As electronically controlled products become more sophisticated in their capabilities, the touchscreen has become an increasingly popular interface, one that places a burgeoning array of features at a user’s fingertips. In a growing number of applications, both large and small, the touchscreen often provides a design engineer with an optimal solution: one that avoids the cost of discrete buttons and switches, while eliminating the complexity of scroll and slew controls.

Designers have had several touchscreen technologies from which to choose, including resistive, capacitive, infrared, and surface acoustic wave. And each has its own set of advantages and disadvantages that need to be considered. But before making the leap to one of the more familiar choices, designers should pause and take a look at a new item on the touchscreen menu.

Recently, the Tyco Electronics business, Elo TouchSystems, Menlo Park, Calif., developed an alternative touchscreen technology based on acoustic pulse recognition (APR). The essence of APR is quite simple. A glass overlay is mounted in front of a display. With the help of sensors and an electronic controller, APR recognizes the distinctive sound that is created when the glass is touched at a specific position.

Distinctive is the key word, because each location on the glass generates a unique sound when touched, a sound detected by four piezoelectric transducers positioned at different places along the edges of the glass. The sound is then digitized by the controller and compared to a list of prerecorded sounds for every position on the glass. When a match is found, the controller knows where the touch has occurred. APR ignores extraneous and ambient sounds, as they do not match the stored sound profiles.

By using a simple lookup table, the APR approach avoids the need to have sophisticated signal processing hardware, which would otherwise be required to calculate the touch location without a frame of reference. This makes APR cost-effective for both large and smaller displays.

Another advantage of using APR is that it possesses many of the desired features found in other touch control technologies. APR has the optical qualities, durability, and stability found in surface acoustic wave and infrared technologies; APR has the dragging properties of capacitive touch sensing technologies; and APR has the low-cost advantages of resistive touchscreens, along with their ability to be activated by a stylus, gloved finger, or fingernail.

In addition, APR is resistant to water and other contaminants on the screen, can be scaled from large to small displays, and provides palm rejection during signature capture. These features can make APR suitable for a wide range of applications including vending machines, kiosks, point-of-sale (POS) terminals, gaming machines, foodservice appliances, medical equipment, business machines, PDAs, mobile phones, and other consumer electronic devices.

These are some of the key features of APR:

 

• -- Uses a simple glass overlay, with nothing more than an anti-glare coating. Glass provides a hard, durable overlay with high light transmission that preserves display colors and clarity. Glass is scratch resistant, chemical resistant, and minimizes reflections. Thicker, or treated, vandal-resistant glass can also be employed where necessary.
  
• Recognizes touch by any type of object. The operator can touch with a finger, fingernail, gloved finger, stylus, pen, credit card, etc.
  
• Resistant to contamination and moisture. The overlay can be sealed to watertight standards, and the glass is resistant to dirt, grease, oils, and cleaning and sterilizing chemicals. The technology will work even if the glass becomes scratched.
  
• Never needs calibration. APR uses a fixed-coordinate system that does not change over time, position, or environmental conditions, so it never exhibits drift.
  
• Exhibits fast response. The system recognizes a quick tap.
  
• Allows dragging. Effectively handles dragging techniques with either a finger or a stylus.
  
• Has narrow borders. Borders for an APR overlay are only 5 mm, including sealing area, making it among the most narrow of any overlay touch technology. Narrow borders permit multiple LCD panels to be placed closely side-by-side, which is becoming more common in some applications.

APR technology also permits specified regions of the screen to be ignored when touched, simply by programming the controller to skip those locations in the lookup table of prerecorded sounds when searching for a match. This feature allows the use of palm rejection for on-screen signatures in POS equipment, which is not easily implemented with other touch technologies. (Palm rejection refers to the tendency of people to rest their palms on a touchscreen while using a stylus to record their signature on the same touchscreen. With some technologies, the secondary touching of the palm can interfere with proper registration of the signature.)

One of the few things that APR cannot do is register a hold. Since the sound is generated by the initial impact of a fingertip or stylus, the technology does not afford any means for recognizing if and when the finger or stylus has been removed. It is, therefore, currently not capable of recognizing touch-and-hold or drag-and-hold maneuvers.

Sound methods

Fig. 1. Conceptual drawing of a bending wave touchscreen.

For those already familiar with surface acoustic wave (SAW) touchscreens, it is important to understand how acoustic pulse recognition (APR) differs, given that they are both acoustic-based technologies.

