Analysis: Difficulties and precautions in capacitive touch screen design

　　The most far-reaching technological change in terms of touch screen performance is the shift from resistive to capacitive touch screen technology. According to market research firm iSuppli, by 2011, nearly 25% of touch screen phones will shift from resistive to capacitive touch screens. The various benefits brought by capacitive touch screen technology will drive the market to grow rapidly.

　　When a traditional resistive touch panel senses a finger or stylus, the top layer of flexible transparent material is pressed down and contacts the conductive material layer below; while a projected capacitive screen has no moving parts. In fact, the projected capacitive sensing hardware consists of a top layer of glass material, followed by X and Y axis components, and an insulator layer of indium tin oxide (ITO) covering the glass substrate. Some sensor suppliers will make a single-layer sensor with embedded X and Y axis sensors and small bridge components in a single layer of ITO. When a finger or other conductive object approaches the screen, a capacitance is generated between the sensor and the finger. This capacitance is quite small relative to the system, but it can be measured using a variety of technologies.

　　One of these technologies uses TrueTouch components, which involve rapidly changing capacitance and measuring the discharge time using a bleeder resistor. This all-glass touch surface gives users a smooth, fluid touch feel. End product manufacturers also prefer glass screens because the glass material gives the end product a sleek industrial design and provides a high-quality capacitive signal for measuring touch. Finally, it is important to not only consider the appearance of the touch panel, but also understand how it works. To design a touch screen product with good performance, the following parameters must be taken into account.

　　Accuracy: Accuracy can be defined as the maximum positioning error in a predefined touch screen area, measured in straight-line distances between the actual finger position and the measured position. When measuring accuracy, a simulated or mechanical finger is used. The finger is placed at an exact location on the panel, and the actual finger position is compared to the measured position. Accuracy is very important, and users expect the system to accurately find the finger position. One of the most criticized shortcomings of resistive touch screens is their low accuracy, which gradually decreases over time. The accuracy of capacitive touch screens has created many new applications, such as virtual keyboards and handwriting recognition without a stylus. Figure 1 shows data from an incomplete touch panel showing wandering finger position, when in fact the simulated finger is moving in a straight line.

　　Figure 1. Example showing inaccuracy or error in touch panel tracking.

　　Finger spacing: Finger spacing is defined as the shortest distance between the center points of two fingers on the screen when the touchscreen controller measures their position. Finger spacing is measured (Figure 2) by placing two simulated or mechanical fingers on the panel and gradually moving the two fingers closer together until the system detects that they are one finger. Some touchscreen vendors measure finger spacing from edge to edge, while others measure it from center to center. A 10mm finger spacing for a 10mm mechanical finger could mean that there are multiple fingers touching the screen or that the fingers are 10mm apart, depending on the touch controller specifications. Without good finger spacing, it is impossible to design a multi-touch solution. Finger spacing is especially important for emulated keyboards, where the fingers are typically very close together on the screen.

　　Figure 2 Measuring finger distance

　　Response Time: Response time is defined as the time between a finger touch on the touch screen and the generation of an interrupt signal by the touch screen controller. It is measured by electronically stimulating a finger touch screen environment or moving a simulated finger across the panel. Response time is particularly important because it directly affects how fast a user can move a finger across the screen; pan or flick; or write on the screen with a finger or pen. Touch panels with slow response times will experience brief pauses and no movement detected. The response time of a touch screen is a portion of the system response time, which includes:

　　X/Y Axis Scan: The time it takes for the touch controller to scan and measure the change in capacitance on the sensor.

　　Finger detection: Compare the panel capacitance change with the pre-defined finger default value. If the change exceeds the finger default value, a finger touch is detected.

　　Finger position: The actual position of the finger is determined by calculating the result data obtained from multiple sensors.

　　Finger Tracking: When multiple fingers are placed on the sensor, each finger must be correctly identified and assigned a unique identification symbol.

　　Interrupt latency: This is the delay between the interrupt indication and the service on the host. In most systems, this delay will not exceed 100 microseconds.

　　Communication: Generally, the system uses I2C at 400kHz or SPI at 1MHZ to communicate with the host.

　　There are many tools on the market that can be used to shorten the response time. The key lies in the intelligence of the touch chip. For example, a more creative method only needs to scan part of the screen to detect the finger position. When the finger is detected, it can be scanned quickly to calculate the actual position of the finger, thereby saving power and time. Another important tool is parallel processing, which uses different hardware components for scanning, finger processing and communication to synchronize these tasks. Using highly optimized algorithms for finger detection, finger positioning and finger identification (ID) can shorten processing and response time.

　　Frame refresh rate: The time between two adjacent frames of touchscreen data in a data buffer when a finger appears on the touchscreen. Low frame refresh rates can cause the system to pause in detecting movement, and the detected movement path will become discontinuous line segments instead of smooth curves. In other words, if the touch panel has a high frame refresh rate, it can provide more data points that can be translated into smooth or complete shapes or movement trajectories. In addition, a high frame refresh rate can also improve the interpretation of gestures. Intelligent touchscreen controllers such as TrueTouch can adjust their frame refresh rate to match system requirements. Hand-drawn or handwritten applications require a very high frame refresh rate, but a mobile phone dial keypad only needs to intercept the host terminal when the user presses or releases a button.

　　Average power consumption: refers to the average power consumption of the touch system, including the time scanning, processing, communication, sleep, etc. when the controller IC is working, as well as the time the main processor receives and interprets touch data.

　　Power consumption is a common performance parameter: the current consumed by the measuring device is multiplied by the voltage, and the power consumption can be inferred. When it comes to the power consumption of touch panels, more sophisticated calculation formulas are required, because different usage modes will produce different power consumption. The standby time of a mobile phone depends on the current consumed by the touch screen in standby or sleep mode.

　　When the touch screen is working, it is also divided into many modes, such as wake-up on touch (WOT) and cheek detection (CheekDetect). For example, when answering a 5-minute call, while checking or entering a phone number, the phone may switch to touch mode for 10 seconds, and then switch to WOT or cheek detection mode for call reminder. Even when sending text messages (SMS), it is still a mixture of WOT mode and actual finger contact. When typing or thinking, the controller IC will switch between various sleep modes.

　　If you don't consider these power modes, it's easy to be misled by the system's power consumption. In most cases, the touch screen switches to cheek detection mode and touch wake-up mode 90-99% of the time. Some systems allow users to set the ratio of processing time to sleep mode, even when the finger is still on the panel. If the system only detects the finger in the same position, it doesn't need a 200MHz screen refresh rate. To develop a high-performance touch screen, you must use a low-power system in sleep mode and work with innovative sleep and wake-up modes.

　　There are many other important factors that system developers need to consider when designing a capacitive touchscreen system:

　　Finger capacitance: refers to the capacitance measured between a finger and a single sensor component. When measuring finger capacitance, a real finger is used instead of a metal mechanical finger to ensure that the measured data is consistent with the actual situation. Factors that affect the feedback capacitance (CF) include the thickness of the lens covering the upper layer and the dielectric constant of the outer covering material.

　　System Noise Floor: System noise floor is the noise measured at the output of the capacitor-to-digital converter and is the input (capacitance) value of the data converter.

　　Signal-to-Noise Ratio: The signal-to-noise ratio (SNR) is the ratio of the finger signal measured by the sensor to the measurement noise. This is an important parameter that designers must understand in order to develop an efficient touch panel. The system must be able to adjust, adapt, and filter out parasitic noise in the mobile system. To achieve high signal numbers and very low noise numbers, consider using accurate analog front-end components for touch functions.

　　Products such as the TrueTouch family of programmable solutions offer many excellent mechanisms for filtering out noise. PSoC programmable analog components can be reconfigured to integrate signals over time to filter out noise. Different signal frequencies, including spread spectrum and virtual random frequencies, can also be used to avoid electromagnetic interference. Standard digital filters can remove 1-2 bits of signal jitter or provide an IIR-like low-pass filter. Smart digital filters can compare samples detected in nearby areas and filter out abnormal samples. Smart filters are only limited by the creativity of the system designer. Figure 3 shows an example of the noise level of a component and the detected touch behavior. In this example, the captured SNR is 5.

　　Figure 3. Signal-to-Noise Ratio (SNR) example

　　Understanding and mastering the key touch screen performance parameters can significantly improve touch screen design. Understanding these criteria also helps in choosing the ideal design partner who has the right technology to properly address the noise and electrical issues of mobile consumer products.

　　The appeal of touch screens lies in their deceptively simple design. Replacing bulky buttons, trackballs, or traditional screens, touch screens offer a whole new mode of operation, creating a user experience that is enjoyable to use. The difficulty in designing a touch screen is that in order to provide a beautiful and simple design, sophisticated hardware, firmware, and manufacturing techniques must be used. Understanding the design essentials, key performance parameters, and trade-offs of touch screen design is the first step to developing a first-class touch screen product.