers the SSX5 full-function pre-emptive kernel and associated development tools, providing PICmicro users with multi-tasking capability. Realogy's SSX5 run-time kernel is purportedly the first RTOS that is able to offer the small footprint and high functionality necessary to provide a viable solution for designers of deeply embedded systems using PICmicro devices. The Real-Time Architect for the PIC18CXXX family will be available in October, with development seats priced down to $7,500.

The microcontrollers with the new PIC18CXXX architecture have a broad range of automotive applications outside the engine and power train areas, such as climate control, seat control, airbag sensors, security systems, lighting controls, cruise control, radar detectors, turn signals, anti-lock braking, active suspension, fuel pump control, fuel injection, sunroof control, and tire pressure monitoring.

In September '99, Microchip unveiled 8-bit microcontrollers featuring an A/D converter, capture/compare/pulse width modulation capsule, and on-chip voltage reference and available in a very small 20-pin package. The 6-channel 12-bit ADC on the PIC16C770 and PIC16C771 and the 6-channel 10-bit ADC on the PIC16C717 provide a level of resolution previously requiring a stand-alone ADC. This capability allows designers to discriminate smaller signal changes, which is important when interfacing with a wide range of sensors. No external circuitry is required for high precision measurement of analog signals.

In quantities of 1,000 (industrial temperature versions), the PIC16C717 is priced at $2.84, while the PIC16C770 and PIC16C771 are priced at $3.69 and $4.11, respectively. Applications for the PIC16C717, PIC16C770, and PIC16C771 include motor controllers, instrumentation, remote sensing, battery charging, uninterruptible power supplies, switching power supplies, automotive actuators, data loggers, environmental monitoring, and medical monitoring.

Microchip's mixed-signal products consist of an expanding portfolio of analog devices that are primarily used in embedded control systems to convert or buffer input and output signals to or from a microcontroller. Their stand-alone analog products were introduced to the marketplace at the beginning of FY 1999 (which ended March 31, 1999). Microchip's analog and interface IC products include system supervisor ICs, operational amplifiers, CAN (Controller Area Network) controllers, and analog-to-digital converters.

In April 2000, Microchip introduced the MCP300X family of ADCs available in 1-, 2-, 4-, and 8-channel versions, respectively. Based on a successive approximation register architecture, the MCP300X ADCs offer 200K samples per second throughput, low power of 500 microamps active (MCP3001) and 5.0 nanoamps standby, a wide supply voltage of 2.7-5.5 volts, extended industrial temperature range of -40° to +85°C, +/- 1 LSB DNL and +/- 1 LSB INL max. at 200K samples per second, guaranteed no missing codes, and a serial output with an industry-standard SPI (TM) bus interface.

In quantities of 1,000, the MCP3001 is priced at $1.75, and the MCP3002, MCP3004, and MCP3008 are priced at $1.83, $1.93, and $2.02, respectively. Applications for the MCP300X 10-bit analog-to-digital converters include data acquisition, instrumentation, and measurement, multi-channel data loggers, industrial PCs, motor controls, robotics, industrial automation, smart sensors, portable instrumentation, home medical appliances, smoke detectors, automotive navigation, and CD motor control.

For FY 2000 ended March 31, 2000, Microchip had net sales of $495.7 million, a 22% increase over the prior year. Microcontrollers and associated application development systems accounted for 80% of overall net sales. Excluding restructuring and acquisition-related special charges/special income, net income in FY 2000 was $100.3 million and diluted earnings per share rose 38% to $1.23 in FY 2000.

For Q2 of FY 2001 ended September 30, 2000, Microchip had net sales of $176.3 million, an increase of 49% from the prior year's second quarter. Diluted earnings per share (reflecting a 3-for-2 stock split effected on September 26) were $0.33, up 76% from that of the second quarter last year. For the six months ended September 30, 2000, Microchip had net sales of $334.047 million, net income of $79.194 million, an diluted earnings per share of $0.63.

Worldwide revenues from shipments of 8-bit microcontrollers totaled $6.881 billion in 1999, representing about a 20.7% increase from the $5.697 billion recorded in 1998, according to Gartner's Dataquest (San Jose, CA, ) (June 2000). Global unit shipments of 8-bit microcontrollers totaled 3,392,035,000 in '99, an increase of about 30.1% from 2,606,397,000 shipped in '98. In 1999, according to Dataquest, Microchip Technology was the number four supplier of 8-bit microcontrollers based on worldwide revenues and the number two supplier based on unit shipments.

In 1999, global revenues for smart sensors with signal conditioning/amplification capability totaled about $3.14 billion, according to SBD. The data refers to pressure sensors equipped with signal conditioning and amplification (which accounted for revenues of about $972 million in '99); flow sensors/ flow meters with built-in, integral, or remote signal conditioning (representing revenues of about $1.9 billion in '99); force sensors/load cells with built-in signal conditioning (accounting for revenues of about $24 million in 1999); and temperature sensors with on-board signal conditioning (representing revenues of about $246 million in '99).

NOTE: On October 27th, Microchip reported that it will acquire TelCom Semiconductor (Mountain View, CA) (Nasdaq:TLCM) in a stock-for-stock transaction valued at about $300 million. TelCom provides linear and mixed-signal integrated circuits solutions, including IC temperature sensing devices. The transaction is expected to be completed during the first calendar quarter of 2001.

Non-Technical Issues Significantly Impact Sensor Product Development

SBD emphasizes that the successful development and commercialization of a sensor product depends on solving issues that do not strictly pertain to the sensor's technical specifications, performance or packaging capabilities. While technical expertise is extremely valuable for realizing cutting-edge yet practical sensing solutions, other factors profoundly influence the opportunities for an innovative sensor product, including knowledge of the key markets/applications for the sensor and the ability to adapt to changing external or internal business conditions.

Dr. David Wadlow, founder and president of Sensors Research Consulting, Inc. (SRC, Basking Ridge, NJ, )--a scientific consulting firm that supports innovation in physical sensors and sensor-based products by providing front-end research with respect to technical problem analysis, new concepts, and prototype development--notes that people often operate under a misconception that the successful solution of technical problems is the key to successful development of an entirely new sensor or sensor-based product.

The brave sensor pioneer can be shocked when he/she discovers that solving technical problems, no matter how challenging, is often less than half the battle, and the remaining problems are usually even more challenging. In such cases, there is the risk that the core technical solution, no matter how innovative, patentable, or technically successful, will subsequently be poorly managed and the project may stall or fail for other reasons.

A long list of factors are involved in transforming a superior idea into a commercial reality. In Dr. Wadlow's experience, performing the research and developing the technical solution for an innovative sensor typically represents only 40% or less of the efforts required for the overall project. Developing a new technology involves not only proving a concept via feasibility analysis and prototype development. Subsequently, there is a long path through engineering development, manufacturing, and product launch. Each stage is at least two- or three-fold more expensive than the preceding one, and the number of detailed tasks expands exponentially.

Moreover, human factors are important. Managing a project can gradually turn into a nightmare for the lone innovator, while, in large corporations, the project may require protection from internal politics and competing claims on the budget as it consumes more "corporate space." Experts can offer advice on product development processes, which can work well for large corporations because they provide a project focus, improve communications, and can relieve internal stresses (i.e., reduce the shock of success).

In a perfect world devoid of corporate politics, the fundamental requirements are still the existence of a potentially lucrative market, adequate business resources, knowledge of the technical requirements and preferences in the areas where the sensor may be applied, and sound technical resources and expertise.

The most important additional ingredient for success is the ability to cope with unforeseen problems that will inevitably arise. This phenomenon dictates teamwork across disciplines and cultures, adaptive planning, and probably some compromises in order to achieve a practical and commercially successful product. Moreover, as the technical solution emerges, unless one intends to license it, one confronts the task of managing a business as well as the project.