Many types of sound waves, or acoustic modes, can be propagated in glass plates. The acoustic mode most efficiently excited by a finger touch is a bending wave. Just as a beam bends under the weight of a load, a glass plate will bend, if only slightly, under the load of a finger or stylus. With APR, these transient touch forces generate waves of bending in the glass that are detected by piezoelectric transducers at the edges. (See Fig. 1.) The resulting signals from the transducers are digitized by electronics and numerically processed to reconstruct touch positions.

Bending-wave touchscreens and surface-wave touchscreens have much in common.  Both are acoustic touchscreens that require nothing more than a glass plate for the touch input area. Both approaches share the valued features of high transparency, absence of wear under normal usage, and stable calibration based on the speed of sound.

However, bending-wave touchscreens and surface-wave touchscreens are only distant cousins. For example, bending-wave touchscreens are completely unpowered signal sources, while surface-wave touchscreens must be powered to constantly generate the waves that illuminate touches. A noteworthy distinction that concerns the effect of contaminants on the touch surface.

Bending waves are difficult to stop once excited because bending wave power is distributed throughout the entire thickness of the glass plate. By contrast, surface wave power is concentrated at the surface of the glass. Another difference is the operating frequency. Bending waves propagate at lower frequencies that are harder to stop. Surface waves propagate at higher frequencies that are easier to stop. As a result of these differences, bending waves are not affected much by contaminants on the touchscreen surfaces, whereas surface waves can experience interference by such contaminants, by water, or even by the simultaneous, inadvertent touch of a palm.

As seen in Table 1, bending wave technology affords a number of advantages over other touchscreen methods, which then raises the question as to why it has only recently been deployed. The answer lies in the engineering challenges of processing the signals produced by the transducers detecting the bending waves.

Measuring the time of flight of bending waves is the most obvious approach for touch position reconstruction, and has served as the basis for all prior research in bending wave touchscreens. The concept is simple.

A finger touch generates waves that are detected by piezoelectric transducers at known locations. The time delay between the touch event and the start of the piezo signals, divided by the bending wave velocity, gives the distance from the finger to each of the piezos. With a sufficient number of piezos, and appropriate math, the touch position is uniquely determined. However, while the concept is simple, the actual implementation contains several challenging complexities:

• Bending waves are highly dispersive, and bending wave velocity increases with the square root of frequency. This means that the faster, high-frequency components of the wave move ahead of the slower, lower-frequency components. This phenomenon makes it difficult to determine the arrival of such a dispersed wave packet, especially since the frequency content of the touch itself is not well controlled.
  
• The piezoelectric transducers receive not only the direct bending wave, but also the bending waves produced by reflection of the perimeter of the glass plate. Glass edges are very efficient bending wave reflectors, and the bending waves typically reflect many times before damping away. These reflections increase the complexity of time-of-flight signal processing.
  
• Finger touches are not ideal wave sources. A finger touch does not occur crisply and quickly like a click. To a person, a rapid finger touch may seem like a quick event, but in the lightning-fast world of signal processing, it is, in fact, a rather slow event that further compounds the complexity of measuring time of flight.

The answer

Enlarge this picture Table 1. Comparison of touchscreen technologies.

The solution to this complexity problem is counter-intuitive. Researchers at Elo TouchSystems decided to look at bending wave signal complexity as a benefit, instead of a drawback. The approach they developed, acoustic pulse recognition, actually exploits signal complexity.

The complex propagation of bending waves yields a unique pattern for every touch location. Every touch point creates a unique acoustic signature. With APR, these signatures are recorded for every possible touch point, along with their corresponding coordinates. The signatures are stored on a microchip in the form of a simple lookup table. When the sensors transmit an acoustic signature to the processor, it looks for a match in the table. Once the match is found, the coordinates of the touch are then known.

With this method, there is no need to measure time of flight or perform sophisticated computations to correct for the effects of dispersion and echoes. There is no need to clean up the signal – it is the very “messiness” of the signal that makes it unique and identifiable. Another analogy is taking a fingerprint and matching it to one in a fingerprint database. Only here the pattern is acoustic rather than visual.

Looking at the photo image of the APR touchscreens, one can see the asymmetrical locations of the piezoelectric sensors. This is done deliberately to actually increase the complexity of the signal and ensure the uniqueness of the acoustic fingerprint for a given location. The small black object near the end of the cable is a 4 Mbyte memory that contains the table of acoustic fingerprints.

Because dragging a finger or a stylus across the glass overlay produces a continuous stream of bending waves, the APR technique also exhibits excellent drag performance, that is, tracking a sliding finger from one location to another.

Elo TouchSystems launched the APR touchscreens earlier this year, initially targeting retail and restaurant point-of-sale equipment, but the technology has the potential to serve a broader array of applications in the future.

For more information, email: eloinfo@elotouch.com