One of the biggest challenges in designing any kind of embedded system is determining the optimal user interface. Simple push buttons and knobs that are reasonably intuitive? Or a scaled-down version of a PC GUI interface with a high-tech look and increased bandwidth for options and graphics? Too often the decision is driven more by the inclination toward more “bells and whistles,” and less by the actual needs of the design. As a result, the designer must try to balance the marketing desire for a flashy interface against the functional needs of the design, as well as the associated production costs.

Fig. 1. Example of LCD graphics user interface.

When making such decisions, there are a number of available and widely known user-interface options for designers to consider, as well as some more recent developments.

Interfaces by definition involve the two-way flow of information. Displays, indicators, and graphics provide information to the user on machine status and available selection options. Input devices, ranging from touch sensors to traditional electromechanical components, convey the user’s intentions to the machine. Both sides of the interface equation have to be carefully considered to make the optimal selection.

Display systems typically fall into a few broad categories, usually based on their basic technology. Examples include LEDs and LCDs. These systems have their pros and cons, but, overall, they are generally capable of producing similar types of displays. The exact format of the displays varies, but the main categories are single indicators, segmented displays, and graphics modules.

LED displays involve time-proven circuits that are microcontroller-friendly. Most microcontrollers (MCUs) utilize general purpose input/output drivers with sufficient current capabilities to drive LEDs directly with only a simple current-limiting resistor because LED displays require only a few mA of drive. LEDs are also available in a wide variety of both individual indicators and segmented displays. However, due to their accumulated heat dissipation, they are generally precluded from larger graphic modules.

LCDs utilize a liquid crystal fluid to either block or pass light through the display. Using this capability, LCDs can be manufactured as anything from indicators to full graphic panels. Segmented and indicator LCDs require low amounts of energy to switch the liquid crystal, making them MCU friendly. Many small MCUs include a peripheral that is capable of driving indicator and segmented displays with few external components.

Larger graphic modules that typically include the necessary drive electronics are available. These modules offer full-color capabilities through a combination of multiple pixels and color filters. The only two drawbacks to LCDs are that they don’t emit light directly, but instead rely on an external light source or a backlight to provide the necessary illumination. LCDs also have a relatively limited temperature range.

Fig. 2. Cross section of resistive touch screen.

Some MCU manufacturers such as Microchip Technology provide complete graphic function libraries to assist in the development of graphic applications. Fig. 1 shows an example of a user interface that demonstrates several types of displays and user inputs. This example makes use of the 16-bit PIC24F MCU to handle all of the inputs and outputs of this reference design.

User input options, much like displays, fall into different primary categories: buttons and knobs, or touch screens. Buttons and knobs enjoy the same simplicity and intuitive use of individual indicators and segmented displays. Touch screens are the equivalent of graphic displays in that they allow the dynamic reallocation of input controls via a graphic display showing different functions. Buttons and knobs can be simple and intuitive. The use of a simple knob selection control for a menu system works well for a handheld PDA and the scroll wheel on a PC mouse.

The danger with a button-and-knob interface is the tendency to overload the user with many functions. In general, it is a good idea to limit the number and complexity of the commands used with a button-and-knob setup, unless the interface is specifically designed to handle the complexity of mice and keyboards.

Touch screens are an alternative to buttons and knobs. In these systems, the entire surface of the display is a user input. These systems allow the designer to redefine the user input on-the-fly by creating multiple, simpler button-and-knob interfaces. The only downside to this system is the greater complexity of the touch-screen interface, and possibly the requirement to scroll through multiple menus to find a specific control.

Most switch and field-effect user-input technologies apply to both button-and-knob and touch-screen systems. Take, for example, the simplest system, one using electro-mechanical contacts. In button-and-knob systems, this involves switches, rotary encoders, and potentiometers. These components bring with them challenges such as mechanical wear and tear, mounting difficulties, and difficulties associated with sealing the interface against dust and moisture. Manufacturers have made great strides in reliability, mounting, and sealing, but the problem of wear and tear remains on any system with moving parts.

Enlarge this picture

Electro-mechanical technology is also used in resistive touch screens. Here, the press from a user pushes two plastic layers together, creating an electrical connection that can be read using current drivers and analog-to-digital Converter (ADC) channels. Fig. 2 shows a cross section of a resistive touch screen. This is a simple system, but it is subject to wear and tear, and it can require system calibration, filtering, or linearization to compensate for physical variances. Electro-mechanical contact systems rate high for ease of interface to a MCU, even though a simple algorithm to debounce the contacts is still a necessity.