SRC, whose clients range from small to mid-size and Fortune 500 companies, as well as non-profit institutions, is thoroughly familiar with the range of issues and vicissitudes involved in sensor product development. For example, one client, a small entrepreneurial company led by an energetic individual, had a clear product concept requiring the development of a specialized, sensitive proximity sensor that would operate an actuator. The novel actuator was being developed by a West Coast engineering design firm.

SRC, which joined the project after it had been underway for two years, invented the sensor where others had failed and then developed a prototype that worked according to requirements. Technical success was achieved in just over one month. Subsequently, however, market influences led the client to alter the specifications for the actuator. The new specifications were significantly different from the current actuator design; and this single change placed the entire project on hold.

This situation illustrates the problem of dovetailing a product concept with changing market conditions without jeopardizing technical feasibility. In this example, an altered objective presented a serious problem in product development. The client broke a traditional tenet of product development by changing the specs halfway through the project. However, during the course of a project, market changes need to be addressed and compromises made so that the initial product concept can evolve and still be commercially successful.

In his previous experience as a research scientist in corporate R&D, Dr. Wadlow encountered numerous instances of difficult, non-technical problems in product development. For example, he led a team that successfully developed a new type of multi-gas analyzer to solve a process control problem for a high-value product. To accomplish the task, he contributed new science involving plasmas.

However, in the corporate giant's home country overseas, others had developed a different solution, which was not as fast or reliable but worked and had a head start. Internal politics and poor interdepartmental communication impeded further development of the multi-gas sensing device. The sensor was patented and used internally on a limited, though lucrative basis, but it was never fully developed and officially commercialized.

Every sensor development project is different, and each stage of the process requires creativity, Dr. Wadlow points out. As the project demands a greater degree of innovation, careful planning becomes more critical in order to accommodate the project's growth and possible changes in market and business circumstances, which may occur up to commercialization. While solving technical problems is imperative, the need to solve the other, significant issues cannot be ignored. SRC considers teaming with the client, and planning and communication, to be as vital for successful sensor development as providing technical solutions.

Civilian, global demand for sensors used in R&D totaled $1.3 billion in 1998 and is projected to reach about $1.8 billion in 2008, representing an average annual growth rate of about 3.3%, according to Intechno Consulting's (Basel, Switzerland, +) Sensor Markets 2008 report.

Humirel Spearheads OEM Opportunities for Humidity Sensors

In July, Humirel SA (Toulouse, France, +) (www. humirel.com)--a specialist in humidity sensor systems based on innovative capacitive polymer relative humidity (RH) sensor technology and a spin-off from Motorola in 1998--reported that they had obtained ISO9001 and QS9000 certification. They also unveiled the HM1500 linear voltage RH module, designed to provide reliable, accurate yet economical measurements in key OEM applications, such as agriculture, environmental (e.g., HVAC, building control), and process control.

ISO9001 is a global standard for the design, manufacture, and sale of products/services. QS9000, which is mandatory for suppliers to the automotive industry, is a more demanding standard, geared toward the auto industry. Humirel is reportedly the first humidity sensor supplier to receive QS9000 certification.

The receipt of ISO9001/QS9000 recognition "is the result of continuous efforts to structure our company towards the highest standards of customer satisfaction and product and services quality," stated Jean-Francois Allier, Humirel's CEO.

"The QS9000 label will reinforce our leadership position in consumer-oriented markets, where you need to be able to incorporate the highest level of quality to massive quantities of products offered at competitive prices," noted Dean Walters, Humirel's business manager for the Americas (Fountain Valley, AZ, ). The QS9000 certification enables Humirel to supply their sensors to tier one automotive suppliers.

Humirel's HM1500 module is calibrated to +/- 2% RH and provides a 1-4 Volt linear voltage output for 0-100% RH, allowing direct interface with a wide variety of microcontrollers. The compact device can be readily attached to a PC board or mounted separately to serve as a probe or remote sensor.

Due to its ease of mounting, the HM1500 is easily adaptable to existing or new designs and provides a convenient mechanical interface. Based on the rugged HS1101 humidity sensor (which has a metal can package and a plastic enclosure for reduced cost in high-volume applications), the HM1500 withstands water immersion and condensation and is insensitive to light. A sample kit, which includes two HM1500s, is available for $125. In quantities of 10, the HM1500 is priced at $39 per unit; and in quantities of 1,000, the HM1500 is priced at $23 per unit.

Allier told SBD that the HM1500 linear voltage module offers a unique, cost-effective alternative to humidity transmitters, which are bulkier and considerably more expensive (e.g., cost hundreds of dollars). The HM1500 is now being used in, for example, agricultural, environmental control, and process control applications. In the agricultural arena, applications for the HM1500 include growing fruits or vegetables in a closed environment, and warning of icy conditions. In environmental applications, the HM1500 helps determine and ensure air quality. Process control applications for the HM1500 include drying processes in the food industry.

Walters noted that promising target markets/applications for the HM1500 in the U.S. include tobacco drying, data loggers/data acquisition systems, and HVAC.

In 1998, Humirel acquired the rights to Motorola's humidity sensor technology (i.e., the equipment and intellectual property pertaining to the packaged humidity sensor product). Under the agreement, Humirel can develop, manufacture, and market humidity sensing devices based on such technology worldwide.

Motorola retained a minority share of Humirel's capital stock in association with Banexi Ventures (Paris, France), a venture capital firm that also owns a minority share. The remainder of Humirel is owned by private investors.

Motorola's humidity sensor technology became available as a result of the cessation of the chemical sensor program (which was not aligned with Motorola's strategic direction) in autumn 1997. (MicroChemical Systems, headquartered in Corcelles, Switzerland, was formed in 1998 as a result of a buyout of Motorola's chemical sensor group.)

Humirel's capacitive humidity sensors use a multi-layer polymer sensing element/chip in greater than 8" diameter wafers, which are diced into separate sensor chips. The proprietary sensor structure and manufacturing process are patented. Compared to their resistive counterparts, capacitive humidity sensors offer such advantages as linearity, rapid response, wide humidity and temperature range, essentially no hysteresis, stability and repeatability, low temperature coefficient, and low cost. The unique benefits of Humirel's sensors, Allier explained, include ruggedness and the ability to be readily produced in very large volumes.

In 1999, Humirel--which provides humidity sensors for automotive, consumer, appliance, and industrial applications--had sales of $1.5 million. Major application areas for Humirel's humidity sensors include automotive (e.g., fog prevention and comfort/climate control); appliances (e.g., humidifiers/dehumidifiers, clothes dryers, dishwashers, refrigerators), and printing and reprographic equipment (e.g., photocopiers).

Humirel's sensors are being designed into European and U.S. vehicles for anti-fog and comfort/climate control applications. Humirel's sensors are expected to be used in MY 2002 vehicles, which will begin to appear by around June of next year.

In automotive comfort/climate control, humidity sensors, used along with temperature sensors, allow for efficiently maintaining an improved comfort zone for passengers, Walters noted. The use of humidity sensors in climate control allows for maintaining a lower level of humidity, which permits a higher temperature in the passenger compartment without causing passenger discomfort. The amount of air conditioning required would be reduced, improving the vehicle's performance. Using humidity sensors to detect the accumulation of condensation on the winshield would allow for automatic defogging before the accumulation reaches a level that impairs the driver's visibility.

Humirel's humidity sensors are also being designed into European- and U.S-built appliances, such as clothes dryers; and the company's sensors are being designed into humidifiers/dehumidifiers that are built in various regions, including the Far East. Moreover, Walters explained that Humirel's sensors are finding opportunities for use in U.S.-built appliances (including dishwashers and refrigerators), as U.S. manufacturers implement enhanced electronic control in appliances, which allows them to cost-effectively add sensing capability.

Humirel's humidity sensors are being used in printing and reprographic equipment worldwide. Walters noted that humidity is key variable that impacts the functionality of such equipment. For example, in large area printers, humidity can affect print quality and the speed of the printer. In reprographic equipment, it is important to know the humidity near the toner cartridge. In an extremely wet environment, the ink has a tendency to smear. A dry environment can lead to an insufficient amount of ink being applied to a page.

