Developing Next-Generation Human Interfaces Using Capacitive and Infrared Proximity Sensing

By Steve Gerber, Silicon Laboratories, Inc.

“High tech/high touch” took on new meaning with the advent of touchscreens. Adding IR proximity sensing takes it to the next level.

Next-generation human interfaces, based on capacitive and infrared proximity sensing technology, can provide dramatic improvements in the end-user experience while increasing system reliability and reducing total cost. In addition to making electronic products easier to use and more visually appealing, these interfaces mask the increasing underlying complexity of electronics, enabling manufacturers to bring products with advanced capabilities more quickly to the mass market.

Advanced sensor-based interfaces are more reliable than traditional mechanical interfaces since they do not contain the moving parts associated with buttons and dials, which are prone to failure over time. Sensor-based control panels and displays are also more flexible, allowing a single set of controls to be reconfigured based on application context so that users are presented with only those choices that are currently active. When combined with gesture recognition and “touchless” technology, developers can add intelligence to device interfaces that anticipates user needs and enables innovative usage models to make products friendlier and more intuitive to use. Flexible firmware can be adjusted quickly and easily as market needs change without having to completely re-architect systems or redesign device packaging.

Next-generation human interfaces

Next-generation products require next-generation human interfaces to differentiate themselves in the marketplace. By making electronic devices more aware of their environment, new features can be implemented which enhance ease of use, improve power efficiency, and reduce system cost. In addition, high sensitivity, low noise, and resistance to moisture ensure reliability, even in the most challenging environments.

Two of the primary technologies driving next-generation interface development are capacitive and proximity sensing. Capacitive sensing detects the presence of a human finger through changes in the capacitive value of the sensing element. It enables advanced controls such as sliders and wheels and is well-suited for close-range interfaces where users are used to physical feedback, such as when they press a button. Proximity sensing uses infrared sensors to calculate the distance of objects up to one meter away using infrared reflectivity techniques. Proximity sensors also can place an object in space, enabling “touchless” gesture tracking.

Using these two technologies together enables finer tuning of user interfaces. Many end users are already familiar with capacitive sensing technology from its use in a wide array of consumer products, most prominently the iPod and iPhone. Until recently, proximity sensing has typically been used for simple tasks such as cheek detection in handsets. However, it can be used for so much more:

  • User detection: Proximity sensing can determine, for example, whether an end user is currently sitting at a PC and can turn off the display when the user walks away from the desk. Given the substantial power required for LCD backlighting, even simple user detection can result in significant power savings across an organization. User detection also can be implemented in devices such as USB chargers or thumb drives so that devices can prepare themselves for sudden removal.
  • Fingerprint-free displays: Many portable devices require users to touch buttons all over the screen, leaving oily marks that are difficult to see through and clean. A touchless interface for a portable multimedia player, for example, would eliminate the need for users to touch the very screen on which they want to watch a video. Similar applications include enabling users to flip through the pages of an electronic cookbook with touchless ease, or allowing doctors to directly interact with touch screen-based systems during surgery without fouling the screen with fingerprints.
  • Automatic backlighting control: Part of the proximity sensing signal path is the use of an ambient light sensor (ALS) to reduce noise from external light sources. This same sensor also can be used to monitor background lighting conditions and automatically adjust the display backlight appropriately to reduce power consumption.
  • Invisible intrusion detection: Reflecting infrared light off the interior door surface of a system allows developers to deploy an “invisible” intrusion mechanism which avoids the unreliability and expense of mechanical switches used for the same purpose.
  • Health and safety benefits: Kiosks, check-out stands, and other public computers present health risks in terms of spreading disease via keyboards or touch screens. In parts of China, for example, laws require that every elevator panel be wiped down once an hour to prevent the spread of SARS. Touchless panels avoid and mitigate these public health issues.

Offloading interface control

One emerging trend in embedded design is to offload user interface management from the primary application processor to a dedicated, 8-bit MCU. Human touch is a relatively slow event to an application processor, and powering the entire system to check if a user has moved his or her finger consumes significantly more power than is required by an 8-bit MCU to accomplish the same task.

Capacitive touch-sense MCUs, such as the F99x family from Silicon Labs, are ideally suited for managing next-generation user interfaces. By providing up to 25 MHz of performance combined with peripherals optimized for the task, F99x MCUs provide the processing and input capabilities necessary to implement intelligent and accurate sensing. When implemented in conjunction with Silicon Labs’ Si114x family of proximity sensors, developers can implement highly efficient user interfaces within a single development environment.

