Incandescent miniature and subminiature lamps are used in applications requiring high brightness, a wide viewing angle, and a variety of display colors. Incandescent lamps generally provide between 6 and 50,000 mcd (millicandellas), while LEDs generally produce between 1 and 40 mcd. Neon lamps typically produce 5 to 12 mcd per milliamp.

Incandescent lamps are sensitive to voltage, shock, vibration, and cycling frequency. A 5% rise in lamp voltage above design level reduces service life by 50%. Many lamp manufacturers state that lamp life varies inversely as the twelfth power of voltage.

These manufacturers suggest using low-voltage filaments, which are short and rigid and have inherently high mechanical resonant frequencies. This helps soften shock and vibration.

Incandescent lamps are also sensitive to flashing rate, especially with high-energy bursts. High peak voltages and temperatures can damage the filament. In addition, flashing rates near the filament mechanical resonant frequency can couple thermal energy into the mechanical system and produce vibration.

At low on/off frequencies, inrush currents shorten lamp life. Inrush currents in lamps operating at room temperature can be as high as 12 times normal operating current and can last 20 to 40 msec. At high frequencies, on the other hand, light filaments cool little between flashes and resistance stays high, causing much lower inrush currents.

Filament heat capacity determines the decay rate of inrush current. The exact relationships for these values must be determined empirically by testing representative samples.

Incandescent lamp temperatures should be kept below 100°C to ensure maximum lamp life. Heat primarily conducts through filament supports and the lamp base, although some heat reaches the bulb through infrared radiation from the filament.

Miniature lamps: Miniature lamps consist of incandescent lamps, neon lamps, and halogen-cycle lamps.

Halogen-cycle lamps are often used where size and space are limited. These lamps differ from incandescent lamps primarily in the gas fill within the bulb. In an ordinary incandescent lamp, tungsten from the hot filament boils off and deposits on the inner wall of the bulb. Over a period of time, the bulb wall blackens, light output decreases, and the filament narrows and weakens. In contrast, a halogen-cycle lamp contains a halogen vapor (such as iodine) in addition to the normal gas fill. This halogen vapor combines with the boiled-off tungsten particles and continuously redeposits the tungsten back onto the filament. Thus, lamp life is increased while light intensity remains virtually constant.

The size of a halogen-cycle lamp is sometimes deceiving. A typical miniature halogen-cycle lamp rated at 62 W produces as much light as a 60-W household lamp. Yet, the miniature halogen-cycle lamp is only one-sixth the size of the household lamp.

Halogens produce high-intensity light with about 90% of the energy being infrared. However, some lamps are used where small amounts of ultraviolet light are required.

Neon lamps:Neon glow lamps combine functional indication with the aesthetics of color illumination. Glow lamps have two electrodes sealed inside a glass envelope containing a gas, typically neon. When enough voltage is applied, current flows through the gas, ionizing it and causing it to glow near the negative electrode.

Color glow lamps offer several advantages over incandescent miniature and subminiature lamps. In-line operation permits direct connection to 110/120 and 220/240 Vac or dc line voltage with appropriate resistors. Low power consumption means that a neon lamp requires only 0.2 W, compared to 1.5 W for an incandescent lamp.

Brightness can be improved by increasing current levels. Color glow lamps can be operated at up to 3 mA to double brightness. Cool operation allows the lamp to be located near heat-sensitive components without any adverse effect. Good shock and vibration resistance results from the absence of a filament. Rated average life of color glow lamps exceeds 30,000 hr.

LED indicators: LED indicators are employed when incandescent lamps give off enough heat to damage nearby equipment. In addition, LED operating life is much longer than that of incandescent lamps. LEDs never burn out under normal use, they merely grow dimmer.

LEDs are generally rated for operation at room temperature. Operating at higher temperatures can cause light intensity to drop to as low as 75% of the rated output. Continuous operation at high temperatures reduces light output at accelerated rates. However, LEDs often operate at low temperatures that would make incandescent lamps fail -- when incandescent lamps are turned off at low temperatures (around -50°C), the glass envelope can crack from rapid cooling.