Capacitive touch is another possible user-interface option. This technology takes advantage of the basic construction of a capacitor — two conductors separated by an insulator. When this field is applied to the iron in the person’s blood, the system capacitively couples every surface of the body with every other surface, including the tips of the fingers with the soles of the feet. Capacitive touch works by measuring the capacitance change caused by a finger touching the cover over a conductive plate. This increase in capacitance is measured and compared against the unpressed capacitance of the plate. Fig. 3 shows the model for capacitive touch. If a sufficient shift in capacitance occurs, then the sensor is considered touched and the corresponding functions are activated. The primary requirements for the system are a sensor plate, an insulating covering such as glass or plastic, and some means of measuring the capacitance with sufficient resolution. Several MCUs have built-in capacitive sense modules.

For the button and knobs format, it is simply a matter of creating the appropriate sensor pads, capacitance-measurement circuitry, and the software to drive it. Some MCU manufacturers supply reference designs and supporting development tools to make this job easy. The only challenge is developing an appropriate averaging algorithm to determine the untouched capacitance of the sensor.

Capacitive touch is also quite prevalent with touch screens, with two main forms currently available — surface-capacitive technology and projected-capacitance technology.

Surface-capacitive technology uses the finite resistance of an indium-tin-oxide layer on the back of the sensor. When the user touches the sensor, the user forms a capacitance to ground that AC shorts the sensor at the point of the touch. The interface then determines the touch by measuring the current drawn by the sensor when each edge of the sensor is driven by an AC waveform. The relative currents are then used to calculate the distance from each edge to the user’s touch. Because the sense signal is in the MHz range, the design is often left to companies that specialize in the technology.

Appliances with inductive touch-sensing user interfaces.

Projected-capacitive touch technology works by creating two layers of touch sensors, one a series of horizontal stripes, and the other a series of vertical stripes. The interface electronics interrogate the system in order to determine which horizontal and vertical stripes are closest to the user’s press and interpolate a final position. Projected-capacitive systems are generally simpler than surface-capacitive methods, but projected-capacitive systems must scan many more inputs to detect a touch, which tends to increase system overhead and slows its response.

Many kitchen appliances utilize finishes that consist of metal, plastic, and glass. Sometimes appliance designers use a combination of these materials to provide the sleek look that appliance consumers want. Stainless steel in particular has become a popular finish seen on many appliances. However, with stainless steel finishes, appliance manufacturers must still use stainless-steel looking plastic or even glass plates to implement the capacitive-touch control. Technology and manufacturing costs can impact the overall feel of the products with this approach.

A solution to this problem can be found in the use of inductive touch sensing technology. Inductive touch sensing enables appliance designers to utilize metal such as stainless steel or aluminum as the main surface of the product. The technology works by sensing slight deflections in the metal when it is pressed. (See Fig. 4.) The sensing is accomplished by mounting a specifically designed circuit board behind the metal, where the buttons are to be placed. A thin spacer layer is used to create a small gap to enable the metal to be slightly deflected. The amount of deflection required is very small, as shown in Fig. 5. The circuit board can also be the main control board for other functions, given that processor requirements and usage are low for this technology.

Enlarge this picture Inductive touch mechanical model.

In addition to providing an aesthetically pleasing user interface, inductive touch-sensing technology works even when liquids are on the surface of the push points. For example, the technology is not triggered by water or oil on the buttons, and it will even operate under water. This is an ideal feature for applications that will encounter liquids, such as a horizontal cook top or commercial appliances subject to washdowns. Even cleaning the buttons will not trigger them as long as the pressure on each button does not deflect the metal. Microchip provides proprietary technology that enables designers to quickly and easily implement inductive touch-sensing applications using its PIC MCUs.

Inductive-touch technology works well with a metal finish, but can also be employed with a dark non-metal finish, such as plastic. In designs where the fascia presents a non-metal surface, inductive touch can be implemented by inserting a small metal target layer behind the surface. A target layer is required to change the value of the inductor by mutual inductance. When the front fascia deflects, it causes the target layer to deflect, as well. The material for the target layer can be copper, aluminum, or another material that is magnetically permeable or electrically conductive. The thickness of the target layer will depend on the frequency used to drive the sensor. This provides the same result as if stainless steel was used. Because a metal target layer that is on one side of the board is measured, the back side of the PCB can be filled with copper to provide an EMC shield that helps with noise immunity and emission.

Designers can select from a wide spectrum of appliance display and user-input technologies. Options for feedback to the end user range from simple LED displays all the way to information rich graphical displays. And input technologies such as capacitive touch sensing or inductive touch sensing can enhance the look and feel of appliance user interfaces. All of these technologies are driven by embedded processors.

For more information, visit:
www.microchip.com/appliance
www.microchip.com/humaninterface