Capacitive sensing performance for the F99x MCUs is further enhanced through a hardware-based, capacitive-to-digital converter (CDC). Developed by Silicon Labs, the CDC includes two current inputs which are digital-to-analog converters (DACs). The first is a variable DAC that measures the current to the external sensor capacitor, and the second is a constant current source for an internal reference capacitor (see Figure 1). Capacitance is measured using successive approximation registers (SAR) – an efficient process immune to DC offset which requires no external components.

Figure 1: The hardware-based CDC enables high performance, 16-bit resolution, high reliability and DC offset immunity − without requiring external components.

The F99x MCU’s 16-bit CDC offers high reliability and accuracy. By performing a two-stage discharge of the external capacitor, the CDC can remove ambient noise energy captured during the discharge process. In comparison, other approaches require additional external components (e.g., series resistors) and more than one I/O per channel (thereby increasing MCU size and increasing routing difficulty).

The CDC’s dynamic range is improved through the use of an adjustable gain. The dynamic range is also enhanced by the ability to reduce the source current to change the charge timing and more directly reflect the voltage at the capacitive sensor when the source current and series impedance are both high (such as when using a touch panel or ESD-protected capacitive pads). Higher sensitivity gives developers greater signal margin, allowing them to use thicker plastics, make electrodes smaller, and operate reliably in noisy environments. The CDC also employs pin monitoring to automatically adjust conversion timing if necessary to eliminate possible interference from high-current switching on nearby pins. In summary, the CDC enables a superior signal-to-noise ratio (SNR) of between 50 to 100 for a typical capacitive sensing implementation.

Unmatched system responsiveness

Proximity sensing employs an infrared sensor and one or more infrared light-emitting diodes (LEDs). The basic operating principle is to illuminate an object and measure the intensity of the reflected light. The number of LEDs required depends upon the application and what spatial information is needed. A paper towel dispenser sensor, for example, needs only a single LED to detect that a person is standing in front of the dispenser. To detect a left/right or up/down gesture, two LEDs are required. To support full spatial navigation, three LEDs are needed. In each case, only a single physical sensor is required. However, each additional sensor increases the processing required to identify the intensity of signals received from each LED and to triangulate the detected object’s position.

Processing is also required to filter noise (i.e., background light) from the received signal. The more powerful the processor or embedded controller, the more samples it can take and the better filtering it can achieve. Increasing the sample rate increases the resolution of the system, while better filtering increases accuracy. Rapid sampling and precise filtering are both required for a robust interface, and developers must balance each approach to optimize their application.

Typically, the extended acquisition times associated with low sensitivity photodiodes allow flicker from sources such as fluorescent lights to reduce accuracy. Silicon Labs’ high-sensitivity photodiode technology – proven in the industry for more than ten years – exhibits excellent immunity to EMI and flicker while reliably detecting objects up to 50 cm away without the use of external lenses or filters. Built upon this robust photodiode technology, the Si114x sensor family supports ambient light sensing capabilities.

The primary power drain in a proximity sensing subsystem is the infrared LED. Silicon Labs’ QuickSense development environment assists developers in defining configuration parameters to optimize accuracy, detection range, and power consumption. For example, advanced control capabilities allow developers to dynamically adjust LED current to tune it for a particular application and detection range. Developers also can control the refresh strobe rate to further minimize LED power consumption. For ultra-low-power operation, developers can use innovative single-pulse proximity sensing to minimize LED on-time and achieve up to a 4000 times improvement in power consumption, as shown in Figure 2.

Reducing system power consumption

With the current emphasis on green, energy-saving electronics, all devices, not just portable devices, are beginning to be designed with power conservation in mind. Part of an effective low-power strategy is to minimize CPU active time while maximizing sleep time for as many components as possible within the system. Silicon Labs has implemented several mechanisms for reducing overall system power consumption in its capacitive touch-sense MCUs:

  • Background scanning: Since the CDC is implemented in hardware, capacitance measurements for channel scanning can be completed autonomously while the CPU operates in its powersaving suspend mode.

Figure 2: QuickSense MCUs offer innovative single-pulse proximity sensing to minimize LED on-time and reduce power consumption by up to 4000 times.

