Capacitive makes sense in gaming

01 December 2006

Capacitive sensing is commonly used in PC touchpads and portable media players for users to play games. Mobile phone manufacturers are also capitalising on its popularity.

Simplified construction, device impermeability and mechanical robustness are capacitive sensing interfaces qualities.

A capacitive sensor is little more than a copper pad on a PCB connected to a controller circuit. The sensing pad and the connecting trace assembly has capacitance to its surroundings. A ground plane, metal bracing, and other electrical and mechanical components influence the capacitance of a sensor. Most often, the capacitance is thought of as capacitance-to-ground. When a conductive activating element, such as a finger, is brought into close-enough proximity to the sensor, the capacitance-to-ground increases. A conductive object provides more possible paths between the sensor and ground. More paths lead to more field lines. More field lines lead to greater capacitance.

The front-end of capacitive sensors consists of switched capacitors, internal current sources, or voltage sources with external resistors. The voltage on the capacitor is put into an ADC (analogue to digital converter) or a charging time measurement circuit consisting of a comparator, and then to a counter or timer. A digital output is used for the dataprocessing and decision-making of a capacitive sensing system. Changes in ADC output or counts are analogous to changes in capacitance.

Implementing a capacitive sensor in a real design is not difficult. A capacitive sensor is merely a conductive pad, usually copper, placed on a PCB. This pad is connected to a controlling device and interacts directly with the activating element, usually a finger. The sensor plate is mounted to the overlay surface directly under the desired sensitive area. It is best to remove all air from between the sensor and the overlay by attaching the sensor substrate to the overlay with a non-conductive material. The controlling circuit is ideally located near the sensor, although it can be some distance away. The mechanical design requirements govern the placement of the controlling circuit. Increasing the distance of the sensor from its controlling circuit increases the native capacitance of the sensor to ground as the trace interacts with its surroundings and adds capacitance. At long lengths, this added capacitance can be significant. Six to 12 inches may be considered a functional maximum.

Substrate choices

The substrates for capacitive sensing applications are not fixed. The most common design is a basic FR4 PCB with copper traces. Flexible PCBs, often on polyimide film, Kapton, with copper are also quite common. Flexible substrates allow for easier mechanical design, especially when working with curved surfaces. Conductive ink, such as carbon or silver, printed onto a flexible material allows for ultra-low-cost construction of the capacitive sensor but requires a control PCB and connector as the flexible material cannot accept solder.

Transparent conductive materials such as indium tin-oxide (ITO) are gaining acceptance in touchscreen applications. ITO sensors are printed on glass or polyethylene terephthalate (PET) film and assembled on the final design. While chip-on-glass is available for controlling such applications, it may be more economical to use a flexible connector or hot bar soldering to a PCB.

There are several types of capacitive sensors. The most basic is a button, a single pad connected to the control circuit. A button has analogue capabilities but its primary output is a digital ‘on’ or ‘off’. The size of the button determines its sensitivity. As a general rule, larger buttons have greater sensitivity. The limit is the size of the finger. A small finger only interacts with a portion of a very large sensor. For standard 7mm diameter buttons, 10mm overlay thicknesses are quite reachable.

Sliders are linear or radial arrays of capacitive sensors. Capacitance changes are measured for the whole array and used to interpolate position to a resolution greater than that of the sensor. Interpolation uses the analogue capabilities of each sensor. A finger interacts with a sub-section of slider sensors. In the centre of the interaction, the capacitance change is high, towards the edges the charge is small. Basic data-processing calculates the centre of the activation from the change values. The resolution of a slider is limited by the capacitance change and the algorithm itself.

The same rule applies to buttons. Larger sensors are more sensitive. It is important that an activating element changes capacitance on multiple sensors. The shape of the sensors can increase the changes of multiple capacitive interactions. A maximum overlay thickness of 4mm is a good reference.

Sensor types

Proximity sensors are basically very large buttons. They use larger sensors to detect larger activating elements at greater distances. A finger can be detected at 10mm of overlay with a button, a hand at 150mm. Proximity sensors do not have the precision of a button or a slider or a touchpad. They also require longer scan times to detect conductive objects that are further away.

The successive approximation method (patents applied for by Cypress Semiconductor) implemented with the PSoC device uses a capacitance to voltage converter and single slope ADC. The capacitance measurement is achieved by converting the capacitance to a voltage, storing it on a capacitor, and measuring the stored voltage using an adjustable current source.

The capacitance to voltage converter is implemented with switched capacitor technology. The circuitry brings the sensor capacitor to a voltage relative to the capacitance of the sensor. The switched capacitor is clocked by the PSoC’s internal main oscillator.

In the first step, the sensor capacitor Cx is connected to an analogue mux bus. The sensor capacitance and the bus capacitance are connected in parallel, sharing their capacitance. A programmable iDAC (current (i) Digital to Analogue Converter) is connected to the analogue mux bus to charge the capacitors. The voltage across the capacitors is equal but the charge is different as defined by:

q = CV

In step two switch SW2 is opened and switch SW1 is closed, bringing the voltage on Cx to zero and reducing the charge on the bus by the value contained in Cx. The analogue mux bus capacitance is, still charged by the iDAC.

These steps are repeated so that the switched capacitor of Cx becomes a current load with a value related to its capacitance:

Ix = foscCxV

While the switched capacitor circuit is running, the iDAC calibration is completed. The iDAC uses a binary search to determine the value at which the voltage on the bus (with Cx attached) remains constant. This yields a voltage that is drawn away from the bus by the sensor of:

Vx = 1

foscCx IDAC

VBUS = VREF – Vx

The successive approximation method
The successive approximation method

The resulting bus voltage is based on the switching frequency, the sensor capacitance, and the iDAC current. CBUS acts as a bypass capacitor that stabilises the resulting voltage. Capacitance can be added to the bus to increase the stability further. External capacitors affect the performance and timing requirements.

With the iDAC value determined for a sensor with no finger present, the sensor capacitor is again connected to the bus by SW2. The iDAC charges the bus and the time required to take the capacitance from an initial voltage to the comparator threshold is measured. Time is measured using a 16bit timer and the internal main oscillator. The new equation for voltage relative to the capacitor is given by:

VP+F = 1

fosc(Cp + CF) IDAC

VBUS = VREF – VP+F

Where CP is the native capacitance of the sensor and CF is the capacitance added to the sensor by proximity of the activating element (the finger).

When a larger sensor capacitor is connected to the bus (by placing a finger on the sensor) the voltage drop on the bus is greater and the voltage at which the measurement charge begins is lower. As a constant current source is used, the time necessary to reach the threshold voltage is greater. The difference in counts between a sensor with and without a finger present is used for decision making.

Sensor activation

Sensor activation status is determined using baseline values and a series of thresholds to provide intelligence to a capacitive sensor. The upper and lower limits of the activation region can create hysteresis to sensor activation. To activate a sensor, it is necessary to surpass the higher threshold. The sensor stays active until the count value falls below the lower threshold.

A third threshold can help to eliminate ambient noise in the system and lessen the impact of a finger on the baseline recalibration. It is important to ignore small changes in counts (capacitance). A noise threshold is used for this. It is can also be used to deactivate the sensor’s baseline recalibration when a finger is on the sensor. If a finger is on the sensor and the sensor recalibration is active, the finger has a disproportionately large effect.

RYAN SEGUINE is product manager, Cypress Semiconductor.


Contact Details and Archive...

Print this page | E-mail this page