Due to the large number of questions I have received regarding LEDs I have decided to publish this on our blog. I originally wrote this article to be published but I decided it’s contain would be better served here so that everyone can have access to it.
With the digital lighting age now upon us, it fills me with great pride to see our hobby embrace green technologies like LED lighting. After 15 years of giving lectures, writing articles, and educating our industry about the benefits of LED lighting, I finally get to tell my colleagues and friends “I told you so.” But seriously after dedicating so much of my life to the continuing development of LED’s, I feel a personal responsibility that the implementation of this technology is done so responsibly.
The purpose of this post is to give hobbyist an in depth look at LED’s, and the physical science that enables them to produce light. As with any new technology there is and will be a lot of rumor, speculation, and erroneous claims about the use of LED’s. My hope is to enlighten readers about the varying quality, and performance of LED’s so that they can evaluate LED lighting products intended for aquarium use.
What is an LED, and how does it produce light?
An LED or light emitting diode is a visible diode. Like conventional diodes, LED’s allow the flow of electrons in one direction, and oppose their flow in the opposite direction. When an electrical current is passed through an LED and its on voltage is exceeded it emits light. As the electrical current flows through an LED electrons are forced to recombine with holes in the semi conductor material. As an electron falls into one of these holes it is stripped of some of its energy. This energy is released in the form of photons. The electromagnetic radiation or visible light emitted during this process is referred to as electroluminescence.
An LED is basically a miniature molecular light engine that runs on electrons. The color of light produced when current passes through an LED corresponds to the energy of the photons, which is determined by the band gap of the semiconductor material used in its construction. By varying the inorganic material (organic material in the case of OLED’s), and subsequently the band gap of the semiconductor material used in the construction of an LED, various wavelengths of light can be achieved. This determines the dominant wavelength of an LED.
One of the common semiconductor materials used to produce various wavelengths of light in the violet and blue region (400-490nm) is Indium gallium nitride (InGan). This semiconductor material is widely used in the production of Royal blue and Blue LED’s which have become so popular in our hobby due to the amazing fluorescence they can produce in a reef aquarium. They are also commonly used in the production of aquarium moonlights. Ingan chips are also commonly used in the production of white LED’s which we will be discussing a little later on.
To better understand LED’s and how they operate, let’s begin by discussing how they are made. A typical LED is constructed from a chip or die of semiconductor material. These dies or chips are produced from larger pieces of semiconductor material called wafers. Wafers are made by taking n-type or negatively charged semiconductor material referred to as the substrate and layering it with p-type or positively charged semiconductor material. The layers are doped (impregnated) with impurities to create a p-n junction and determines the band gap. The operating temperature of this junction is referred to as the junction temperature of an LED. In some cases p-type material is used as a substrate, these are referred to as p-type LED’s.
An n-type LED is constructed by placing the LED die n-side down into a small reflector cup built into a lead frame. Lead frames are the legs you see protruding from common thru hole LED’s. This first lead frame not only acts as a mount for the LED die but also serves as the negative electrical connection known as the cathode. Via wire bond which is so small it is barely detectable to the human eye the positive or anode lead frame is connected to the top of the LED die (p-side). This entire assembly is then encapsulated in epoxy to protect the LED and help distribute the light emitted during operation.
While the construction of high power LED’s is considerably more complex, their basic assembly, and the principles on which they operate are the same. In an n-type LED, electrons flow through the n-type material in the die and forced to jump across the band gap in order to recombine with holes in the p-type material producing light. It is pretty obvious that the output and performance of an LED relies on the quality and construction of the LED die itself.
One problem faced by hobbyist is that it is often difficult to determine the performance of LED’s in products they purchase because their performance lies heavily in the quality of the LED dies used in their construction. To compound matters, even if an LED die comes from a reputable manufacturer, it does not mean that its performance is absolute. Thousands of dies can be produced from a single wafer. Due to unavoidable anomalies in the production of a wafer it can have numerous flux, color, and forward voltage bins. The desirability of these bins varies by application and relies primarily on the importance of color consistency and output.