  • Autonomous auto scanning: Rather than scan and convert all capacitive sensing channels, only active channels are scanned and converted.
  • Channel bonding: Scanning several channels together using a single input consumes less power than handling multiple conversions to check channels individually. For example, the system can scan an entire slider using a single input and wake the CPU if any active channel is touched. Once awake, the CPU can then scan each channel separately to determine which channels were touched and begin interpretation of pending gestures.
  • Integrated LDO regulator: The F99x MCU’s integrated low drop out (LDO) voltage regulator provides linear response while maintaining a constant, ultra-low active current at all voltages. In addition, the F99x has special circuitry to retain RAM when the LDO regulator is disabled in sleep mode.
  • Flexible operating voltage: For many MCUs, the CPU must be operated at a lower frequency as the operating voltage is reduced, thus increasing operating time and power consumption. MCUs limited to 2.2 V operation also waste 20 percent battery life when using AA/AAA batteries. With full operating capabilities at 25 MHz down to 1.8 V, the F99x maximizes battery powered efficiency across applications.
  • Most MCUs are designed to optimize either active or sleep power efficiency. The F99x architecture was designed from the ground up to offer the industry’s lowest power in both active mode and sleep mode (see Table 1). An internal power management unit (PMU) limits leakage, resulting in active and sleep currents less than half that of the F99x’s closest competitors.

    Mode Current Consumption
    Active 150 μA/MHz*
    Sleep with brownout detection disabled 10 nA
    Sleep with brownout detection enabled 50 nA
    Sleep with internal RTC operational 300 nA

    Table 1: F99x active and sleep mode power consumption.

    *When operating at 0.9 to 1.8 V, C8051F99x MCUs achieve even greater power efficiency through the use of an internal boost converter.

    Fast wake-on-touch

    An important technique for reducing power consumption is to turn off a device’s display and control interfaces when they are not in use and then place the entire system into sleep mode. A key element of interface design is how responsive the system is to a user when it is in transition between sleep and active modes, (i.e. how fast it can wake up). With a capacitive sensing-based system, when the system is asleep there is no backlighting to indicate to the user which function each capacitive button or slider currently represents. Thus, the first button press is limited to simply waking up the system.

    When proximity sensing is available, a system can detect a user from up to one meter away. This capability allows the proximity sensor to wake the system as the user approaches or reaches for the device so the system can wake and the display can be fully functional by the time the user is ready to press a button. In practical application, this changes how users can interact with devices by making systems more intelligent and friendly. For example, devices such as car stereos or set-top boxes can have black panels that “disappear” from sight when not in use but turn on with full awareness when a user’s hand is near.

    Wake-up time is measured as the time interval between recognizing the need to wake and when the first instruction is executed. Wake-up time depends upon many factors, including regulator stabilization and analog settling time. For events such as reading a capacitive or proximity sensor, among the first actions the CPU must take is an analog measurement. If the analog peripheral is not ready to be read, this increases the effective wake-up time. Wake-up time not only defines system responsiveness, it also affects power efficiency. During wake-up, the MCU performs no work but still consumes power. Shorter wake-up time, therefore, reduces the power wasted as the CPU wakes up.

    To complicate evaluating wake-up time, vendors measure wakeup time using different criteria. Some MCUs wake to trigger an interrupt service routine (ISR) and must wait before taking an analog measurement. Wake-up time, in these cases, is measured from the wake event to either an MCLK valid on the appropriate pin or when the interrupt vector is fetched. To obtain a wake-up time, equivalent to measurements made until the execution of the first code instruction, developers must add several μs/CPU cycles to the measurement.

    The wake-up time of the F99x MCU has been speed optimized, resulting in a wake-up time from sleep of 2 μs. In addition, its analog settling time is only 1.7 μs, 15 times faster than competitive MCUs. Thus, the effective wake-up time from event to first analog measurement is less than 4 μs, up to seven times faster than the closest competitor.

    In addition to fast responsiveness, F99x MCUs have the industry’s lowest power capacitive touch sensing available on the market today. They provide outstanding performance – 150 μA/MHz – over the entire 1.8 to 3.6 V operating range, as well as the industry’s lowest power wake-on-touch at less than 1 μA. Fourteen CDC channels provide ultra-fast 40 μs acquisition time, 16-bit accuracy, built-in averaging to increase reliability, and interference immunity to low frequency noise and DC offsets. The F99x MCU’s CDC is the fastest and most sensitive capacitive-to-digital converter available today, and other devices with equivalent sensitivity take more than 1000 times longer to sample. The highly programmable F99x MCUs enable developers to dynamically adjust active and inactive thresholds to accommodate changing environmental factors for superior sensing reliability (see Figure 3).