LED indicators generally require a series resistor to limit current. When operating from ac sources, LEDs sometimes require an additional diode in series with the resistor to limit reverse breakdown voltage. LED breakdown voltage is typically from 3 to 6 V.

Unlike incandescent lamps, where light spreads out uniformly in all directions, LED light radiates almost entirely straight ahead. Manufacturers generally provide beam patterns for specific LED styles because light patterns vary greatly by package type. Lenses are also available for inclusion on LED packages to modify light output patterns.

Electronic bar-graph displays, arrays of LEDs molded in a common package, are widely available. Bar graphs can be powered by standard array driver ICs. The general technique is to drive arrays by converting analog, binary-coded decimal (BCD), or binary signals to decimal form. Displays can also be microprocessor or computer controlled.

Bar-graph arrays often have 10 to 15 LEDs molded in a plastic housing. These packages can be stacked to produce arrays of any length. One of the largest arrays has 101 LEDs in a single housing. Each LED in an array is behind a light scrambler. Scramblers are light pipes, composed of diffusive epoxy that provides uniform illumination at a light-emitting surface. The light-emitting surfaces are thin rectangles, each forming a section of the bar graph.

All LEDs in an array should have identical luminous intensity. Each package is marked to indicate average intensity, so that multiple packages stacked together can have similar luminous intensity.

Color is inherently uniform for red and high-efficiency LEDs. Thus, arrays using these colors need not be color matched. However, people are often sensitive to slight color variations in yellow LEDs. Therefore, yellow bar-graph packages are marked to indicate the dominant wavelength emitted.

To drive bar graphs from analog signals, LEDs are generally powered by standard analog bar-graph array drivers. These ICs are nothing more than analog-to-digital (a/d) decoders with sufficient output to drive LEDs. They can be fed directly from most analog transducers without intervening amplifiers or other signal conditioning. Analog array drivers have 5 to 16 outputs and are cascaded to accommodate large LED arrays.

Reference voltages for analog bar-graph drivers typically are preset with 50 to 200-mV steps of equal value. Driver outputs are linear with respect to input voltage.

Reference voltages for some drivers, however, increase logarithmically, typically in 2 or 3-dB steps. A decoder with ten 2-dB steps, for example, having a first-step output in response to a 25-mV signal, produces a full display with a 2-V input. This is a 0 to 18-dB scale.

Improved LEDs: At one time, watch and calculator LEDs were made of GaAsP. Often referred to as standard red, these devices emit light at a wavelength of 655 nm. Although GaAsP is still used in low-end displays, today's technology offers displays that are considerably brighter. In the mid 1970s, so-called high-efficiency devices (GaAsP:N) were commercially introduced. A mixture of GaP, rather than GaAs is used for the substrate. One reason efficiency is improved is because GaP, unlike GaAs, is transparent. Color is determined by the ratio of As to P and doping of the epitaxial layer. Available colors are red (635 nm), yellow (583 nm), and yellow-green (566 nm).

In the early 1980s, manufacturers began offering devices doped with a liquid-phase epitaxial process. The advantage is that the resultant LEDs have a higher efficiency than those doped with a conventional vapor-phase process. When used with a GaP substrate, an extremely bright device results.

When these LEDs are coupled with an antireflection, circular-polarized filter, they provide sufficient luminous contrast to meet military requirements for sunlight readability. In 10,000 fc ambient light, a luminous contrast ratio better than 2.0 is possible. The displays are available in yellow-green (566 nm) and emerald green (555 nm). LEDs offering the highest efficiency use AlGaAs and a transparent substrate. These devices typically produce infrared wavelengths in the 800-nm region.

Although blue LEDs have been available for several years, many of the manufacturing bugs have only recently been ironed out. Several semiconductor materials were initially investigated, including ZnS and ZnSe. The problem with these materials is that they produce high-resistance p-n junctions. Therefore, they need relatively high operating voltage.

Several LED makers have turned to SiC to circumvent this problem. Although SiC has a bandgap that is less efficient for light generation than the other materials, it is easily doped. The result is an LED that emits blue light (480 nm) with an average luminous intensity of 6.0 mcd.