Electronics such as tablet and smartphones are increasingly moving toward a bezel-less design. Bezel is a term used to describe the outside frame of a computer, monitor or any other computing device. The bezel serves to hold the screen in place, and perhaps also reduce unintentional inputs on a touchscreen. Newer device designs, such the iPad, have extremely narrow bezels, allowing for more screen. This has become an increasingly important design factor for consumers.
Unfortunately, every pleasure has its price, and zero-bezel is no exception. Multiple issues arise from this feature, both constructional and operational. Some of these problems can often be solved with software design. For example, when the iPad mini was introduced, its thin bezel led Apple to adapt its software to ignore inputs when a user is simply resting a finger on the screen. This adjustment was made to cancel touch event detection when the user just holds the phone .
However, some problems cannot be fixed via software and must be catered via hardware. Touchscreen hardware design has become highly challenging during the past decade because less and less space is available. Inarguably the most vulnerable part of these devices is the touch sensor.
The hardware design of a touch sensor is crucial to the system’s overall success. This step of the process will determine how difficult is to get a working, robust application. Despite the fact that manufacturers seem to take this into account, it is quite common practice to base their decisions on certain rules-of-thumb. Those guidelines, based on practice rather than theory, are broadly accepted, yet not always accurate.
However, experience has shown that guidelines are not a cure-all. In order to make confident decisions, you have to be 100% sure of what you are doing, and to do so, you have to examine each hardware design separately. This document’s purpose is to provide some technical details on bezel-less products hardware design and outline the importance of several design factors, specifically the traces.
Touchscreens consist of a sensor and a controller. Typically, sensors are made up of two transparent electrode grids, usually transmitting Y, and receiving, X. When a finger comes close to the screen’s surface, the electrodes are “excited,” translating the touch event to an electric signal. The controller takes this signal to acquire information about the user input.
What are those traces?
Traces (or tracking) are conductive busses that connect the sensor’s electrode grids to the controller pins. In order to save valuable space and material, traces are not actual wires (as one would expect). Instead, they are “printed” on substrates (forming PCB-like patterns) and then glued over the LCD.
The following figure depicts a typical smartphone bezel on the left, and a touch screen sensor with its tracking on the right. The traces begin from the electrodes’ edges, circle around the sensor and end on the bottom of the screen, where the controller is located.
Figure 1. Typical smartphone bezel (left) and touch sensor with tracking (right).
Since traces and electrodes are connected, the sensor’s overall “electrical state” is influenced by the traces also. This “state” (more scientifically, the “capacitance”) depends solely on the geometry of the conductive bodies under examination. High values of capacitance mean either wide surface areas or small distance between the conductors (or both reasons).
Although traces surface is very small (compared to the electrodes), when the distance between traces gets really small (as in most cases), capacitance value skyrockets. Narrow tracking (due to zero bezel construction) can, therefore, impose great variances in mutual capacitance between two electrodes. The proximity of the traces forms capacitive couples, and the addition of their capacitance the sensor’s overall state is called traces coupling.
In mutual p-cap touchscreens, when a finger touches the screen, the controller “sees” a decrease in “mutual capacitance” of two electrodes: one X and one Y. “Mutual capacitance” could be described as the “electrical state” between two exclusive objects. Based on this point of view, each time a touch event takes place, only two electrodes and two traces are participating.
What happens if there is strong coupling between these two traces? The following example exhibits this situation. Two cases are examined: the finger touches the screen a. on the corner and b. on an interior point of the screen.
Figure 2. Left: Corner node (X1-Y1) touch event. Right: Inner node (X2-Y2). Node is the overlapping section of the electrodes. The participating traces are denoted with red color. Inactive electrodes are black, and the pointer is green.
In case the user touches the screen’s corner, the participating traces are next to each other: the coupling is strong. The controller cannot perceive the finger’s presence: the sensor is insensitive in this area. In case the user touches an interior point of the screen, the participating traces are not directly next to each other: the coupling is weak. The controller “sees” the finger: the sensor is sensitive in this area.
And there’s more
These undesired effects are enhanced by traces shielding. Most often p-cap touch sensors are shielded, using a thin grounded metal layer underneath the sensor (figure 3) and/or by using grounded traces that run around the transmitting and receiving electrodes (figure 4). The reason for this construction is to reduce EM noise (similar to coaxial cable shielding). This results in more charge building up on the tracking surface, leading in even higher total capacitance. This coupling effect is more important between adjacent traces, i.e., of a corner node.
Figure 3. Tracking cross section showing trace ground capacitance (left) and trace fringe capacitance (right). GND can be either grounded traces or traces of neighboring electrodes. Wider traces, as well as a more crowded tracking (less space between them), results in a stronger coupling .
Sensor tracking varies greatly in shape and size. Each company uses its own tracking patterns, just like in electrode design. As a result, defining which parameters are considered as tracking specifications can be quite tricky. To name a few of this parameters:
- traces width
- spacing between traces of the same group (X and Y)
- spacing between trace groups (X and Y)
- distance between the electrodes and the trace closer to the sensor
- traces total length
- controller pin distance
- grounded trace surface and routing
- traces thickness
- traces parallel routing length
- And so on
Moreover, the importance of each specification varies from application to application. For example, traces spacing may be of the utmost importance in self – capacitive sensors, but rather insignificant in double layer mutual – capacitive sensors.
A full investigation of traces influence on p-cap operation exceeds this document’s purpose. The goal was to outline that, in the situation of traces, rules-of-thumb do not help! Each pattern is unique and needs special treatment.
Since prototyping is money consuming, simulation is the only way to decide on the hardware design. However, simulation is not a “cure-all.” Traditional simulation software may be tried and tested, but they also have some issues. Specifically, in the case of traces investigation, where multiple simulations must be run with small adjustments each time, this process can be extremely time-consuming. Moreover, their cost may be comparable to that of prototyping.
Touchscreen simulation tool developed by Fieldscale, Sense, helps designers to evaluate several sensor designs regarding their performance for all possible tracking specifications. By utilizing Sense’s Import Tool, any geometry can be imported and tested. Based on simulation results, the specifications of the tracking can be adjusted, to result in the desired capacitance variations, leading to optimal performance.