Cover Story: November 9, 1995

If you want a user-friendly interface for an embedded system, touchscreens offer the most intuitive interaction. By placing the touchscreen over a display or graphic, you let users activate the system by touching the displayed options. Touchscreen technology has improved its basic offerings to eliminate its early reputation of being expensive and unreliable. Vendors have also extended touchscreen capabilities, increasing the technology’s ability to meet diverse application needs.

Touchscreens best suit designs that require frequent interaction with nontechnical users or must work in dirty environments. The devices are easy to use and can tolerate dirt and moisture that would quickly disable a keyboard or a mouse. Some touchscreens can work through a 2-in.-thick barrier. This feature can protect both the system from user abuse and the user from the system’s environment. Compact designs can also benefit from touchscreen technology. Because touchscreens are integral to the display device, the screens eliminate the need for a separate keypad in your design. You can make handheld devices with touchscreens as small as the display itself.

Touchscreen technology comprises a variety of options that let you match the technology to your application. For example, a touchscreen’s sensor uses one of five mechanisms: resistance, capacitance, acoustics, optics, and mechanical force (Table 1). The first step in successfully applying touchscreens, therefore, is to understand these options.

Resistive touchscreens use a thin, flexible membrane separated from a glass or plastic substrate by insulating spacers (Figure 1). The substrate surface and the facing membrane surface have transparent metallic coatings that meet when the user’s finger or a stylus presses on the screen, thus closing an electrical circuit. Four- and five-wire designs are available for sensing the position of the touch.

In a four-wire-resistive touchscreen, electrode arrays at opposite sides of the substrate can establish a 1-D voltage gradient across the substrate’s resistive indium-tin-oxide (ITO) coating. Similar electrodes can establish an orthogonal gradient across the membrane’s ITO coating. Both sets of electrodes also allow the ITO coatings to act as high-impedance probes. When a user touches a four-wire-system screen, the controller establishes a gradient across the substrate. The controller then measures the voltage at the point of touch using the membrane as a probe. Similarly, the controller establishes a gradient across the membrane and uses the substrate as a probe. The two voltages provide the x and y coordinates of the touch point.

In a five-wire system, the substrate has a resistive ITO coating and electrodes on all four sides. The membrane has a single electrode and a conductive coating. When a user touches the screen, the controller establishes first an x-axis and then a y-axis gradient across the substrate. The controller uses the membrane as a probe at all times. The two voltages that the probe senses reflect the point’s x and y coordinates.

Capacitive touchscreens use resistive-ITO sensor coatings but have no membrane. The ITO coating lies below a protective layer of glass (Figure 2). With analog-capacitive screens, electrodes at the corners establish an ac field on the coating, while the controller monitors the current flow through each electrode. A user touches the screen with his or her finger or a conductive stylus, and capacitive coupling between the coating and the finger or stylus draws a small current from the screen. The controller then calculates the touch coordinates from the ratio of the four currents.

Another form of capacitive touchscreen, digital- or zone-capacitive touchscreens, also depend on capacitive coupling between the user and the sensor coating. Digital-touchscreen coating is not continuous, however. The coating forms an array of isolated touch zones. The controller scans through the touch-zone array, establishing an ac field at each zone and measuring the current. The user’s finger draws additional current from the touch zone by capacitive coupling. This current signals the controller that a touch has occurred within the zone.

Acoustic touchscreens rely on absorption of sound energy by the finger or stylus touching the screen. Such touchscreens operate by launching bursts of high-frequency (5-MHz) acoustic energy along two edges of the screen (Figure 3). Reflector arrays along the edges divert the acoustic energy across the screen and redirect the energy to sensors. Because the speed of sound in the glass is constant, the energy’s arrival time identifies its path. A touch causes a dip in the received-energy waveform for both axes. A firmer touch causes a greater dip, providing acoustic systems with a third (z) measurement axis: pressure. The timing of the dips indicates the x and y touch-point coordinates.

You can use either surface-acoustic-wave (SAW) or guided-acoustic-wave (GAW) touchscreens. SAW screens confine most of the acoustic energy to the screen’s surface. GAW screens channel the acoustic energy into the full volume of the screen material.

IR touchscreens use an array of photodiodes on two adjacent screen edges with corresponding photosensors on the opposite edges. These diode/sensor pairs establish an optical grid across the screen. Any object that touches the screen breaks the optical-grid lines that cross the touch point, causing drops in the corresponding sensor-output signals. These drops indicate the touch-point coordinates. Rather than simultaneously establishing all grid lines, IR screen controllers typically scan through the array. Force-sensing touchscreens come in two types: strain gauge or platform.

