Touch the Future: Projected-Capacitive Touch Screens Reach for New Markets

Spurred by successful implementation in devices such as the Apple iPhone and the LG Prada phone, projective-capacitive touch screens seem poised for mass adoption in various applications. Here is an overview of the technology and how to decide which type of projective-capacitance touch screen to use in your product.

by John Feland

THE LAUNCH of the LG Prada phone in March 2007 [Fig. 1(a)], followed by the Apple iPhone in June 2007, iPod Touch in September 2007, and the Samsung Yepp YP-P2 in October 2007 [(Fig. 1(b)], signaled to the world that transparent projected-capacitive touch screens are ready for mass adoption. Prior to 2007, transparent projected-capacitive was a niche technology with little impact.

Total worldwide sales of projected-capacitive touch screens in 2006 were estimated to be less than $20 million; sales in 2008 could be five times that, as several varieties of this technology make their way into various platforms, and consumer-electronics companies use it to transform the end-user experience across multiple markets.

As companies seek to leverage this maturing technology in their products, several questions arise, such as which type of projected-capacitive sensor is right for my application? What is the trade-off between glass and PET substrates? These and other questions will be addressed below.

Capacitive Touch Screens

Two main types of touch screens use capacitive sensing as the main input method: surface-capacitive and projected-capacitive. Surface-capacitive touch screens use a sheet of indium tin oxide (ITO) with at least four electrodes around the periphery. These electrodes sense the change in the surface capacitance when a grounded object, such as a finger, approaches. This method has been used for kiosk touch screens for quite a while, with 3M MicroTouch being one of the major suppliers of such technology.


Fig__1a_tif (a) Fig__1(b)_tif (b)

Fig. 1: The LG Prada phone (a) and the Samsing Yepp YP-P2 media player (b) are two of the first consumer-electronics products to use projected-capacitive touch screens.


However, there are some limitations to surface-capacitive touch screens. They can only recognize one finger or touch at a time. Also, given the electrode size, small screen sizes such as those used on handheld platforms are impractical.

Projected-capacitive touch screens are named for the electrostatic field lines these sensors project from their electrodes. There are two types of commonly used projective-capacitive sensing technology: self-capacitive and mutual-capacitive.

The most widely used method, self-capacitive (also called "absolute capacitive"), uses the object being sensed as the other plate in a capacitor. This object is sensed typically by driving a charge between the sensing electrode and the sensed object, and then measuring the charge stored in the resulting capacitive coupling. Figure 2 illustrates how this principle works.

Mutual-capacitive (also called "transcapacitive") sets up a capacitive coupling between neighboring electrodes. When the sensed object approaches the field lines projected from one electrode to another, the change in the mutual-capacitance is sensed and reported as position. Mutual-capacitance sensors have been used extensively as conductivity sensors for oil conditioning in automotive applications.

Millions of self-capacitive solutions are used by people every day for position sensing, namely, the touchpad that is ubiquitous on today's notebook computer. The typical notebook touchpad uses an array of X and Y sensing electrodes to form a sensing grid. When a finger approaches the touchpad, the electrodes push a small amount of charge between that finger and the sensor electrodes. Algorithms then process the signals from this row-column sensor to resolve the location of the sensed object, in this case a finger.

In both types of projected-capacitive sensing, the sensor electrodes can be designed in a fashion so as to be able to detect more than one finger on the sensor at any given time.

The basic theory of operation for opaque projected-capacitive sensing on devices such as touchpads and projected-capacitive touch screens is the same; the differences lie in the sensor electrode materials, sensor substrates, manufacturing methods, and many other items in the solution stack. Touchpads can be made of opaque materials and use metallic or carbon-based electrodes in the sensing area. Projected-capacitive touch screens must be transparent and therefore are often made with the same transparent conductor found on resistive touch screens, i.e., ITO.

However, unlike resistive touch screens, projected-capacitive touch screens do not require an air gap between layers or the abilityto deform any of the layers; hence, a sensor can use rigid glass or a PET substrate. Another key difference between the construction of projected-capacitive and resistive touch screens is the requirement for the ITO to be patterned onthe former rather than deposited in a continuous film as in the latter. This extra complexity is well worth the trouble, given the benefits in using projected-capacitive touch screens.

Synaptics, for example, uses a patented diamond pattern on the multiple layers of its ClearPad sensor. The sensors in the X-axis form one layer, the sensors in the Y-axis another layer, then a ground or shield layer rounds off the stack-up, as shown in Fig. 3.

There is no clear choice between glass and PET substrates. Both can be laminated to a plastic or glass lens (screen cover), depending on the OEM's product design. Glass tends to be a bit thicker, heavier, and more expensive but offers greater overall stiffness, potentially reducing some costs elsewhere in the device. Glass has higher transmissivity than PET, though both are superior to resistive touch screens of the same size. PET sensors are thinner and easier to laminate to the product lens (because laminating a flexible material to a rigid material is easier than laminating two rigid surfaces). Both glass and PET substrates can be used to manufacture self-capacitive and mutual-capacitive touch screens, since the manufacturing methods are very similar.

Most sensor suppliers utilize a continuous batch sputtering process to etch the ITO pattern onto the substrate.

