How capacitive turbosensing reduces noise & signal error in touchscreen apps
This "Product How-To" article focuses how to use a certain product in an embedded system and is written by a company representative.
Capacitance measurement, widely used to detect proximity of human touch, creates a characteristic capacitive sensing chain like most of environment-sensitive proximity sensing systems. A key part of this sensing chain is the capacitive sensor (Figure 1, below).
This sensor is the low-layer hardware and software, which drives and monitors the charging process at the same time. Many low-cost systems use only single threshold logic with simple driving, using the constant current charge sensor or charging through a parallel resistor.
|Figure 1. Capacitive Sensing Chain|
Hardware sensors typically provide a fixed or configurable oversampling feature, where multiple subsequent measurements are taken and processed as one value (either a sum or an average). Oversampling is time consuming but offers precision. Software sensors provide discrete sample values, where oversampling may be controlled.
For this article, sensing is the “waiting” time when the driven capacitive signal reaches a known detection threshold (VIH). When oversampling is counted, the waiting time varies in the microseconds range.
Capacitance-to-time conversion (CTC) means that a capacitance added (by a finger) to the base capacitance of the measuring circuitry (Figure 2, below) increases the measured conversion time. The base capacitance of the measuring circuitry is called parasitic capacitance, which determines the minimum conversion time, called baseline.
|Figure 2 Capacitive Interface & Capacitances|
Depending on the layout, route thickness, and other board- and input pin-specific parameters, the baseline may easily reach tens or even hundreds of microseconds. The finger capacitance can extend this conversion time further.
Without suitable elimination of parasitic capacitance, this baseline can be very high and may even be higher than the finger capacitance being measured. A problem with capacitive measurement is instability and difficulty to estimate hardware parameters and conditions with almost random parasitic capacitance in the range of tens or hundreds of picofarads.
Today's systems have adjustable sensitivity through a single configurable current or voltage source. This setup always requires minimal or at least stable parasitic capacitance in the measuring chain.
These systems don't allow independent control of sensitivity and charge time for various parasitic capacitances and for different physical conditions. This means that higher sensitivity always means longer conversion time with all consequent disadvantages.
The patent-pending capacitive turbo sensing (CTS) approach (Figure 3, below) allows independently controlled sensitivity and capacitance-to-time conversion. The main advantage of CTS is the fast capacitance-to-time conversion independent of the sensitivity, board design, electrode size, dielectric distance, and other environment conditions.
|Figure 3. [RCT/CTS-sensor] HW, CTC, algorithm, function, i/o|
The CTS circuitry arrangement has two voltage or current sources where the first eliminates the hardware conditions such as parasitic capacitance by using quick charging/discharging control, and the second is used for precise measurement of the remaining capacitance.
The method speeds up this capacitance measurement in range of 1 fF to 100 pF by using general-purpose I/O (GPIO) processor pins with internal pull-up/pull-down resistors being used as second current or voltage sources. The CTS algorithm allows significantly faster measurement and guaranteed maximal conversion time for typical parasitic hardware conditions.
The hardware required by the CTS-sensor is a microcontroller‘s GPIO pin with an internal pull-up resistor, an external resistor, a suitable touch channel, and the touch interface. The pull-up resistor is used by a software turbo-control algorithm to shorten the measurement time. The measurement sensitivity is then achieved by an external resistor.
Sensitivity of the CTS-based sensor comes from the standard resistor-charge-transient (RCT) mathematics (see Equation 1 below). The microcontroller measures the time to reach the known detection threshold. The ratio (see Equation 3 below) may be used to calculate the characteristic gain factor (See Equation 2 below) of RCT.
Also, the leakage current affects the target sensitivity. The main rule is to use devices with very low leakage current (in the nanoamp range or better). Sensitivity of the hardware specifies basic sensing behavior.
dt[us] =eR[MΩ) x dC[pF] x GRIH “sensitivity”(Equation 1)
GRIH= -1n(1 – RIH) “gain factor” (Equation 2)
RIH = VIH[V]/(Vdd[V] + eR[MΩ] x IL[uA]) “Relative VIH” (Equation 3)