The strain-gauge touchscreen measures at each corner the stresses that a touch to the screen produces. The ratio of the four readings indicates the touch-point coordinates. The platform touchscreen doesn’t use a screen. Instead, the monitor or display device rests on a platform with force-measurement sensors at the corners. A touch to the display device translates to forces at the platform’s corners. The platform’s controller performs the vector calculations that determine the touch point from the four force measurements through rigid-body mechanics. The controller tracks out static forces, such as gravity, and repetitive forces, such as vibration.

Once you understand the touchscreen mechanisms, your next step is to consider the conditions under which your system must work. Moisture, grease, dust, and other contaminants that have no effect on some touchscreens can cripple others. You need to know if a bare finger, a gloved finger, a stylus, or any combination will activate your system. Keep in mind, also, the type of handling the system will receive. A system in a hospital will receive more reasonable care than a public-information kiosk, which is more susceptible to vandalism. Other factors may also have a bearing on your applications. (See “Don’t forget the details.”)

By understanding basic touchscreen technology and your operating environment, you can begin to evaluate the range of choices. Table 2 shows some of the trade-offs for various touchscreen options. Although most of the technologies are comparable in price, four-wire resistive touchscreens are typically the least expensive option. MicroTouch, for example, offers the SimpleTouch clip-on touchscreen kit for 14- to 17-in. CRTs for $199. The four-wire resistive touchscreens, however, are highly susceptible to wear-out and surface damage because of their flexible outer layer.

Touchscreen vendors, such as Elo Touchsystems and MicroTouch, have addressed the damage and wear-out problems of resistive touchscreens with the creation of tougher outer coatings and the adoption of the five-wire approach. Resistive touchscreens wear out because repeated flexing of the membrane can crack its resistive coating. Four-wire touchscreen membranes need uniform surface resistivity for the sensor to accurately determine touch points. Cracks quickly ruin that uniformity. Five-wire screens, in contrast, require only that the membrane be somewhat conductive. Cracking must be severe to affect the five-wire system’s membrane. The five-wire screens, therefore, have 35 million touches/point, whereas four-wire systems offer only 1 million touches/point.

Five-wire screens are also less affected by surface damage, such as cuts, than are four-wire screens. A cut in a four-wire screen’s membrane compromises the membrane’s ability to establish a uniform voltage gradient, making accurate touch-point determination impossible. The five-wire system’s membrane, however, is merely a probe, so cuts, and even holes, do not affect its accuracy. Cuts and holes may create dead spots on the screen but don’t totally disable the system.

Unlike resistive touchscreens, capacitive touchscreens are all glass and have no flexing parts. As a result, capacitive touchscreens are highly resistant to surface damage and don’t wear out. Their main drawback is that the screens depend upon the user’s conductivity for operation and typically don’t work if the user is wearing gloves. Vendors have developed some capacitive systems, however, that work with gloves and other insulating barriers.

One product line that allows a user wearing a glove to use the screen is MicroTouch’s ThruGlass. The ThruGlass products divide their touchscreens into zones. Each zone projects an electric field as far as 2 in. above the screen surface. Conducting objects that enter the field disturb it enough that the controller recognizes a touch. Intervening nonconducting materials, such as window glass, have no effect on the sensor’s ability to recognize a touch. Thus, a user can activate the touchscreen while as much as 2 in. of glass separates the user from the screen. A variation of the ThruGlass technology lets you place 0.25 × 0.25-in. pads onto the back of a window. The pads transform the window into a touch-sensitive input device.

Capacitive touchscreens had other drawbacks, such as accuracy limits and drift (the change of sensor calibration over time). Vendors have corrected both limitations. MicroTouch, for example, improved the accuracy of its capacitive screens using a Smart Touch Screen (STS) system. STS applies a 25-point calibration procedure built into the controller and software. The calibration procedure prompts the user to touch a series of points on the display screen. The controller then calculates correction factors that accommodate irregularities in both the touchscreen and the display. The correction factors improve position accuracy to within 1%. The company also claims to have eliminated drift from its capacitive touchscreens.

Drift is not a factor with acoustic touchscreens, as it is with capacitive and resistive touchscreens. Acoustic touch-point determination depends not on the more variable electrical characteristics of the screen and user, but on the relatively stable speed of sound in glass. Because acoustic touchscreens use speed, and, therefore, timing, to determine touch point, the devices can identify multiple, simultaneous touches. Capacitive and resistive touchscreens respond to multiple touches by calculating the single touch point that produces the same electrical effect. Acoustic systems sense several dips in the sensor waveform, signaling the presence of multiple touches.