3M MicroTouch announced in 2007 the availability of a roll-to-roll method of manufacturing projected-capacitive sensors. In the past, etching such a pattern, though clear, caused a difference in the reflectance of the surface of the touch screen, causing the pattern to be visible as light played across the surface. Recent advances in reflectance matching have rendered the sensor pattern all but invisible.

While surface-capacitive touch screens have practical limitations on how small they can be produced, projected-capacitive touch screens have limitations on their maximum size.

The sensor electrodes have to be close enough so that the finger can affect the field lines of at least two electrodes to interpolate the position of the finger. As such, the number of sensor electrodes needed increases geometrically as the screen size increases. As projected-capacitive touch screens increase in size, the number of sensor electrodes that need to be routed back to the controller increases rapidly, forcing the inactive border of the sensor to increase as well. There are a few tricks to create larger projected-capacitive touch screens, but none of these schemes have been tested as a real product as of yet.

Controllers Are the Key

Without a controller, a sensor is just an inert piece of glass or PET. Compared to the broad proliferation and integration of resistive touch-screen controllers in everything from application processors to MP3-decoding chips, projected-capacitive touch screens still require specialized silicon to drive the sensors and decode the position of the finger or fingers on the screen.

The approach that Synaptics takes in the ClearPad modules on the market today uses the self-capacitance technique, borrowed from the millions of notebook touchpads already in service. The sensing scheme used by the Synaptics controller polls each sensor trace on the X-axis and then each sensor trace on the Y-axis, looking for the maximum capacitance point on each axis. This technique provides good rejection of unison noise such as changes in moisture, temperature, or even an external noise source such as 60-Hz line noise.



Fig. 2: This illustrates how self-capacitive touch screens work.


Apple uses the mutual-capacitance technique for the iPhone and iPod touch screens. The sensing scheme used by the Apple/Broadcom controller excites each line on the Y-axis one at a time. For each Y-axis line, the controller measures the capacitance at the intersection of that line with each X-axis line. The result is an "image" of whatever is touching the surface at each of ~700 X–Y intersections. However, this technique is very sensitive to environmental noise, much more so than the self-capacitive technique. Since none of the other controller suppliers listed above are in any shipping products yet, it's unknown what technique they use.

Rich Palette of Gestures

The feature that is driving the adoption of projected-capacitive touch screens in general is the rich palette of gestures now possible. The user-experience and user-interface designers on many OEMs' product-development teams are hungry for this new capability. The use of intuitive gestures holds tremendous promise in reducing the complexity of today's consumer devices.

One question that's often asked is where should the processing of gestures take place? Gestures can be processed and decoded in four places: in the touch-creen controller, in a separate CPU or DPS, within the touch-screen driver on the host CPU, or in the application that's running on the host CPU. As in the "glass vs. PET substrate" question, there is no single correct answer – each architecture has tradeoffs. In the Apple iPhone, the touch-screen silicon consists of two separate chips: a Broadcom analog controller that processes the raw analog signals from the sensor and converts them into a digital data stream of multiple X and Y points, and an NXP (Philips) ARM-7 CPU that decodes the digital data stream into gestures. In the Apple iPod Touch, these two pieces of silicon are combined into a single, second-generation Broadcom chip that includes both analog and digital cores. One reason that Apple chose to process the gestures on a separate CPU (rather than the host CPU) was to ensure the fastest possible response to gestures. The iPhone includes a total of five or six separate ARM cores; it's clear that the overall product architecture is that of distributed computing.

Synaptics used a different approach in the Samsung Yepp YP-P2 media player. With much simpler functionality than a smartphone, a media player generally only has one CPU, which limits the range of possible gesture-processing choices. The Samsung Yepp YP-P2 media player uses Synaptics ChiralMotion gesture as the main method of searching through the various applications. (ChiralMotion is an intelligent virtual-scrolling gesture that allows the user to control the direction and speed of scrolling by varying the speed and radius of arcs through which he moves his finger.) The touch-screen controller outputs a digital data stream of single X and Y points from the touch screen. The recognition of the ChiralGesture takes place in the touch-screen driver running on the host CPU. The driver notifies the UI application of the user's intent, so that the UI application can allow the user to "tunnel" through the other applications on the player for a nice blend of eye-candy and ease-of-use.



Fig. 3: Synaptics's patented diamond pattern used on its ClearPad sensor.



Projected-capacitive touch screens are one of the most promising interaction methods for tomorrow's handheld devices. Thinner, more robust, and optically clearer than resistive, supporting the use of multiple-finger gestures, and pushing the industrial design envelope, this next generation of touch screens is likely to see a broad adoption. Already the portfolio of available options (self-capacitive, mutual-capacitive, under glass, under flat or curved plastic, etc.) is giving OEMs tremendous flexibility in how they integrate this maturing technology.

Already we are seeing numerous experiments as to how handheld device manufacturers are using projected-capacitive solutions to differentiate not only their design, but also the user's experience. We expect to see more of these devices in the market soon with successful product launches. •


John Feland is a human interface architect at Synaptics, Inc., 3120 Scott Blvd., Santa Clara, CA 95054; telephone 408/454-5383, e-mail: