Graphical explanation of electromagnetic interference problem of touch screen

Developing mobile handsets with touchscreen human-machine interfaces is a complex design challenge, especially for projected capacitive touchscreen designs, which represent the mainstream technology for multi-touch interfaces today. Projected capacitive touchscreens are able to pinpoint the location of a finger touching the screen by measuring small changes in capacitance. A key design consideration in such touchscreen applications is the impact of electromagnetic interference (EMI) on system performance. Performance degradation caused by interference can have an adverse effect on touchscreen designs, and this article will explore and analyze these sources of interference.

Projected capacitive touch screen structure

A typical projected capacitive sensor is mounted under a glass or plastic cover. Figure 1 shows a simplified side view of a two-layer sensor. The transmit (Tx) and receive (Rx) electrodes are connected to transparent indium tin oxide (ITO) to form a crossbar matrix, and each Tx-Rx junction has a characteristic capacitance. The Tx ITO is located below the Rx ITO, separated by a polymer film or optical adhesive (OCA). As shown in the figure, the direction of the Tx electrode is from left to right, and the direction of the Rx electrode is from outside the paper to inside the paper.

Figure 1: Sensor structure reference.

Sensor Working Principle

Let's analyze the operation of a touch screen without considering interference factors for the moment: The operator's finger is nominally at ground potential. Rx is held at ground potential by the touch screen controller circuit, while the Tx voltage is variable. The varying Tx voltage causes current to flow through the Tx-Rx capacitor. A carefully balanced Rx integrated circuit isolates and measures the charge entering Rx, and the measured charge represents the "mutual capacitance" connecting Tx and Rx.

Sensor status: Not touched

Figure 2 shows a schematic diagram of magnetic field lines in an untouched state. Without a finger touching the surface, the Tx-Rx magnetic field lines occupy a considerable space inside the cover. The edge magnetic field lines are projected outside the electrode structure, hence the term "projected capacitance".

Figure 2: Magnetic field lines in an untouched state.

Sensor Status: Touch

When a finger touches the cover, magnetic lines of force are formed between the Tx and the finger, which replace a large amount of the Tx-Rx fringe magnetic field, as shown in Figure 3. In this way, the finger touch reduces the Tx-Rx mutual capacitance. The charge measurement circuit recognizes the changed capacitance (△C) and detects the finger above the Tx-Rx node. By measuring △C at all intersections of the Tx-Rx matrix, the touch distribution map of the entire panel can be obtained.

Figure 3 also shows another important effect: capacitive coupling between the finger and the Rx electrode. Electrical interference can couple to the Rx through this path. Some degree of finger-Rx coupling is unavoidable.

Figure 3: Magnetic field lines under touch.

Special terms

Interference in projected capacitive touch screens is generated by coupling through parasitic paths that are not easily perceived. The term "ground" is often used to refer to both the reference node of a DC circuit and the low impedance connection to earth: these are not the same terms. In fact, for portable touch screen devices, this difference is the root cause of touch coupling interference. To clarify and avoid confusion, we use the following terminology to evaluate touch screen interference.

Earth: Connected to the earth, for example, through the ground wire of a 3-pin AC power outlet.

Distributed Earth: The capacitive connection of an object to the earth.

DC Ground: The DC reference node for portable equipment.

DC Power: The battery voltage of a portable device, or the output voltage of a charger connected to a portable device, such as the 5V Vbus in a USB charger.

DC VCC (Direct Current VCC Power Supply): A regulated voltage that powers portable device electronics, including the LCD and touch screen controller.

Neutral: AC power circuit (nominal at ground potential).

Hot: The AC power supply voltage, with electrical energy applied relative to the neutral line.

LCD Vcom is coupled to the touch screen receiving line

Portable device touch screens can be mounted directly onto LCD displays. In a typical LCD architecture, the liquid crystal material is biased by transparent upper and lower electrodes. The multiple electrodes below determine the multiple individual pixels of the display; the common electrode above is a continuous plane covering the entire visible front of the display, which is biased at the voltage Vcom. In a typical low-voltage portable device (such as a mobile phone), the AC Vcom voltage is a square wave that oscillates back and forth between DC ground and 3.3V. The AC Vcom level is usually switched once per display line, so the resulting AC Vcom frequency is 1/2 of the product of the display frame refresh rate and the number of lines. A typical portable device may have an AC Vcom frequency of 15kHz. Figure 4 is a schematic diagram of the LCD Vcom voltage coupled to the touch screen.

Figure 4: LCD Vcom interference coupling model. The dual-layer touch screen consists of a separate ITO layer with a Tx array and an Rx array separated by a dielectric layer. The Tx lines occupy the entire width of the Tx array pitch, and the lines are separated by only the minimum spacing required for manufacturing. This architecture is called self-shielding because the Tx array shields the Rx array from the LCD Vcom. However, coupling can still occur through the gaps between the Tx strips.

