How capacitive turbosensing reduces noise & signal error in touchscreen apps

Libor Gecnuk

June 14, 2010

Libor Gecnuk

Using CTS effectively
The TSS approach (Figure 4, below) integrates a CTS-sensor and is royalty free (and can be downloaded from the Freescale web site). Today, it supports the HCS08 MCU family of microcontrollers and applications developed with CodeWarrior Development Studio for Microcontrollers.

Figure 4. Freescale Touch Sensing Solution (TSS)

As the full TSS solution provides a lot of other useful algorithms, detectors, and electrode decoders, the low-level CTS measurement calls may be utilized directly from the application when appropriate.

In an example showing the integration of a simple CTS-sensor into custom embedded code (Figure 5, below), the CTS-sensor doesn't use an MCU module or RAM. Required flash ROM is typically 100 bytes per each GPIO pin that's used as a capacitive channel. Each channel requires just one external pull-up resistor.

Figure 5. [RCT/CTS-sensor] coding example

Running the CTS-sensor on an HCS08 device at fBus = 20 MHz results in an average sample time of about 9 μs. Integrating the CTS-sensor is straightforward and its capacitive sensing algorithm can coexist with custom applications on the same platform.

Results
Test results of the CTS-sensor behavior can be seen with different turbo-control values (Figure 6 below, left). The tested capacitive interface had various parasitic capacitances in the range of 15 to 35 pF, serial resistances in the range 20 to 60 Ω, and a glass dielectric distance of about 4 mm with an outlined electrode area of about 15 by 15 mm.

The figure shows the main difference of measurement time between a simple RCT-sensor (emulated with a CTS-sensor and its CTS_arg=0) and the real CTS-sensor (CTS_arg greater than 0) (Figure 6, right). Both have been measured on the same capacitive interface and capacitive channel, with the same environmental conditions.

Figure 6: [RCT/CTS-sensor]: results and comparison

The results show that the CTS sensor allows for a fast capacitive measurement using the turbo-control. It also shows that with a different conversion time, the sensitivity does not change. The CTS sensor provides a guaranteed conversion time for a wide range of parasitic capacitances using optimized turbo-control.

The CTS algorithm allows for significantly faster capacitance measurements and a guaranteed maximal conversion time for typical hardware conditions. The CTS method offers noise reduction, low accumulated signal error, measurement accuracy, and reduced power consumption.

It's suited for embedded capacitive sensing, which means that the sensor, the sensing software, and the custom application can all coexist in the same firmware. This approach could significantly reduce the cost of the capacitive sensing interface design.

Libor Gecnuk holds an M.Eng. degree in textile engineering from Technical University, Liberec, Czech Republic. He joined Freescale in 2004. His current research interests include embedded systems and sensors, with a focus on automotive, medical, and custom applications. He can be reached at R63174@freescale.com.