Let's compare lighting effectiveness of 3 bulb types:
- Fluorescent;
- Incandescent (Tungsten);
- Light Emitting Diodes (LED).
Technical introduction:
- Tungsten bulb lits by overheating of metal wire. The wire is heated by electric current. For long life, the air in the lamp is replaced with argon. Since heating is main process in Incandescent lamp, it's considered only 5% of energy is spent for lighting. the rest is the heat.
- Fluorescent lamp is one of few gas-discharge lamps. Visible light comes from hit of invisible ultraviolet photon into luminophore. Different luminophores produce different colors of emitting light: cold white, warm white, blue, red, green, etc. What is a source of ultraviolet photons (UVP)? In common words (avoiding entering interesting physical processes), UVP is created by collision of accelerated electron and atom of inert gas. Electron is accelerated thanks to high voltage on electrodes in the tube ends. Of course, by collision not only UVP are produced. Other wavelength photons (i.e. "colors") are born in dependency of tube filled gas. For instance, Neon gives red color. Mercury vapors added into the tube allow significant reducing of discharge voltage. To make the lamp smaller, the tube is sometimes bent into U-shape, Spiral, circle, etc. These bent bulbs are called Compacf Fluorescent Lamps (CFL).
- LEDs are based on ability of special semiconductors to create (emit) photons when the current is applied to them. The color of light depends on chemical mix of the semiconductor. "Natural" LEDs give red, green, blue, and orange light (orange is a mix of red and green materials). White LEDs, actually, aren't white. Their semiconductor is covered by luminophore. These LEDs emit UVP which are transformed into "white light" by luminophore.
I haven't illuminometer. So I tried to compare bulbs by stills camera.
It's known that picture brightness is defined by:
- Light sensibility of film or sensor (ISO);
- Exposure duration (parts of second);
- Aperture opening (relative number).
Adding "focus distance" and "required picture brightness" we can compare bulb illumination by one of parameters above. I chose compare the illumication by exposure. Other conditions are the same for all types of lamps and all types of shots.
Why do I compare three types of bulbs?
* I don't like tungsten bulbs by 3 reasons:
1. Their light is yellowish. IMHO, it's weak light, sad, depressing...
2. They heat. It's unpleasant to feel lamp heat on one side of body and cold environment on other side. I am sensible to uneven heat.
3. They waste the energy. Although their price is very low.
* I don't like CFL by 3 reasons:
1. Their illumination activity depends on environment temperature.
2. Maximum light is expected only 10-15 minutes after power on: when the gas inside the tube will hot enough.
3. Although claimed MTBF more than 5000 hours, their electronic circuit usually passes away earlier. No problem to replace burnt CFL in the store. Just it's annoying to go to store every 2-3 months and hoping stupid administrator will allow (or not) to change the bulb.
* Light emitting diodes - LEDs... This is the youngest lighting industry. It's developing these days. Development of LEDs is concentrated on effectiveness of luminophore and on cooling methods. Well, I am ready to donate an effort for this knowledge area. :) So:
1. LED has concentrated light beam. Optics required for dissipation.
2. LED heating kills the semiconductor crystal and luminophore. I.e. older LEDs emit less light than new ones with the same power applied.
To be fair, both LED and CFL contains electronic circuit which is the weak point of them. In CFL it's electronic ballast for lamp start. In LED bulb it's power supply unit. Bad circuit planning/build quality reduces the bulb reliability (MTBF). Therefore, in terms of the most unreliable part, the primitive tungsten bulb is more reliable than CFL and LED.
I use more than 30 CFL. Lamps, which work 2-6 hours daily, burn their ballast circuit within 2-6 months. Other lamps, which turned on and off more times but staying lit less time, work already few years.
Among 10 LED spots I use every day only one exploded its power supply unit. It was beautiful :) I replaced PSU ($2.5) and the LED works until now. It's a time to move bigger lamps ;)
Comparison results in the table:
Lamp/bulb
Power
75 W
20 W
15 W
13 W
10 W
6 W
Lamp view
Lampshade view
Illumination intensity within 1 meter
Relative illumination per Watt*
1
3
27 (inadequate evaluation due to beam direction)
5
4
5
Comments
Yellowish light, hot.
Non-reliable for long-time use.
Industrial design (ugly). Too concentrated light beam.
Longer than lampshade, bluish light.
Lampshade is is half-shaded.
Almost perfect view, weak light, significant delay when turning on (about 1 sec).
* How did I calculate Relative illumination per Watt?:
I suppose the most objective parameter for illumination comparison is exposure for photograph shot of object within 1 meter distance from the bulb. I excluded of considerations the shots of bulb itself and lampshade because:
- Lamps' construction is too different;
- Lampshade has aesthetic effect only;
- We use bulb not to see its light but to see around it.
15W lamp has too directed (non dissipated enough) light beam. Therefore, calculated illumination value can't be considered as objective.
The formula for Relative illumination per Watt is:
Illumination = 10 / Exposure / Power.
10 is a coefficient for comfort reading/perception :)
13W and 6W shot are underexposured relatively to other shots. So I artificially "increased" their exposure for calculation. If I will calculate "as is" they would absolute champions for illumination. :)
Conclusions:
1. 20W CFL illuminates a bit weaker than 75W incandescent bulb. I.e. CFL is greener than tungsten only 3 times, not 5 time as every CFL manufacturer claims.
2. 13W LED corn illuminates as well as 75W incandescent bulb. And it's stronger than 20W CFL. Looks like LEDs passed green level of CFLs. Although in some reference books LEDs are defined as less effective for lighting, and in other handbooks vice versa.
Meanwhile I run LED bulbs for about month. I will publish long-run result after year of run :)