Global revenues for humidity sensors at the component level (excluding humidity instruments) totaled on the order of $50 million in 1999, of which about $18 million was accounted for by North America, according to SBD.

The worldwide market for automotive HVAC temperature sensors totaled $33.6 million in 1998 and is projected to increase at a 16.6% compound annual rate to reach $72.3 million in 2003, according to Strategy Analytics (Boston, MA, ). The global market for automotive HVAC pressure sensors (which are used to monitor refrigerant pressure) totaled $8.8 million in 1998 and is forecast to rise at a 29.6% compound annual rate to reach $32.3 million in 2003.

Strategy Analytics in Luton, England (+) notes that humidity sensors have opportunities for use in new, more efficient automotive air conditioning systems, where the air entering the passenger compartment is wetter and has a higher relative humidity, which increases the risk of fog appearing on the windshield. Moreover, there is even a greater risk of windshield fogging in newer automotive climate control systems, which include occupant temperature sensors and have faster climate control loops leading to more rapid changes in temperature.

Sensor Companies Gain By Offering Intelligent, Value-Added Solutions

SBD notes that sensor manufacturers and providers can significantly expand their business opportunities and served markets by offering enhanced sensor solutions that, in addition to sensing per se, provide such attributes as intelligence (e.g., signal processing/conditioning electronics), superior packaging, and streamlined communications capability.

Moreover, the proliferation of the Internet as a vehicle for communication of data, as well as the growth in industrial sensing and control networks, are driving opportunities for wireless sensing. In applications where multiple data points need to be monitored, wireless sensing solutions, an embryonic area at present, have key potential for eliminating the costs and inconveniences associated with wired networks.

CrossNet(TM), an integrated hardware/software solution released in August by Crossbow Technology, Inc. (San Jose, CA, ) (www.xbow.com), leverages Bluetooth wireless technology for economical, short-range radio links to enable users to create wireless sensor networks. Such networks can link dozens or hundreds of sensors of diverse types and brands with data acquisition/analysis systems, such as handheld devices, Internet-enabled laptops, or desktop PCs, and eliminate the expense and difficulties involved in hand-wired sensor applications.

Michael Dunbar, director of CrossLink strategic partnerships at Crossbow, explained that multiple sensors, which need to be cost-effectively and reliably interconnected to a control point, are increasingly being implemented in industry, driven by the accessibility of the Internet and the expanding use of industrial networks. CrossNet simplifies and reduces the cost of connecting multiple sensors and allows more convenient access to real-time sensor data. For example, CrossNet users would be able to access machine condition data via a Palm Pilot, laptop computer, or other device.

The initial CrossNet offering, priced at $1,895 for the basic system, is a general-purpose solution. Subsequently, Crossbow plans to release CrossNet products for specific applications. Target markets for CrossNet include general data acquisition, machinery health monitoring and building environmental monitoring.

Smaller, smarter, and more rugged sensors are finding expanding opportunities in test, measurement and control, and security applications. Generally, the more complex the manufactured product, the greater the number of sensors and the complexity of the wiring involved.

In large industrial facilities (such as power plants and paper mills), sensors are employed to monitor machinery health and environmental conditions. The food and agricultural industries also rely significantly on sensors to monitor conditions of perishable items in warehouses, stores, and during transport. Sensors are used for real-time environmental and security monitoring in commercial buildings housing valuable computer facilities and hundreds of workers. In addition, a promising field for networked sensors entails the consumer arena (e.g., home safety, automation, and security).

In many applications, each sensor must be individually wired to a data monitoring, acquisition, or control device. Irrespective of whether the device is an intelligent system (a PC or workstation) with analytical and communication capabilities, or a standalone box that stores data for subsequent collection and analysis, the wiring of sensors, Crossbow notes, is a tedious, manual task, typically performed by lab technicians or test engineers whose time could be more productively employed elsewhere.

Although wiring can involve considerable effort (including entering hazardous or uncomfortable spaces), it is often temporary and must be torn down or reconfigured as soon as a particular test is completed. When the application involves dozens or hundreds of sensors that monitor the status of a large building or environmental conditions in a building, the costs associated with cabling can exceed the price of a sophisticated data acquisition system. The possibility that faults in wired connections will invalidate days of expensive testing increases in as more sensors are used in the wired application. Even when all connections are correct, the wiring can act as an antenna or otherwise distort incoming data.

In sensor product test and development applications, time is a vital factor, and prototypes are available for testing only during a limited window of opportunity. The ability to achieve reduced product development cycles and compressed time-to-market can profoundly boost a company's profitability and competitive position. A primary drawback associated with wiring sensors is the loss of time, due to making, checking, and changing connections. Another key problem is the high cost of manpower and materials. Furthermore, cabling can be impractical in certain applications, such as running wires in harsh environments or connecting sensors on rotating equipment.

To counteract the limitations associated with wiring sensors, a number of companies have developed proprietary wireless solutions. Such proprietary wireless protocols typically operate in industrial/scientific/medical (ISM) bands for license-free transmission. Popular frequencies utilized to date include 40 MHz, 900 MHz, and 2.4 GHz. Crossbow notes that cost is the major barrier for proprietary wireless systems. The cost of such systems ranges from about $100 per transmitter to $10,000. However, since each sensor in the network must have its own transmitter, the cost of even a low-end proprietary system can rapidly escalate.

Owing to the high cost of proprietary wireless systems, Bluetooth--a new,standards solution, designed to provide small, low-cost radio links between computers, mobile phones, and other portable devices--is being embraced. Bluetooth devices have significant potential for replacing RS232, parallel, USB, and other types of cables with a single, standard wireless connection.

Since Bluetooth is a global standard and uses the universally available, unlicensed 2.4GHz radio frequency spectrum, Bluetooth-certified devices will communicate in the same way anywhere around the globe. For instance, a Bluetooth-certified PDA or cell phone will work with any PC that is equipped with a Bluetooth-certified card, regardless of the manufacturer of either product.

The Bluetooth Special Interest Group (SIG)--which includes such promoter companies as 3Com, Ericsson, IBM, Intel, Lucent, Microsoft, Motorola, Nokia, and Toshiba, and nearly 2000 adopter companies--is spearheading development of the technology. Cahners In-Stat Group (Scottsdale, AZ, ) expects worldwide demand for Bluetooth-enabled devices to grow to at least 1 billion units by 2005. Due to this proliferation, the costs for Bluetooth components are expected to decline substantially, enabling expanded implementation of Bluetooth into everyday products.

The benefits of a low-cost wireless solution for sensors will be particularly attractive in the following types of applications: speed-critical applications requiring rapid, convenient set-up; temporary applications necessitating frequent reconfiguration; applications where it is difficult or impossible to use wires; applications in harsh environments; applications where cabling costs can be prohibitive; and mobile applications.

For example, in new product test and development activities in manufacturing companies, the time saved by using wireless networks to test large, complex products accelerates time-to-market, thereby boosting profitability. Test labs for large companies that manufacture a variety of product lines are perpetually involved in setting up or tearing down wires. Wireless networks would conserve time and increase productivity for companies that produce a wide range of consumer appliances and electronic devices.

In certain industrial applications, such as machinery health monitoring, there are practical barriers to wiring sensors, such as the tendency of wires to distort measurement results or to become tangled due to the rotation of the test object. In applications involving extreme heat, cold, humidity, or corrosive conditions, wireless solutions offer greater robustness and reliability than lengths of exposed cabling.

Running cables to sensors dispersed throughout a large facility to monitor environmental conditions, energy usage, or security can be very expensive, driving the need to go wireless. In mobile applications, such as the transport of food and other perishables, wireless networks can provide cost and time advantages in monitoring environmental conditions.