Typically a lighting designer or manufacturer like myself will specify the exact flux, dominant wavelength, and forward voltage required for a given lighting application. We will then purchase LED’s from very specific bins. This binning process is how the luminous flux (output), color (dominant and wavelength), and forward voltage (performance) of LED dies is categorized and sorted. The less desirable bins are typically used in less demanding applications where variances in light output are not as critical or even noticeable. These variances are usually not discernible to the human eye and can even escape detection by most light meters.
A key factor in any LED’s performance especially in the case of high power LED’s is its packaging which is the physical housing the LED chip or die is mounted in. There are two primary factors to the packaging of LED’s. The first is its optical design and determines the pattern of light an LED will produce. This will also determine an LED’s angle of radiation which represents its ability to spread light over a given area.
For aquarium lighting I prefer the use of LED’s with a wide angle of radiation as they work well with specialized reflectors that can enhance an LED’s output. Some LED’s incorporate the use of secondary optics which collimates (focuses) the light produced by an LED. This will typically produce a spot or flood effect depending on the optic used. The optic design can also affect the amount of light emitted from an LED versus the amount of light produced or absorbed by an LED package.
Since most of the materials used for the production of LED dies have high refractive indices (including optics), a certain amount of light produced will be reflected back into the LED at the air surface interface or absorbed by the LED itself. The second important factor regarding the packaging of an LED, and probably the most important in terms of reliability, color consistency and life expectancy, is its ability to dissipate heat. This is referred to as its thermal design. The ability of an LED package to dissipate heat during its operation is considered its thermal resistance. This rule also applies to heat sinks.
The higher the efficiency of an LED package (or heat sink) to disperse heat generated by the LED die, the lower its thermal resistance expressed in degrees Celsius per watt (C/W). The thermal threshold at which an LED die will start to fail is referred to as its maximum junction temperature. IF maximum junction temperature of an LED is exceeded it will cause permanent damage to the LED and result in premature failure. Now before you start raiding your ice box to save this poor LED, certain high power LED’s can have maximum junction temperatures exceeding 100 degrees Celsius.
This is why high power LED’s incorporate the use of metal slugs and or aluminum or ceramic housings to transfer the heat generated by the LED die to help maintain its junction temperature. They are then mounted to a metal core printed circuit board which helps transfer the heat from the LED to a heat sink, aluminum extrusion or other heat conducting surface.
An efficiently designed thermal package can not only help maintain the junction temperature of an LED, it can also help increase the amount of current the device can effectively handle. But before you start hot wiring your LED power supplies or drivers there is a catch. Depending on the size of the LED die which determines the maximum amount of current a device can handle, increasing current can reduce its efficiency resulting in less light output. The effect is known as droop and occurs when increased current produces more heat than light output.
Lower power LED’s that use their lead frames to transfer and conduct heat from the LED die are especially susceptible to increased drive currents. Precise current control is integral to any LED’s performance. Even if an LED is operated at its recommended drive current, improperly designed heat sinks, and poorly designed fixture housing can all rob an LED of performance. If an LED fixture incorporates the use of cooling fans, be sure to check them regularly. If a cooling fan fails this can cause a fixture to overheat causing permanent damage to the LED’s. While an LED fixture system that incorporates the use of cooling fans should have some type of thermal protection to shut the system down in the event a cooling fan stops working, it still means down time for the fixture until the fixture can be serviced. This is why most of our LED fixture systems use passive cooling to manage heat removal. Passive cooling (fan-less) uses the surrounding air as a medium to transfer heat from the fixture. A properly designed LED fixture can operate for over 25,000-50,000 hours so it is not uncommon for components like cooling fans or power supplies to fail at some point. It is for this reason many LED lighting systems for aquarium use come with remote power supplies so that they can easily be serviced or replaced.
