The challenge of this project is to create a model of a capacitive touch screen, including physical modeling of a user’s finger, the touch screen materials, and the touch-detection circuitry. A capacitive touch screen works by measuring the change in capacitance between two conductors caused by a nearby finger. In the touch screen, two transparent conductors, which run in perpendicular directions, are separated by the screen. These conductors act as a parallel-plate capacitor where they overlap; the capacitance is determined by numerous physical parameters. For example, the conductor width, length, resistivity, and separation, along with the screen material properties and thickness, all affect the coupling between the conductors.
To operate a single pixel of the touch screen, a particular frequency is transmitted along one conductor and received by the other. The capacitive coupling between the conductors causes a particular phase shift in the received signal. When a finger is close enough to where the conductors overlap, it changes the capacitance between the conductors because the finger appears to be grounded. This causes an additional phase shift, which is detected by a simple demodulator circuit and converted into a DC signal. When the DC signal exceeds a given threshold, a “touch” is detected by the pixel. This process occurs for all of the touch screen pixels to determine which locations are being touched.
The traditional approach to modeling a touch screen uses a purely circuit-based design that uses components whose values must be precomputed in another environment. Changes in the design parameters would require the recalculation and re-entry of these and other values into the circuit model before subsequent simulations; this process makes experimentation and fine-tuning of the model difficult and time-consuming. For example, the coupling capacitance between the transmitting and receiving conductors would have to be calculated and fed into the circuit simulation ahead of time. Similarly, the effect of a grounded finger would have to be incorporated as some time-dependent capacitance whose time-domain waveform is predetermined and loaded into the simulator. Also, the demodulator, which is essentially the multiplication of two sine waves, would have to be implemented using a full demodulator circuit as opposed to simply multiplying two signals.
Contact Maplesoft to learn how MapleSim can be used in your projects.