The flexible, modular CrossNet solution is adaptable to a wide range of test, measurement, monitoring, and security applications. The three building blocks of the solution are sensors, nodes, and hubs.

CrossNet's universal architecture is designed to support a wide range of diverse types of sensors. In addition to the sensing element, the sensor, which is connected to the node, consists of signal conditioning circuitry (including calibration and temperature compensation), and the Transducer Electronic Data Sheet (TEDS). These features can be built directly into the sensor or included in a CrossNet Smart I/O (SI/O) sensor-interface cable that contains the required circuitry (microcontroller for node communications plus memory for TEDS configuration information).

CrossNet cables are available for sensors with voltage, current, thermocouple, or resistive bridge outputs. Specialized CrossNet cables can readily be developed for specific sensor configurations or technologies, and can include sensor signal conditioning and communications capabilities.

The IEEE 1451 family of standards defines the functional boundaries and interfaces required to enable smart transducers to be readily connected to a variety of networks. The standards define the protocol and functions that give the transducers interchangeability, self-identification, and network independence. One of the key elements of IEEE 1451 is the TEDS, which identifies the sensor's characteristics and allows sensors to be interchangeable in networked systems. Based on information contained in the TEDS, a host computer can recognize a temperature sensor, pressure sensor, or other sensor type, along with the measurement range and scaling information.

In a typical CrossNet application, the user mounts sensors on a machine or in a facility at the number of points to be monitored. The individual sensors are connected to CrossNet nodes. Each node controls up to four sensors and incorporates a Bluetooth radio link for wireless communication with a computer or handheld device. The maximum transmission distance is 10 meters (about 30 feet), with an optional capability to 100 meters. Since they are compact, light-weight, rugged units (3.5" x 2.5" x 3/4"), the nodes can be mounted close to the sensors they monitor. Any number of nodes can be mounted in a building or on a machine.

The CrossNet node collects data from multiple sensors and transmits the information via Bluetooth wireless communications to a network hub or Internet appliance, such as a desktop, laptop, or handheld computer. The node can supply excitation to each sensor, or sensor power can be supplied externally. Up to four channels are available on each node for analog inputs as well as digital output. The sensor's signal is digitized (16-bit A/D resolution) for transmission along with the TEDS for each sensor. This scheme allows each channel to identify itself to the host system. The node can operate from an external power supply or an attached battery.

The node can initiate operation in various ways: data can be acquired and returned upon request from the host device; an event trigger can be established to begin acquiring data based on one of sensor channels (enter or exit an alarm region or alarm threshold); or an external TTL signal can be activated to initiate data sampling.

The data can consist of a single measurement from each channel, an average of a number of measurements, or information that has been logged over a user-determined amount of time and user-defined sampling rate. In the data-log mode, a time stamp is included with the data. The initial CrossNet product will support a sampling rate of up to 100 Hz for each channel. Future versions will support higher-speed sampling in excess of 5 kHz.

The CrossNet hub, which can be any Bluetooth-enabled device for data acquisition or analysis, downloads data (including the TEDS information for sensor self-identification) from multiple nodes. Application software running on the host system displays the data in a user-defined format or links the data to a number of popular Windows-based data acquisition and software packages, such as National Instruments' (Austin, TX) LabVIEW (TM) or Excel.

The hubs can range in complexity from single computers or handheld devices that log or display data to Web servers that communicate using a TCP/IP Internet communications protocol for wide area networking. Using the TCP/IP protocol, the hubs can be connected to other wired or wireless network architectures, including Ethernet and LonWorks, and other protocols. Crossbow is developing a family of network hubs, or gateways, that permit multiple CrossNet nodes to be connected to the Internet or other network architecture.

Crossbow developed Windows-compatible software for users of laptops or desktop PCs that is capable of receiving data and transmitting commands from or to any number of sensors attached to CrossNet nodes. The data is presented in a format that can be readily converted into any application program based on LabVIEW, Visual Basic, or C++. Remote users connected via the Internet can have real-time access to the CrossNet data, enabled by a WebWare program located on a PC. For PDA users, a compact program has been developed that configures the PDA to receive data from any number of CrossNet nodes. Most programs are available to the customer free of charge.

Crossbow's CrossLink partnership program is designed to bring together companies interested in contributing hardware or software technologies to the CrossNet solution. For instance, sensor companies that combine their sensors with the TEDS or incorporate the node functionally into their product (to form a sensor featuring wireless data communication) can provide data input to the network. Vendors of data acquisition systems can configure CrossNet systems for specific customers, providing a vital systems integration function. Software partners can tailor applications software to meet the needs of specific customers.

The CrossLink program also includes Internet appliance manufacturers who offering innovative ways of accessing and displaying sensor data. Various joint marketing and engineering support arrangements are available within the program. Target members of the CrossLink program include sensor manufacturers (who gain the ability to offer wireless communications by joining the program), systems integrators or large users of data acquisition systems, and providers of related hardware/software products or services (such as software for machine health monitoring).

Founded in 1995, Crossbow Technology provides low-cost, intelligent, digital sensor solutions, which integrate silicon micromachined (MEMS) sensing elements with digital signal processing. The sensors, which are purchased from other companies, incorporate Crossbow's proprietary SoftSensor (TM) embedded firmware, which includes algorithms for stabilization and navigation applications coupled with internal compensation and communication functions.

Crossbow's products include a wide range of acceleration, tilt, inertial, and magnetic direction and orientation sensors, as well as sensor-based analog and digital subsystems. At present, the company's primary sensing products are inertial measurement systems/gyros used for such applications as aerospace/military and industrial. Crossbow's products are used by over 1,000 customers in demanding industries. Applications include autonomous vehicle control, machine health monitoring, precision farming, smart munitions guidance, automotive testing, avionics, and marine platform stabilization.

In 1997, U.S. consumption of industrial sensors with embedded circuitry that are directly connected to the various buses totaled 58,000 units, according to Venture Development Corporation's (Natick, MA, ) report on The U.S. Market for Industrial Automation Products Incorporating Device/Sensor Buses: ASI, CAN (DeviceNet, SDS), Interbus-S, LonWorks, Profibus DP, and Seriplex, Second Edition. U.S. consumption of industrial sensors with embedded circuitry that are directly connected to these buses is projected to rise at about a 60.8% compound annual rate to reach 624,000 units in 2002. The data predominantly pertains to wired devices.

Quantum Expands Markets for CO Detection

Quantum Group (San Diego, CA, /) (www.qginc.com)--a developer and provider of carbon monoxide (CO) detectors and sensors that use the company's biomimetic sensing technology, which react to CO similar to the way hemoglobin in the blood reacts--is expanding into key markets that offer growth opportunities, including the automotive after-market and personal monitoring arenas.

Under a licensing agreement with GM, Quantum will manufacture portable CO detectors for vehicles that are sold under the AC Delco brand name. The CO alarms have been designed to meet GM's requirements for durability, temperature range, and humidity range. Opportunities for the product will be boosted by its availability through GM's dealer network. Shipments of the AC Delco-branded CO detectors, which will be sold primarily through mass merchandisers and auto supply retailers (e.g., Kragen), are targeted to begin late this year. The suggested retail price for the CO alarm for vehicles is expected to be $49.95.

The CO detectors sold under the AC Delco brand name will have the same specifications as Quantum's Costar® model P-1 carbon monoxide alarm for vehicles, which was announced last November and is being test marketed through retail outlets (such as Longs Drugs) in limited regions of the U.S.

The Costar P-1 is one of several Quantum 9-volt battery powered CO alarms that incorporate the company's solid-state infrared reservoir (SIR) sensor system. The SIR system provides high selectivity to CO, rapidly regenerates to its original state when CO is eliminated from the environment, has a six year warranty, is reliable and versatile (can be powered by a 9V battery, 12-24 DC, and 120 AC), and has been designed to meet the revised Underwriters Laboratories (UL) 2034 standard (effective October 1, 1998) for single and multiple station CO detectors and the requirements of International Approval Services (IAS) currently in effect. Other battery powered Quantum CO detectors that incorporate the SIR sensor technology include the Costar model 9RV for recreational vehicles and Costar 9SIR residential CO alarm.

The Costar P-1 personal carbon monoxide alarm for vehicles, described as the first CO detector for automobiles, is capable of operating from -40°F to +151°F (-40°C to +66°C) and within a relative humidity range of 15% to 95%. The product, which can be mounted in the vehicle's visor or in a pocket or belt, has a 5 year limited warranty. The P-1 has an audible alarm of 65 dB at 10' (3.3 meters).

The Costar P-1 personal CO detector (as well as Quantum's 9RV and 9SIR CO alarms) is designed to be CO specific, eliminate nuisance alarms, and not activate an alarm unless there is a sufficient quantity of CO present. The P-1, 9RV, and 9SIR will not activate an alarm when CO levels are 30 ppm (parts per million). They will activate an alarm when the following dangerous concentrations of CO are present: 70 ppm for 60-120 minutes; 150 ppm for 10-15 minutes; and 400 ppm for 4-15 minutes.

Under normal conditions, the P-1, 9RV, and 9SIR models will flash a red LED every 30 seconds, indicating that the alarm is powered. If they detect harmful levels of CO, the LED turns on for two seconds and off for three seconds, with four short beeps for one second and four seconds of silence. In the event of a fault, the LED double flashes and a beep is emitted every 30 seconds. When the battery is low, one chirp occurs every 30 seconds. In the intelligent self-test using the microprocessor, there is one chirp, then the LED flashes 4 or 5 times, followed by two alarm signals.

Angelo Eftimeo, Quantum's national sales manager, noted that the AC Delco personal alarm for vehicles has very high market potential. Once the product is widely distributed through mass merchandisers and auto retailers, it is anticipated to generate sales of about one million units annually. He underscored that Quantum's CO detector for vehicles is the only feasible CO detector for this application, since it was designed specifically to meet the requirements of the automotive marketplace.

While the 9SIR residential CO detector, which retails for under $40, and the 9RV recreational vehicle CO detector, which retails for around $54, have received UL2034 (Oct. 1, 1998) listing, the P-1 has been submitted to UL, and Quantum and GM have the goal of receiving UL approval for the vehicle CO alarm. However, Eftimeo explained that UL approval of the CO detector is not a significant factor in the automotive after-market.

Eftimeo added that Quantum's personal CO monitoring detector technology has opportunities in industrial and worker safety applications. A slightly modified Costar P-1 can benefit, for example, fork lift operators and individuals operating machinery that can emit harmful levels of CO. Quantum's portable CO detector has potential for creating new and expanded applications in the occupational safety and health area. Since Quantum's CO detector does not respond to hydrogen, it is not subject to false alarms in, for example, HVAC applications.

The SIR CO detection technology is described as the first solid-state infrared CO sensor system that is able to respond over the relative humidity (RH) and temperatures found in vehicles.

The core of the solid-state infrared CO sensor system consists of two sensing elements, which are enclosed within a plastic housing. The sensor subassembly is mounted on the PCB block, which contains the optical monitoring components (i.e., IR light emitting diode (LED) and photodiode). A reservoir system, on the top of the PCB block, contains a proprietary chemical system (including a chemical pellet) that controls the relative humidity and air quality within the sensor's micro environment. The SIR sensor is designed to prevent fogging, due to rapid change in temperature or humidity.

The sensing elements change their optical properties in response to CO. The LED emits photons at 940 nm, which pass through the sensing elements in the absence of CO, and the photodiode monitors the change in photon transmittance. When exposed to CO, the sensor absorbs photons in proportion to the concentration of CO in the surrounding environment. The circuit pulses the LED periodically to measure the extent of photon absorption and rate of change of the sensors.

In the alarm, a microprocessor analyzes the change in voltage from the photodetector; and an alarm can be set off at virtually any programmed time. The intelligent chip tests the CO alarm every 10 minutes, and provides a signal that indicates when the battery needs to be replaced.

The SIR sensor system, introduced into the marketplace around autumn 1998, uses Quantum's S34 biomimetic CO sensor, which has two patented sensing elements (one red, one yellow). Each sensing element has a nanotechnology substrate composed of 99.9% silicon dioxide; and the sensing elements are made of a porous transparent disk coated with a supramolecular organometallic complex.

The complex is formed by a self-assembly process that generates a sensing element that mimics hemoglobin. By monitoring the rate of change in the amount of light transmitted through the sensing elements, the concentration of CO in the surrounding environment can be determined. The CO response characteristics of a particular sensor can be designed to address specific standards or applications by controlling the formulation process.

The SIR sensors--which are priced at about $4-$5, depending on quantity-- are used in residential CO detectors that are provided by such companies as First Alert, North American Detectors, Universal Security Instruments, and Maple Chase. Such sensors are also used, for example, in residential security systems provided by ADI.

Founded in 1982 by Mark Goldstein, Ph.D., who serves as the company's president and CEO, Quantum's CO sensors include a "synthetic" hemoglobin, the oxygen-carrying substance in red blood cells that binds to carbon monoxide. The sensors darken proportionally to the concentration of CO in the environment. Quantum, which began manufacturing and selling CO safety products in 1989, is purportedly the first company to commercialize a biotechnology-based gas sensor.

When inhaled, CO combines with hemoglobin, which blocks the hemoglobin from picking up oxygen from the lungs. A lack of oxygen causes body cells and tissues to die. Once in the bloodstream, CO is released very slowly from the body.

Quantum's sensors contain molecules that mimic the binding and release of carbon monoxide. Some of the CO is converted to CO2 (carbon dioxide) and some of it is released into the air in the same manner as in the human body. Since Quantum's sensor directly monitors the toxicity of ambient CO, they can protect consumers from long-term, low-level, and peak high-level CO exposure. The "artificial hemoglobin" at the core of Quantum's sensor is fabricated with the help of genetically engineered organisms. Such organisms produce molecular structures that resemble a "keyhole;" and CO is the key that fits the hole. An LED and photodiode detector are used to measure the transmittance of light through the keyhole to detect the presence of CO.

According to the Centers for Disease Control, CO is the leading cause of accidental poisoning in the U.S. The U.S. Environmental Protection Agency reported in 1991 that close to one million people are affected by CO poisoning annually. According to a study published in the Journal of The American Medical Association (1991), about 5,600 fatalities per year can be attributed to CO. This study, moreover, stated that 57% of the unintentional deaths caused by CO were from motor vehicle exhaust. In addition, alcohol and CO are supra-additive.

Sources of carbon monoxide in the home include clogged chimneys; wood-burning or gas-powered fireplaces; gas logs or burners; portable heaters; gas stoves and range tops; grills, barbecues, or hibachis used indoors; faulty water heaters, furnaces, and gas clothes dryers; and airtight, energy-efficient homes (which reduce the fresh air supply leading to less efficient combustion and a greater CO risk). An energy efficient home, with insulation to reduce heat loss, can produce a negative pressure build-up, creating a backdrafting effect that would draw fumes into the home instead of exhausting them to the exterior.

Moreover, the National Highway Traffic Administration (NHTSA) has advised that most accidental deaths associated with automobiles occur when cars are left idling in areas without adequate ventilation, such as garages. Holes in mufflers or exhaust pipes can cause CO to enter the vehicle and rise to hazardous levels even outdoors.

Quantum notes that multiple gar garages, such as those found in apartment houses and condos, are especially dangerous; and a commercial CO detector that activates ventilation controls is beneficial for such structures. Furthermore, drivers in vehicles sitting in congested traffic can be exposed to CO; and even low levels CO have impaired drivers' reaction times and psychomotor coordination skills. In Australia, legislation has reportedly been proposed to have a CO detector installed in the vehicle.

