Fluorescent tubes

First attempts to produce light using electric discharge were made at the end of 19th century and both Tesla and Edison were involved.
In 1901, Peter Cooper Hewitt obtained a patent for his low pressure discharge mercury vapour tube.

There has been enormous progress in the technology of phosphors since the first fluorescent tubes appeared in nineteen-thirties. The development of colour TV and the discoveries of lanthanides, or rare earth elements - not much know before then, had a great effect on this.

There are about 10 different basic types of phosphors. The desired spectrum of a lamp can be obtained by mixing these powders in the right quantities. A CRI of about 90 can be achieved in this way.

Using special phosphors and techniques, partly patented and partly kept secret by their producers, a CRI of up to 98 can be achieved.

Phosphors are powder mixtures of complex salts of elements such as (for example): aluminium, antimony, barium, calcium, cerium, europium, gadolinium, germanium, magnesium, strontium, terbium, tin, yttrium and chlorine, fluorine and phosphorus.

Thicker layers of phosphors are sometimes used to reduce the peak light of the mercury discharge to further improve colour rendering, for the sake of a bit of efficiency.

A great invention in this field was the “hot spot/cool spot” arrangement. The temperature difference works to prevent the migration of the mercury atoms caused by direct electric current, keeping the lamp in a good shape even when operated on steady or pulsating DC (Direct Current).

The smaller the diameter of the tube, the better is the efficiency, and less material is needed to build a lamp. Between the 30's and 80's, T12 (1 1/2") tubes dominated the market. The 80's brought T8 (1") to the mainstream and T5 (5/8") is the current trend today.

Incandescent bulbs

The development of the incandescent bulbs has come a long way since 1802, when Humphry Davy first lit a platinum wire by passing an electric current through it. These bulbs lit the path to the industrial development and we are currently witnessing their dusk.

The abilities of classic incandescent bulbs to resemble sunlight are limited. Their Colour Rendering Index is ultimate: 100. But the trouble is the CCT, the colour temperature, which is only about 2900 K. We see that as a warm yellow light.

This limitation comes from the melting point of tungsten, the metal from which the filaments are produced, which is 3695 K.

The filament of an ordinary bulb operates at a temperature of about 2900-3000 K. Increasing the temperature leads to a dramatic shortening of the bulbs life. For example, projector bulbs with filaments operating at 3300 K are rated for just several tens of hours of service.

A revolutionary approach here is using a filter to reduce the excess yellow light seen. A suitable material to do this has been discovered: neodymium oxide, Nd2O3. This powder is mixed into the glass of the bulb, giving it a blue-violet tint. The unique properties of the NdIII+ are also used enhance contrast in astronomy optics and, most prominently, in infrared lasers.

The filter absorbs part of the light energy at wavelengths around the colour yellow - which we see predominantly in the traditional bulb. The result is a "less yellow", whiter light that can be very well combined with fluorescent lamps.

The Nd2O3 glass fluoresces at near infrared (~1300 nm). Part of the yellow light is converted to heat. The filter absorbs 20-25% of the total luminous flux produced by the filament.

The yellow colour is a transitional stage between red and green. As the Nd filter reduces yellow it increases the colour contrast between red and green.

As the filtered bulb is a "coloured" light source (off-Planckian locus), the CRI evaluated by the classical formula only reads about 75-80, which is pretty misleading. A new formula called CQS is being worked on for evaluating the colour quality of LED diodes that would follow the viewer's impression of the given light more closely than the CRI formula currently in use.

LED – light emitting diodes

History of these light sources dates back to 1907 when H. J. Round, PA to G. Marconi, observed electroluminescence on crystals of silicon carbide. O. V. Losev described this phenomenon thoroughly in 1927. K. Lehovec provided deep theoretical explanation in 1951.

Several infrared-emitting semiconductors were created in 50s. In 1962, N. Holonyak presented the first red light emitting diode. Yellow LED followed soon and the efficiency was growing. LED price dropped dramatically in the beginning of 70s so they were used as indicators and displays of portable calculators. Green LED followed after a while, but the blue one was invented only in 1989. First white LED appeared soon after this discovery. Inventions made in 90s improve the efficiency of LEDs dramatically, thanks to multi-stage, multi-layer or resonant structures. The power of LEDs grown from miliwatts to units of watts and their efficiency has already beaten fluorescent tubes and HIDs.

LEDs can be found as indicators in most of the electronic appliances, they backlit the displays of cellular phones, hand torches and are even used as a source of light in some TVs and data projectors. LEDs are new to lighting industry and they can be used in exterior, interior, architectural and transport lighting.

LED is a semiconductor source of light. Layers of silicon-based materials emit light as electric current passes through them. The spectrum of the emitted light is a bell-shaped in a relatively narrow band around a dominant wavelength. So LEDs are primarily sources of colour light.

White light can be achieved by combining several colours, blue LED with yellow phosphor being the most commonly used. The blue light partly passes the phosphor layer and partly gets converted to the yellow light. The resulting light looks white. LEDs of warm to daylight colour can be obtained, however, the Rais only about 75 to 80. Special phosphors can be used to push Rato about 90 for sake of efficiency. Another approach is to use a (ultra)violet LED and a mixture of colour phosphors. Raof 97 can be achieved and high lm/W are expected. Yet another approach is mixing the white from several colour LEDs. Using the most common RGB, Raof about 70 can be achieved. Using newer components with four colours, RGBA (A for Amber), Raaround 90 can be reached. Ra of 97 has been reached with RGBAW (W for White). The colour components can be driven independently which offers a wide range of applications. LED/Light/Engin units with an RGBW matrix, cooler and driver introduce a complete solution for creating a comfortable light with variable parameters, dimming, chromatic temperature changes, mood, decorative or biodynamic light.

The advantage of LEDs over other light sources is their great efficiency and a potential to increase it even more. The fluorescent and HIDs have reached their maximum around 100 lm/W while LEDs are already at 150 lm/W and it’s expected that more than 220 lm/W can be reached. Another advantage is longevity: LEDs live 3 to 5 times longer than fluorescent tubes. LED’s start-up time is much shorter than that of incandescent bulb, which makes LEDs great for car brake lights. Fast switching (PWM) can be used to control the average luminous flux of LEDs from almost zero to 100% with tiny losses.

The basic drawback of LED is high initial cost, which is not yet balanced by their longevity and efficiency. The LED way mostly equals new luminaires too. There are some retrofit lamps with LEDs on the market, but it’s often just a halfway solution. A paradoxical disadvantage is their potential efficiency in combination with a long lifetime. Some investors rather choose a cheaper traditional solution and wait. The longevity of LEDs is often overrated and the expected 30,000 to 100,000 hours can be achieved with expensive cooling system. A quaint con can be their high efficiency too – LED traffic lights freeze and get covered with ice while the old incandescent still work.

LEDs are often criticized for low colour rendering. Tests show that people judged some LED lights better than a plain Racomparison would suggest. It is possible that wider use of LEDs will require a new metric for quality of colours that would reflect the subjective feeling from a given light better than Ra, which does not work very well for LEDs.