To reduce the architectural cost and obtain better transparency, single-layer touch screens mount the Tx and Rx arrays on a single ITO layer and connect each array in turn through separate bridges. Therefore, the Tx array cannot form a shielding layer between the LCD Vcom plane and the sensor Rx electrode. This may cause severe Vcom interference coupling.

Charger interference

Another potential source of touch screen interference is the switching power supply of the mains powered mobile phone charger. Interference is coupled to the touch screen through the finger, as shown in Figure 5. Small mobile phone chargers usually have AC power live and neutral inputs, but no ground connection. The charger is safety isolated, so there is no DC connection between the power input and the charger secondary coil. However, this still produces capacitive coupling through the switching power isolation transformer. The charger interference forms a return path through the finger touching the screen.

Figure 5: Charger interference coupling model.

Note: In this context, charger interference refers to the voltage applied to the device relative to ground. This interference may be described as "common mode" interference because it is equal to the DC power supply and DC ground. Power switching noise generated between the DC power supply output of the charger and DC ground may affect the normal operation of the touch screen if it is not adequately filtered. This power supply rejection ratio (PSRR) issue is another issue and is not discussed in this article.

Charger coupling impedance

The charger switching interference is generated by coupling through the transformer primary-secondary winding leakage capacitance (about 20pF). This weak capacitive coupling effect can be compensated by the parasitic shunt capacitance that appears in the charger cable and the powered device itself relative to the distributed ground. When the device is picked up, the shunt capacitance will increase, which is usually enough to eliminate the charger switching interference and avoid interference affecting touch operation. A worst-case interference generated by the charger will occur when the portable device is connected to the charger and placed on the desktop, and the operator's finger is only in contact with the touch screen.

Charger switch interference component

A typical mobile phone charger uses a flyback circuit topology. The interference waveform generated by this charger is complex and varies greatly from charger to charger, depending on the circuit details and output voltage control strategy. The interference amplitude also varies greatly, depending on the design effort and unit cost invested by the manufacturer in the switching transformer shielding. Typical parameters include:

Waveform: includes complex pulse width modulated square wave and LC ringing waveform. Frequency: 40~150kHz under rated load, when the load is very light, the pulse frequency or skip cycle operation drops to below 2kHz. Voltage: can reach half of the peak voltage of the power supply = Vrms/√2.

Charger power supply interference component

In the charger front end, the AC supply voltage is rectified to generate the charger high voltage rail. In this way, the switching voltage component of the charger is superimposed on a sine wave of half the supply voltage. Similar to the switching interference, this supply voltage is also coupled through the switching isolation transformer. At 50Hz or 60Hz, the frequency of this component is much lower than the switching frequency, so its effective coupling impedance is correspondingly higher. The severity of the supply voltage interference depends on the characteristics of the shunt impedance to ground and also on the sensitivity of the touch screen controller to low frequencies.

Figure 6: Example of charger waveform.

Special case of power supply interference: 3-pin plug without grounding

A power adapter with a higher power rating (e.g., a laptop AC adapter) may have a 3-prong AC power plug. To suppress EMI at the output, the charger may internally connect the ground pin of the main power supply to the DC ground of the output. Such chargers typically connect Y capacitors between the live and neutral wires and ground to suppress conducted EMI from the power lines. Assuming that the ground connection is intentional, such adapters will not cause interference to the powered PC and USB-connected portable touchscreen device. The dashed box in Figure 5 illustrates this configuration.

For a PC and its USB-connected portable touchscreen device, a special case of charger interference occurs if a PC charger with a 3-pin power input is plugged into a power outlet without a ground connection. Y capacitors couple the AC power to the DC ground output. Relatively large Y capacitor values ​​couple the power supply voltage very effectively, which allows large power frequency voltages to couple through a finger on the touchscreen with relatively low impedance.

Conclusion

Projected capacitive touch screens, which are widely used in portable devices today, are susceptible to electromagnetic interference, where interference voltages from the inside or outside are coupled to the touch screen device through capacitance. These interference voltages can cause charge movement within the touch screen, which can confuse the measurement of charge movement when a finger touches the screen. Therefore, the effective design and optimization of touch screen systems depends on the understanding of the interference coupling path and how to reduce or compensate for it as much as possible.

The interference coupling paths involve parasitic effects such as transformer winding capacitance and finger-to-device capacitance. Proper modeling of these effects provides a full understanding of the source and magnitude of the interference.

For many portable devices, the battery charger is the main source of interference to the touch screen. When the operator's finger touches the touch screen, the generated capacitance allows the charger interference coupling circuit to be turned off. The quality of the charger's internal shielding design and whether there is a proper charger grounding design are key factors affecting the charger interference coupling.