    Figure 3: Developers can dynamically adjust active and inactive thresholds to accommodate changing environmental factors for superior sensing reliability.

    Silicon Labs’ QuickSense portfolio includes a wide variety of sensing devices. In addition to the F99x MCUs, Silicon Labs’ F8xx and F7xx MCU families provide advanced capacitive sensing capabilities, optimal performance, efficient power consumption, and low cost for a wide range of applications. For proximity sensing, developers can use Silicon Labs’ Si114x infrared proximity and ambient light sensors. These proximity sensing devices offer power-saving, single-pulsing technology and touchless gesture support. Silicon Labs’ infrared-based proximity sensors are the fastest sensing devices on the market, offering the longest sensing range without compromising power or efficiency.

    Advanced development environment

    As embedded applications continue to become more complex, designing a robust implementation requires not only proven hardware but also production-ready software and world-class development tools. To aid developers, Silicon Labs offers the QuickSense Studio, a comprehensive suite of hardware, software, and development tools to enable developers to quickly and easily introduce capacitive and proximity sensing to any application.

    From an application perspective, both capacitive and proximity sensors can be thought of as simple inputs to the system. By abstracting their implementation through an API, developers can access user interactions, regardless of their source, as touch or gesture events which can be easily mapped to specific functionality, thus substantially simplifying application and interface development. The easy-to-use, GUI-based QuickSense Configuration Wizard accelerates development by generating all application configuration code and firmware drivers required, eliminating the need for developers to understand or write any lower level code for MCU peripherals used to monitor the sensors. Industry-proven firmware manages the different capacitive interface options – including touch button, slider, and wheel – and capacitive proximity sensors. Developers have full control over important sensor characteristics such as sensitivity, operation thresholds, responsiveness, and code size.

    QuickSense Studio also automatically calibrates sensors and provides full debugging and performance analysis capabilities to ensure that designs are responsive and robust. Even if several switches are the same size and shape, for example, their location on the PCB will affect their active and inactive state capacitance given their proximity to other conductive elements, ground planes, and the presence of electrical interference. Each switch will need to be calibrated during development or production with the results programmed to flash memory. In addition, if effects from environmental factors such as temperature, humidity, supply voltage, and contaminants are large enough, incorrect measurements can lead to false sensor events. QuickSense Studio enables systems to account for the dynamic nature of these factors by periodically reconfiguring themselves.

    QuickSense Studio is the only development tool on the market that supports both capacitive and proximity sensing, enabling developers to design a complete user interface using a single development environment. In addition to the Configuration Wizard, QuickSense Studio accelerates the design of the following:

    • Infrared proximity sensing
    • Ambient light sensing
    • Capacitive buttons and sliders
    • Capacitive proximity sensing
    • Complex algorithms
    • Gesture recognition
    • MCU control and communications
    • Capacitive touch screens


    Effective human interfaces involve a combination of pleasing aesthetics and revolutionary ways to interact with electronic devices. Manufacturers seeking to differentiate their products with as little impact as possible on system cost and power consumption can turn to next-generation interfaces based on capacitive and proximity sensing to provide an easier, more intuitive user experience. By combining the capacitive and proximity sensing capabilities of devices such as Silicon Labs’ C8051F99x capacitive touch-sense MCUs and Si114x proximity sensors, developers can introduce next-generation gesture and touchless interfaces to any system quickly and easily using industry-proven hardware and firmware.

    For more information about capacitive touch and proximity sensing solutions from Silicon Labs, visit

    Powered By
    Convergence Logo

    Disclaimer: The opinions, beliefs, and viewpoints expressed by the various authors and/or forum participants on this website do not necessarily reflect the opinions, beliefs, and viewpoints of Digi-Key Corporation or official policies of Digi-Key Corporation.

    Discover the benefits of becoming a My Digi-Key registered user.
      •  Enjoy faster, easier ordering with your information preloaded.
      •  View your order status, web order history
      •  Use our BOM Manager tool
      •  Import a text file into a RoHS query