Mobile device manufacturers want to provide consumers with a natural and easy-to-use user interface, which is as convenient as using paper and pen. At the same time, they also want to have the high flexibility of computers. Such advanced features can help manufacturers achieve product differentiation. The small pointy passive stylus with palm malfunction exclusion enables manufacturers to provide low-cost solutions for new applications such as handwriting, editing, signature capture, and precise navigation. However, the implementation of the above functions also faces some challenges. The developers of passive stylus must meet a series of performance requirements of capacitive sensing technology on larger touch screen. Specifically, we need to use advanced algorithms and induction methods to detect the small signal from the stylus, and eliminate the large useless signal caused by the user’s palm.
With the increasing size of capacitive touch screen, the writing device similar to paper and pen becomes more intuitive and convenient. The most common way for manufacturers to support the stylus function is to use an active or passive stylus. The active writing pen uses electronic components, needs a power supply, and transmits signals to the host device. The active stylus can support hovering on the display screen, pressure sensing, key support, erasing and other advanced features. The passive pen is made of conductive material, which is equivalent to the extension of the user’s body. The capacitive coupling of the user’s hand supports the passive stylus to send signals when touching the screen. There is no active communication between the stylus and the host platform, so how to distinguish between the finger and the passive stylus is a difficult problem.
In many cases, if both active and passive pens can achieve the same characteristics, there is no need to add extra cost to the system. The additional components and power requirements of the active pen make it difficult to open a market, while the poor performance and / or bulky head of the passive pen will bring unnatural handwriting experience. Therefore, if the pen head of the passive pen is 1 to 2 mm, the user’s palm can be placed on the screen when writing, while maintaining enough speed and accuracy, and ensuring that the contact point is just the “inking place” of the “ink”, the user experience of the passive pen can be improved.
In order to create a practical implementation scheme, which can support finger and passive pen operation at the same time, we must consider a variety of different use cases. For example, developers should consider how fast the system will switch between detecting finger and stylus input. They also define how the system reacts when the stylus is in front of, behind, or touching the screen at the same time. Other important factors include how close the stylus is to the hand, the signal of the stylus will not be detected. Passive stylus detection is a complex problem for touch control engineers. The root of the problem lies in the “pen paradox”. The so-called “stylus paradox” means that the signal of the passive stylus is much smaller than that of the normal finger touch input, while the user thinks that the stylus tip is so thin that it should be more accurate than the finger.
The accuracy and linearity are proportional to the SNR of the system. Since the noise floor will not change with the input, signal weakening will have a great impact on the signal-to-noise ratio. The signal level of capacitive touch screen depends on the coverage area of touch input. That is to say, the signal strength of a 2 mm passive stylus is 25 times less than that of a typical 10 mm finger touch. This gap in signal strength causes many problems for touch engineers. Even if there is a large touch signal, the firmware must be able to detect a small stylus signal, which often requires different sensor scanning modes, and the noise immunity and refresh rate are affected. In addition, the passive stylus is most suitable for large touch pad, but the refresh rate of large touch pad is low, or the use of large distance sensors, both of which will affect the performance of the system.
Fundamentally speaking, two problems need to be solved to deal with the signal strength gap. First, although the signal strength is very low, the stylus must be detected first. Second, once a stylus is detected, it must be reported accurately. These two problems have their own difficulties. Conceptually, the most reasonable method is to maximize the sensor signal. Generally, the problem is solved by minimizing the dynamic range from the sensor to the signal level, or even by using software multiplication and filtering. However, the high gain system can easily reach saturation through large input such as normal finger touch, so it is necessary to deal with normal touch and small stylus signal seriously. A common method is to perform two independent scans at each expected signal level to distinguish the normal touch from the stylus input.
This mode switching is easily affected by error detection, so the error detection must be filtered out. A typical example is a finger moving close to or away from a touch screen. When the finger approaches, the signal level is very low (in the passive pen area), and when the finger leaves, the signal level is also very low, so other judges must be used to confirm any detected pen input.
When a stylus is detected, it must be reported accurately. Unlike a typical finger touch, the tip of the passive stylus allows the user to see exactly where it is placed on the LCD. Therefore, although the signal-to-noise ratio is significantly reduced, users have higher expectations for the accuracy of pen manipulation. In addition, linearity is also a key factor because the stylus is usually used for writing. The key problem related to the accuracy and linearity of passive stylus is “dead zone”. “Dead zone” refers to the area on the touch screen where the reported signal level does not change even when the input stimulus is transferred to a new position. For example, a 2mm passive stylus tip can be completely surrounded by a typical 5mm sensor on the touch screen.
It is difficult to detect the small movement of the stylus in the center of the sensor, but for the sensor, the input is usually quantized to the center of the component. Therefore, when the movement of the stylus is limited within the range of the sensor, it will be reported to be in a fixed position, which is the so-called dead zone.
The general way to solve this problem is to analyze all the surrounding sensors and use them to create a look-up table index to correct the position of the report, so as to better grasp the actual position of the pen tip. Therefore, in the final analysis, the problem of the accuracy and linearity of the passive pen is to generate the above location index through a very creative method, or design a more advanced look-up table, because the dead zone is usually an insurmountable physical problem, so we must find an appropriate correction method.
The early implementation of passive stylus only supports a single input type at a time, and normal finger touch has higher priority. If there are normal finger touch on the screen, including holding the edge of the mobile phone or tablet computer, or putting the palm on the screen, the stylus system will not work properly. However, these two situations are very common when using a stylus on a large screen. In order to facilitate the use, when the stylus works on the screen, we must eliminate this kind of false touch and false action, so as to improve the user’s satisfaction.
The influence of touch screen on the performance of stylus also depends on the signal gap. Touch screen will cause its signal to spread to multiple sensors, while peripheral sensors are usually in the signal level of the stylus area. The signal level of normal touch is much higher than that of stylus. It’s like having two flashlights in a dark room, one very bright and the other very dark. The stronger the flashlight, the harder it is to see the darker one. In addition, normal touch can also produce common mode noise. Therefore, if the noisy touch and stylus share the same sensor receiver, the input of the stylus will be difficult to detect.
These common mode noise problems are another big problem. In general, we can solve this problem by scanning only the specific sensor of interest to isolate the desired signal of the passive stylus. At this time, we assume that we can initially detect the stylus and track its movement on the screen, so that the first touch of the stylus is the weakest. However, once the sensor subset tracks the stylus, most of the annoying touch problems can be solved.
Although most of the above problems seem to be difficult to solve, the current development of touch controller makes our products not only have enough sensitivity to detect small pointed passive pen, but also have intelligent function to filter noise and other interference objects on the screen. From the user’s point of view, intelligent touch controller can deal with many input related problems of detecting and tracking touch objects. From the system level, the key to success lies in the development of related applications, so that users can better use, create and control programs on their own devices, so as to improve work efficiency and bring more natural control experience.