In tests performed at Lawrence Berkeley National Laboratory (LBNL) under the sponsorship of Quantum, which compared the performance of CO detectors that use biomimetic, electrochemical cell, and metal oxide (tin oxide) semiconductor CO sensor technologies, the biomimetic detectors outperformed their electrochemical (EC) and metal oxide semiconductor (MOS) counterparts with respect to interferent resistance and CO selectivity.

The research was conducted to independently investigate differences among the sensor technologies and to explore the effects that a variety of common household vapors have on CO detectors. The results of the tests were reported in July 1998.

In the interferent resistance tests, the detectors were exposed to the sequence of interferent gases to ascertain the detector's resistance, or lack of response, to gases other than CO. In the selectivity tests, the same detectors and interference gases were again tested; however, CO in concentrations of 100 ppm was introduced.

A biomimetic model was the only detector completely unaffected by all 13 intereference gases. It did not exhibit any false alarms, premature responses, or failures. The interference gases were methane (500 ppm), heptane (500 ppm), butane (300 ppm), ethylene (200 ppm), toluene (200 ppm), trichloroethylene (200 ppm), ammonia (100 ppm), nitrous oxide (200 ppm), carbon dioxide (1000 ppm), ethanol (200 ppm), isopropanol (200 ppm), acetone (200 ppm), and ethyl acetate (200 ppm).

A total of 9 false alarms to ethylene in 30 tests were observed; however, the biomimetic models did not exhibit any false alarms. Detectors with MOS sensors false alarmed in nearly half the tests in which ethylene was present. An EC model continuously failed CO selectivity tests; and this model failed to respond to CO alone following exposure to interferents. In the selectivity tests, the biomimetic detectors did not exhibit any failures (they responded to 100 ppm of CO), and such detectors exhibited the fewest number of premature responses (2). In most cases, premature responses (those occurring in less than five minutes) were correlated with shortened recovery times.

In addition to RVs, residential, and vehicle use, Quantum's CO detectors are used for such applications as appliance safety shut-off and ventilation fan control.

ASHRAE (American Society of Heating Refrigeration and Air Conditioning Engineers) has drafted a residential ventilation standard (62.2) that calls for a CO detector somewhere in the home. The standard will soon be available for public review. While ASHRAE standards do not have the force of law, they can serve as a standard of care for professionals designing ventilation systems, and such standards can be adopted by regulatory bodies.

Moreover, the credit card size Quantum Eye® multi-level CO detector, with a minimum 18-month operational life, monitors indoor air quality and indicates whether the surrounding air is safe from CO via a color change. This device is finding expanding opportunities in such areas as personal aircraft. The Quantum Smog Detector® is a passive, self-regenerating monitor that visually indicates smog (ozone). Darkening of the strip depends on the smog (ozone) level at the time of exposure. If air quality is satisfactory, the strip retains its normal color. Graying of the strip after one hour indicates moderate air pollution; and further darkening within an hour or less indicates a severe air pollution. The detector regenerates in clean air overnight.

Presently, a major market for Quantum's CO detectors is RVs. The company's 9RV detector is widely installed in RVs that are manufactured by such companies as Winnebago, Fleetwood, and Holiday Rambler. The ANSI standard requires a CO detector in an RV with a power generator or with a generator prep. The Recreational Vehicle Industry Association (RVIA, Reston, VA), which reportedly represents about 95% of RV's produced in the U.S., has adopted the ANSI standard. Every member of RVIA is committed to complying with the ANSI standard.

Dr. Goldstein told SBD that, on average, Quantum's sales have been growing about 30% annually over the past five years. In 1999, over 50% of Quantum's sales were accounted for by CO sensors as opposed to CO detectors. By 2001, the CO sensor business should account for less than 50% of Quantum's overall sales. Dr. Goldstein noted that key new markets with strong growth opportunities for Quantum's CO detectors include vehicles, medical (e.g., breath diagnostics for detecting a disease that causes jaundice in newborns, detecting CO in anesthesia gas monitoring), and commercial (e.g., ventilation control in parking lots and warehouses) applications.

SBD pegs revenues for the North American market for CO sensors used in residential CO detectors at about $33.75 million, representing 6.75 million units.

Heraeus Sensor-Nite Catalyzes Exhaust Gas Temperature Sensing

As requirements for vehicle exhaust emissions become increasingly stringent, exhaust gas temperature sensors (EGTS) are finding expanding opportunities in engine development/diagnostics, as well as in on-board automotive catalyst monitoring/diagnostics and engine management.

The use of exhaust gas temperature sensors for detecting and monitoring catalyst light off performance (the period before the catalyst is operational) can allow for enhanced monitoring and reduction of HC (hydrocarbon) emissions during cold start, as a result of monitoring the catalyst's performance during this vital phase. Moreover, the use of a catalyst temperature sensor to evaluate the light-off characteristics of the catalyst can facilitate NOx (oxides of nitrogen) efficiency monitoring.

The prevailing dual oxygen sensor approach for on-board catalyst monitoring/diagnostics--which indicates the HC conversion efficiency of a warmed-up catalyst by determining its oxygen storage capacity-- reportedly has limitations for adequately evaluating the catalyst's performance during cold start, when a significant portion of emissions for low-emission vehicles occur.

Furthermore, the California Air Resources Board (CARB) has noted that, in OBD II demonstration vehicles, NOx emissions can be high when HC emissions reach the malfunction threshold. CARB has proposed, with respect to its OBD II regulations, that, beginning in MY 2004 on LEV II applications, the catalyst system shall be considered malfunctioning when its conversion capability decreases to the point that either: HC or NOx emissions exceed 1.75 times the applicable FTP (Federal Test Procedure) standard; or the average FTP Non-Methane Hydrocarbon (NMHC) or NOx conversion efficiency of the monitored portion of the catalyst falls below 50%.

CARB has also noted that a significant portion of the weighted emissions of an FTP test for low-emission vehicles is generated during the initial two minutes of the cold start portion of the test. They are concerned about the performance and monitoring of the components/strategies used to reduce cold start emissions or accelerate catalyst light off; and they believe that confirmation of the accelerated or warm-up conditions is necessary. CARB has recognized that some manufacturers may incorporate exhaust or catalyst temperature sensors into their emission control systems; and a manufacturer may develop a diagnostic strategy using such sensors to verify that the conditions to accelerate catalyst light off were achieved or catalyst light off temperature was reached within a specified time period.

CARB will have a proposal to monitor cold start strategies, for implementation on applications certified to LEV II emission standards, beginning in MY 2004. SULEVs receiving a partial ZEV allowance shall comply with such requirements, likely starting in MY 2003. The details of the input and output components to be monitored are being discussed.

CARB's maximum exhaust emission standards for new 2004 and subsequent model year passenger car and light duty truck SULEVs, at or below 120,000 miles, are: NMOG (Non-Methane Organic Gas)--0.010 g/mi; CO (carbon monoxide)--1.0 g/mi; NOx--0.02 g/mi; formaldehyde--4 mg/mi; and particulate from diesel vehicles--0.01 g/mi.

In 2004, CARB's fleet average NMOG requirements for passenger cars produced and delivered for sale in California by a manufacturer other than a small manufacturer is 0.053 g/mi. The fleet average NMOG requirement for light-duty trucks in 2004 is 0.085 g/mi.

In addition to catalyst diagnostics and the detection of catalyst light off characteristics, the EGTS catalyst monitoring sensor can be used as an input for various engine management functions, such as closed-loop catalyst temperature control, catalyst over-temperature protection, secondary air injection, air/fuel control, fuel enrichment, ignition timing, idle bypass air flow, oxygen sensor heater control, regeneration control and over-temperature protection for lean NOx traps in gasoline direct injection engines, and control of injection of the reducing agent in active lean NOx catalysts in diesel engines.