In settings in which the system must be water- and contamination-resistant, acoustic touchscreens prove troublesome. Unless you apply it carefully, any sealant that you use to make the touchscreen water-resistant dampens the acoustic signal. The sealant may also make the screen inoperable. Similarly, grease and water drops on the screen’s surface can disturb the acoustic signal enough to generate false-touch indications.

Vendor discoveries have reduced these contamination and sealing problems for SAW touchscreens. Elo Touchsystems, for instance, uses the acoustic screen’s multitouch-determination attributes to make SAW touchscreens less sensitive to surface contamination and damage. By having the controllers continually scan to determine a background signal level, the Elo screens can ignore false touch points that remain fixed. Elo Touchsystems has used careful assembly of SAW touchscreens to solve the sealing problem.

The introduction of GAW touchscreens from companies such as Carroll Touch and Computer Dynamics has served to eliminate the effect of sealants and surface contaminants on acoustic touchscreens. GAW screens are relatively insensitive to surface effects, because the acoustic energy travels within the glass rather than at the surface. As a result, GAW systems perform reliably in applications in which acoustic touchscreens were previously unable to operate.

Similar to GAW touchscreens, the force-sensing platform is a recent innovation that extends the applicability of touch as an input mechanism. (The TouchMate platform won EDN’s Innovation award for 1993.) The platform has the unique ability to sense touch both within and outside the display area, including points on the bezel and housing. This ability lets you utilize permanent touch zones on the display housing to save space on the display screen. The platform sensors can also make a touch-force measurement, providing you with 256 levels of pressure sensitivity and adding a third dimension to the touchscreen response.

The platform touch sensors are the simplest to install in the field. You simply set the display on the pedestal. Their drawback is the need for frequent calibration. When you place the display on the platform, you must then calibrate the sensors. The sensors need recalibration whenever the display changes position on the pedestal. The sensors are also subject to false readings based on shock and vibration to the pedestal.

One technology that appears to be declining is the IR touchscreen. IR screens are susceptible to dust and other contaminants, need special bezels for daylight use, have more possible points of failure, and cost in scale with the array size. IR screens also suffer from parallax problems when used with curved screens. Because the light must travel in a straight line from source to sensor, the optical grid that forms the touch-sensing surface must be flat. If used with CRT displays having curved faces, the optical grid floats above the screen. This displacement causes parallax between the display and the optical grid, resulting in possible misalignment between the measured and intended touch points if the user views the screen from an angle. The effect is most pronounced at the corners. Having the optical sensor grid float above the screen also causes the touch sensor to register a touch before the user’s finger reaches the display screen. This feature deprives the user of tactile feedback and could result in unintended activation of the screen.

The primary advantages of IR touchscreens are their clarity and immunity to drift. The LED and sensor arrays of an IR touchscreen need no substrate, so the touchscreen doesn’t place anything between the display and the user that might reduce the display’s brightness. Further, because the array forms a stationary optical grid, the touch-point positioning cannot drift. These advantages may outweigh IR’s limitations in many applications.

The many trade-offs between touchscreen technology options are subject to change. Vendors are constantly improving their technology to reduce or eliminate the drawbacks inherent in each choice. Vendors are also creating packaging that extends the utility of their touchscreens. For example, by including a louvered filter in its touchscreen assembly, MicroTouch can produce a touchscreen that only the user can view. This restricted-view angle adds a measure of privacy and security to transactions on a touchscreen system. By repackaging its TouchMate force-sensing platform, the company has created an electronic white board that mounts on a wall. The board’s touch sensitivity allows it to capture drawings for later printing or transmission to a remote display station.

The most recent wave of innovations in touchscreen technology includes the use of both pen and finger touch, with an ability to differentiate between the two. This differentiation is vital to the successful combination of pen and touch. Without this feature, if the user’s hand were also resting on the screen, the system would not be able to accurately read the pen. MicroTouch, Symbios Logic, and Philips Semiconductor have announced three systems. Each system uses capacitive sensing for finger touch, but each uses a different scheme for pen touch.

The MicroTouch TouchPen was the first to enter production. Designed for CRT use, the TouchPen employs a tethered pen that injects current into the touchscreen’s partially conductive surface. Because finger touch capacitively draws current and pen touch injects current, the controller can tell the difference between the touches. This ability, in turn, allows the controller to reject finger touches if the pen is in use. Resting your hand on the screen while writing does not cause unwanted touch responses. A TouchPen kit for 14-in. monitors costs $795.

The Symbios Logic WriteTouch system also uses a pen that injects current into the sensor. Symbios uses the technology of digitizing-pad manufacturer Scriptel (Columbus, OH) to provide a cordless pen that generates a modulated electrostatic field. The field induces current into the sensor layer. Because the user’s hand can act as an antenna for the pen signals and distort the pen-touch-point measurement, the WriteTouch system incorporates a conducting surface layer that grounds the user’s hand. A by-product of the WriteTouch pen-sensing mechanism is an ability to determine pen proximity. The pen begins inducing current into the sensor as the pen approaches the screen, allowing the system to recognize that the pen is near. The system uses this proximity sense to disable finger touch before the pen contacts the screen. The system can, thus, ignore a pen-wielding hand that is resting on the screen even without the pen in contact.

To differentiate between the pen near the screen and a pen touch, the WriteTouch pen incorporates a tip switch that changes the pen’s signal modulation upon touch. Similarly, the pen can include a barrel switch that provides its own unique modulation when activated. The combination of proximity, touchdown, and barrel switches allows the WriteTouch pen to emulate a two-button mouse. Moving the pen above the screen controls the cursor, touchdown acts as a left click, and the barrel switch acts as a right click.

The WriteTouch system, which is now entering production, works in flat-panel displays. An evaluation unit for the system is available for $3000 and includes a display, a sensor panel and pen, a controller card, and software. The OEM version of the system costs $95 (1000) and includes only the sensor panel, a controller IC, software drivers, and a stylus cartridge. The cartridge resembles a standard ink cartridge and contains the electronics for generating the pen signal. By packaging the electronics in the cartridge format, Symbios hopes to free OEMs to develop their own pen designs.

The third pen-and-touch system, the Philips Advanced Interactive Display (PAID), comes from Philips Semiconductor. The PAID system functions only with flat-panel displays. The system’s cordless pen generates an electromagnetic (EM) signal that works with an opaque sensor array mounted behind the display. A glass plate mounted above the front of the display serves as the capacitive touchscreen.

The PAID system does not simply attach to your existing display, however. The EM sensor needs careful matching with the display panel to resist electrical noise radiated by the LCD, so the sensor must be custom-manufactured by Philips for your specific display. The company will start production by year’s end and estimates that the sensors, controller IC, stylus, and drivers will add $35 to the cost of an 8 × 10-in. flat-panel display in volume production.

Similar to the WriteTouch system, the PAID system offers pen-proximity detection and pen-tip and barrel switches. The PAID system, however, allows the pen and the touchscreen to sense 64 levels of pressure. The pen senses pressure through its tip, changing its radiated frequency to signal the pressure level to the sensor. The touchscreen capacitively senses pressure through a change in separation between the display and the touchscreen.

The availability of pen-and-touch systems and other innovations has greatly expanded the range of applications that touchscreens can service. The durability of resistive screens makes them more economical for public-information kiosks. The projected capacitance technique allows touchscreens to operate away from the user. You can, therefore, create shopping-display kiosks behind store windows, games embedded into bar counter tops, and capacitive systems that work with gloved users.

The pressure-sensing ability of acoustic touchscreens adds a degree of freedom to touchscreens. Industrial-control panels with variable settings can use pressure, rather than a slider or multiple touches, to control changes. For example, pressing harder on the throttle touch point of an engine-controller touchscreen can rev the engine.

Pen capability in touchscreens provides an additional level of control and interaction. The pen allows use of a denser touch-point pattern than does the finger, so electronic forms with check lists become an option. The duality of pen and touch allows the system to do even more by offering separate behaviors for the two entry methods. A system can use finger touch for selecting and activating options and use the pen for annotations, drawings, and checklists. The system could also use finger touch to erase pen markings. The pressure-sensing capability of the PAID pen adds another dimension to screen operation. For example, depending on pen pressure, paint programs could vary line width or brush pattern and allow for electronic calligraphy.

One application that suits pen-and-touch systems is electronic-transaction validation. The touchscreen provides the user with an intuitive interface for browsing through options; even 2-year-olds know how to point at what they want. The pen enables the system to capture a signature and validate the transaction. The pressure-sensitive pen of the PAID system can add the security of signature verification. By basing verification on a combination of measurements that include the signature’s shape, speed of inscription, and pressure variations during writing, the PAID system can virtually eliminate the potential for forgery. Philips has such verification algorithms under development.

With these improvements, touchscreen technology has become a viable user interface for many embedded systems. The inclusion of electronic-ink services in Windows 95 indicates that touchscreens will become a dominant interface. You need only to carefully match the technology to the application environment.