Heraeus Sensor-Nite (Amsterdam, The Netherlands, +), whose U.S. headquarters is located in Fenton, MI ), has been at the forefront of vehicle exhaust gas temperature sensing. Their ECO (Energy Control and Optimization) TS-200 thin-film platinum RTD exhaust gas temperature sensors--which ensure that optimal temperature conditions are maintained in NOx absorber systems and provide vital data for on-board diagnostics of the three-way catalytic converter--have been finding increasing opportunities for use in production vehicles.

Joseph R. Griffin, vice president-automotive at Heraeus Sensor-Nite in Fenton, MI, notes that closed-loop catalyst temperature control--where a single temperature sensor provides closed-loop catalyst light off control, as well as OBD catalyst monitoring--allows for reducing HC, CO, and NOx tailpipe emissions and for enhanced, more robust catalyst monitoring. The system, which is especially applicable for ULEV systems with small-volume lightoff catalysts, allows for catalyst monitoring that is not restricted by driving cycle, monitored catalyst volume, catalyst washcoat technology, air/fuel ratio control strategy, or fuel sulfur poisoning. Moreover, the closed-loop catalyst temperature feedback monitoring method focuses on the major cause of increased HC emissions; i.e., degraded catalyst lightoff performance.

The three primary parameters that must be controlled in a three-way catalyst to achieve optimal performance are air, fuel (A/F ratio), and energy (temperature). Presently, feedback from an oxygen sensor provides closed-loop control of the A/F ratio; however, there is no feedback control for the energy (temperature) of the catalyst. Such energy includes a combination of exhaust gas energy (wasted heat from combustion) and exotherm energy generated in the catalyst.

Measuring the rate of change of the catalyst thermal mass temperature offers a direct measurement of the combined energy (exhaust gas plus catalyst exotherm) in the catalyst, while taking into account exhaust feedgas temperature, mass flow, and any stored energy in the catalyst from earlier operation. Catalyst temperature measurement provides a feedback signal during the cold start period to control the quantity and timing of energy and excess oxygen supplied to the catalyst, and also serves as a catalyst diagnostic.

Closed-loop catalyst light off control, using direct temperature measurement for feedback, produces the fastest possible catalyst light off and optimizes catalyst heating in a manner that is nearly independent of the driving cycle from idle conditions to moderate accelerations. Moreover, the closed-loop catalyst temperature feedback signal provides a vital input parameter to control the start of closed-loop A/F (air to fuel ratio) control, or the start and end of secondary air injection if required.

This ECO strategy controls ignition timing and idle bypass air flow at a programmedloop A/F ratio during the catalyst light off period to produce a constant rate of energy for heating the catalyst that is virtually independent of the driving cycle from idle to moderate accelerations. In this energy control and optimization strategy, the combination of a rich A/F ratio, increased spark retard, and increased idle bypass airflow allows a high catalyst heating rate at idle without increased idle RPM.

Furthermore, the closed-loop catalyst temperature control system is compatible with a small volume (approximately .25L to 0.3L), low oxygen storage capacity (OSC) light off catalyst, which Griffin has noted is attractive for achieving low-emission vehicle requirements. The dual oxygen sensor approach for catalyst monitoring, he noted, is not readily compatible with close-coupled, small-volume light off catalysts, since the dual oxygen method requires the monitored catalyst volume to be larger than 0.3L for many engine/vehicle configurations. Moreover, the OSC component in a close-coupled light off catalyst may have to be minimized to avoid excessive exotherms during high speed and load driving conditions, which could negatively impact the dual oxygen sensor catalyst monitoring method.

In addition, an enhanced Heraeus Sensor-Nite's exhaust gas temperature sensor (with a response time in the range of 800 ms) makes possible a method for indirectly measuring differences in A/F during the initial cold start phase of engine operation, affording it opportunities for use in detecting the cold start volatility characteristic of the fuel. During initial cold start, high driveability index (DI) (i.e., low volatility) fuel causes theloop A/F ratio to shift lean, which contributes to unstable combustion. To compensate for the lean shift and allow acceptable driveability with high DI fuels, the commandedloop A/F ratio must be richer than necessary. Such excessive enrichment leads to increased cold start HC emissions, particularly when the vehicle operates on higher volatility fuels.

DI fuel detection via exhaust gas temperature measurement exploits a primary effect of combustion during cold start: the relationship between exhaust gas temperature and the A/F ratio. Using the rapid-response exhaust gas temperature sensor, it is possible to detect high DI fuel via measured exhaust gas temperature during the first three seconds after a cold start. During the initial two or three seconds after cold start, once the maximum exhaust gas temperature is reached, the DI of the fuel can be calculated. The presence of high DI fuel can be calculated by the deviation from a predicted exhaust gas temperature during cold start (which is derived from the temperature characteristics of low DI fuel, engine speed, load, coolant temperature, air charge temperature, and throttle position).

With the ability to estimate the fuel DI during cold start, the vehicle manufacturer is able to optimally calibrate theloop fueling to meet driveability and HC emission requirements. To meet stringent emission requirements, some vehicle manufacturers do not fully compensate for the use of very high DI fuels. If fueled with high DI fuel, such vehicles can experience engine misfire and incomplete combustion during cold start, leading to increased tailpipe HC emissions. Detection of fuel DI in such vehicles would allow for enhanced fueling, resulting in more complete combustion and reduced HC emissions.

Beginning in May of this year, Heraeus Sensor-Nite will produce in Europe exhaust gas temperature sensors for gasoline direct injection engines. The sensors will help provide on-board diagnostics of the three-way catalyst, regeneration control for lean NOx traps (absorbers), and over-temperature protection for lean NOx traps.

By 2002, Heraeus Sensor-Nite's exhaust gas temperature sensors will be used in the U.S. in light-duty diesel engine active lean NOx control. The sensors will measure catalyst operating temperature and provide feedback with respect to when the catalyst is in the targeted temperature window (approximately 180-280°C) for controlling injection of the reducing agent.

Heraeus Sensor-Nite's exhaust gas temperature sensors are supplied as RTD assemblies to automakers. In high volume, such sensors will be priced below $15.

Griffin believes that, by 2004, exhaust gas temperature sensors will be required on SULEV (super ultra low emission vehicle) applications in the U.S. for catalyst cold start heating control/diagnostics. He noted that the dual oxygen sensor method will be insufficient for diagnosing catalyst performance at SULEV emission levels.

Moreover, Heraeus Sensor-Nite's new ECO-MCT08.10 exhaust gas temperature analysis system is a rapid-response, robust plug-and-play development tool that provides an engine designer, calibration engineer, or dynamometer laboratory with real-time exhaust gas temperature measurement for up to eight cylinders on any engine.

Key features of the ECO-MCT08.10 system include: eight ECO-TS-200 thin-film platinum resistance temperature sensors, which have a temperature range of -40 to +1000°C; an electronic enhancement system allowing for a response time of up to 800 ms; displays of maximum temperature and delta temperature, with store and hold function; eight 0-5V linear analog outputs for linking each channel to a data acquisition system; programmable over-temperature alarm with indicator lights and a voltage output for protection strategy initiation; real-time data acquisition via an RS-232 serial port and the included software program; 12V DC power; and mounting hardware (e.g., stainless steel fittings with caps, wiring harness (interface cable), power cord). The complete system is priced at $1,750.

Target markets for the ECO-MCTO8.10 include OEM engine development and aftermarket applications (such as racing). In engine dynamometer development, the system is designed to replace type K thermocouples used for measuring exhaust gas temperature. The ECO-TS-200 temperature sensor used in the ECO-MCT08.10 has a significantly faster response to temperature change than a 1/8" thermocouple, and it has a longer life (of 10 years or 150,000 miles) than the type K thermocouple. The ECO-TS-200 also detects a higher steady-state temperature than the thermocouple. Measurement error in the ECO-TS-200 is minimized, as a result of minimizing heat transfer from the sensor element to the sensor body (i.e., exhaust pipe).

Heat transfer issues can be a major source of measurement error at exhaust system temperatures above 600°C. Thermally de-coupling the sensing element from the sensor body is required to minimize conduction heat loss (e.g., from exhaust pipe walls), which induces measurement error. Thehousing design and small-mass sensing element of the exhaust gas temperature sensor used in the ECO-MCT08.10 system optimize convection heat transfer for enhanced performance in dynamic conditions.

Griffin noted that the ECO-MCT08.10 exhaust gas temperature analysis and display system is highly adaptable to the user's data acquisition system or engine design system, and it can be used as a development tool by any vehicle manufacturer.

In the OEM engine development area, applications for the ECO-MCT08.10 include individual cylinder performance monitoring, air/fuel distribution analysis, and air/fuel ratio calculation. The system, moreover, provides over-temperature protection of the engine. In catalyst aging, the system performs diagnostics to ensure that the engine performs as required in order to age the catalyst appropriately. The system provides feedback on engine performance, and it provides an early warning so that the test can be stopped before a catastrophic engine failure occurs, which would impact the catalyst aging process.

Racing applications for the ECO-MCT08.10 include system monitoring for adjustment to changing conditions, over-temperature detection, and over-temperature protection strategy initiation (i.e., drive warning light, solenoid, or relay). Moreover, the system can be used to determine if there is a balanced air/fuel ratio among the cylinders by having an exhaust gas temperature sensor in each cylinder exhaust runner.

Key markets/applications for Heraeus Sensor-Nite's thin-film platinum RTDs (PRTDs) include automotive, energy management (e.g., calorimetry, HVAC), medicine and laboratory, and household appliances/white goods (e.g., ovens), as well as electronic components, communication, and industrial plants. The company delivers more than one million sensors annually to major European heat meter producers.

Since PRTDs are able to detect minimum temperature differences, they are also used where information about other (temperature-related) parameters is required, such as measuring flow rates, tension, wear, level, and detecting leaks. Another key application for such sensors is temperature compensation of electronic components.

Heraeus Sensor-Nite--which was established in 1997 as an independent spin-off from Heraeus' sensor division--offers a variety of automotive sensor products, including the ECO family of gas sensors for measuring concentrations of hydrocarbons, NOx, and exhaust gas oxygen (lambda value); mass air flow sensors; oil temperature sensors; coolant water temperature sensors; and sensors for air conditioning (e.g., measuring external temperature, measuring temperature at different locations within the car, and measuring temperature and air humidity in the car's interior via a dew point sensor).

The direct injection gas engine appears to be favored as a means of reducing automobile carbon dioxide emissions, Heraeus Sensor-Nite notes. Their ECO-TS-200 high-temperature sensors ensure that optimal temperature conditions are maintained in NOx absorber systems and furnish data for on-board diagnostics of the three-way catalytic converter. The company's ECO NOx sensor, which uses thin- and thick-film technologies, allows for improved monitoring of NOx absorbers.

Since 1993, over one million automobile fuel injection engines have been equipped with Heraeus Sensor-Nite's heaters and temperature sensors for mass air flow meters. Emission of pollutants is reduced by optimizing the air-fuel mixture in the induction manifold and precise air metering in the exhaust gas recirculation system.

Heraeus Sensor-Nite's microbridge mass air flow sensor--which combines a micro heating element with a temperature sensor--allows for precisely measuring the required air mass. The microbridge mass flow sensor system--a membrane system based on microsystems technology--has a thickness of only a few micrometers and a response time of only a few milliseconds. The device's pressure pulse recognition capability allows the mixture to be precisely controlled across the entire range of engine speeds. The microbridge sensor's rapid measurement and pressure pulse recognition abilities drive improvements in exhaust gas recirculation, leading to a significant reduction in NOx emissions.

Heraeus notes that, in motor management systems, thin-film platinum RTDs are used to measure coolant water and oil temperatures to help optimize safety, cost, and environmental compatibility. Such temperature sensors facilitate determining oil quality. An on-board computer calculates an index, which signals when an oil change is due, taking into account individual driving habits. This scheme eliminates the need to adhere to rigid, predefined servicing dates and allows for needs-oriented maintenance, which is more economical and environmentally-friendly and reduces oil consumption.

Thin-film platinum resistance temperature sensors help safeguard against uneconomical and environmentally unsafe operating conditions in vehicles with a continuously variable transmission (CVT). While CVT systems help reduce fuel consumption and exhaust gas emissions, they are subject to increased friction when operating at high power levels. As a result, the temperature and viscosity of the oil rise, which leads to reduced transmission efficiency.

Low-temperature thin-film platinum RTDs provide data about external temperature and temperature at different locations within the car's interior to allow the most economical and environmentally-friendly settings to be selected for the air conditioning system.

Moreover, a dew point sensor, based on thin-film platinum technology, allows for measuring temperature and air humidity (a key parameter for controlling the micro-climate of the car's interior). The car's air conditioning system adjusts the measured temperature and humidity parameters to ensure that the windows remain clear and the air quality contributes to good driving. In auxiliary heating systems, measurement of exterior and interior temperatures and the air humidity in the car facilitates economically and ecologically optimized control and the avoidance of heat build-up.

Heraeus Sensor-Nite's thin-film platinum RTDs--which provide high long-term stability and reliability, and are aimed at markets/applications with a high degree of automation--consist of a sealed, photolithographically structured platinum thin-film and an Al2O3 substrate. The sensors are manufactured using techniques adapted from the semiconductor industry (e.g., vapor deposition, photolithography, laser trimming). The following models of PRTDs are offered: FK-series (unhoused with connection leads); FR-series (sealed into a ceramic pot); SMD-series (with axial soldering pads); and TO92 and SOT223-series (sealed in a plastic housing with rigid connection leads).

The plant in Kleinostheim, Germany (+) is responsible for manufacturing the basic platinum thin-film temperature sensor elements required by other business units within Heraeus Sensor-Nite; and Kleinostheim also supplies the group's production sites with other elementary temperature sensors. p> Heraeus Sensor-Nite's high-temperature and exhaust gas sensors for the automobile industry are manufactured in Leuven, Belgium. Their facility in Fontegnay-Tresigny, France manufactures glass wire-wound platinum RTDs used for measuring temperature in textile machinery and motors and thermocouples for use in the semiconductor industry. The plant in Sao Paulo, Brazil produces wire-wound ceramic platinum RTD elements, used in power stations and for process control in the chemical industry. The plant in Ellwood City, PA manufactures temperature sensors for domestic appliances (e.g., oven temperature probes) and platinum chip sensors for use in other consumer products.

In fiscal 1999, the Heraeus Group had sales of EUR 4.583 billion (about $4.17 billion at the current exchange rate). Heraeus Sensor-Nite is a business unit of Heraeus Electro-Nite (Houthalen, Belgium). The latter had sales of EUR 234 million (about $212.7 million) in 1999.

Worldwide demand for automotive temperature sensors totaled $221.5 million in 1998 and is projected to reach $259.8 million in 2003, representing a compound annual growth rate of about 3.2%, according to Strategy Analytics (Luton, Bedfordshire, England, +). Worldwide unit volume for automotive temperature sensors totaled 120.3 million units in 1998 and is expected to reach 173.2 million units in 2003, representing a compound annual growth rate of about 7.6%. The data includes HVAC, transmission, engine coolant, engine air, and engine fuel applications.

Worldwide demand for automotive oxygen sensors totaled 71.8 million units in 1998 and is projected to reach 90.8 million units in 2003, representing a compound annual growth rate of about 7.6%.

Civilian, global demand for automotive sensors totaled about $6.4 billion in 1998, according to Intechno Consulting's (Basel, Switzerland, +) Sensor Markets 2008 report.

According to SBD, the North American market for RTDs totaled about $173.2 million in 1999.

Sensor Business Digest Index | Excerpts from Sensor Business Digest | Order Form | Home | This site is best viewed with version 4.0 or later of either Netscape or Internet Explorer. Click the appropriate button below to update your browser