While LED lighting can offer infinite control-ability options, I recommend hobbyist be practical in their purchase of an LED system. The more options or controls a system offers the more likely that system will fail at some point. For this reason I recommend that the control-ability of LED lighting systems is best achieved by means of external controllers so your lighting system can be easily serviced in the event of a problem. I would like to point out that the quality of light produced from an LED fixture is far more important than its ability to turn your aquarium into a discotheque. A simple appliance timer or an on/off switch still works just fine.
A common question I often hear regarding LED light systems is “what is the wattage of the LED’s”. For decades (the dark ages) the aquarium industry has used radiometric power (watts) to compare the light output of aquarium lighting products. Radiometric power refers to the total amount of radiant energy produced by a light source. Depending on the light source only a given portion of this radiant energy is in the visible region where photosynthesis occurs. Typically a good portion of this radiant energy is in the form of ultraviolet, infrared and heat.
In the case of LED’s, while all of the light they produce is usually in the visible region a good deal of the radiant energy they produce is in the form of heat or absorbed by the LED itself as we discussed earlier. So, while wattage is used to indicate the radiometric power of an LED, the amount of light produced in photo-metric terms, and the amount of power it consumes (wattage) are two completely different things. Consider an automotive engine for example. Its performance is gauged by the amount of horsepower or foot pounds of torque it can produce at certain RPM, not its displacement (cubic inches). The wattage of an LED indicates the amount of power it can handle and is not indicative of the amount of light it will produce.
Actually the higher the wattage of an LED, the less efficient it is at producing light. AS we discussed earlier, when power exceeds the amount of heat a devise can effectively handle you lose output and not the other way around. You also considerably decrease the life expectancy and lumen maintenance of the device. For this reason you are actually better off using multiple lower wattage LED’s. You will get a greater light spread, and depending on the efficiency of the devices used you will get more overall light output.
So please keep in mind the visible light output of LED’s is expressed in luminous flux or lumens. Their performance is gauged by their luminous efficacy not by their wattage. Luminous efficacy is the quotient of total luminous flux emitted by the total power consumed (lm/w). By comparing luminous efficacy not only can we compare the luminous efficiency of LED’s, we can compare any light source in simple terms of lumens per watt. Sounds simple enough, but there is a catch. In terms of reef aquarium lighting, or planted aquaria use, this still does not give us a complete picture of the amount of usable light produced.
While a light source may appear to produce a lot of light it can still lack the necessary amount of photosynthetically active radiation (par) required to meet the spectral requirements of photosynthetic corals, invertebrates and plants. One of the major benefits of LED’s is that they can be used to target specific wavelengths of light known to be beneficial to photosynthetic organisms. With the use of LED’s not only can you color tune the light produced by your existing lighting, you can supplement any spectral deficiencies it may have as well.
LED’s make an excellent choice for supplemental actinic lighting, exciting fluorescence in corals and invertebrates, and mimicking moon light for night time viewing. Because they so closely resemble a point source of light (although LED’s by definition are technically not true point sources of light) the rippling effect they can create in an aquarium is absolutely breathtaking.
White LED systems can be a great choice for general illumination or full spectrum use. I have left the discussion of white LED’s for last because this is where things get a little tricky. There are multiple ways to produce white light with the use of LED’s. One is to use individual red, green, and blue LED’s. The light they emit is blended together with the use of an RGB controller to produce white light. A second way to produce white light with LED’s is with the use of di, tri or tetrachromatic LED’s commonly referred to as multi-die LED’s. Their name is pretty self explanatory. Multiple individual color dies are incorporated into a single LED package.
The problem with these two approaches is that while the light emitted may appear white, it can have poor color rendering washing out certain colors in an aquarium. The biggest problem with these two approaches is that they lack spectral bandwidth and therefore the spectral power distribution required by photosynthetic organisms. The best method in producing full spectrum white light for aquarium use is with phosphor based LED’s. They produce white light by exciting a phosphor coating that is applied to a blue LED die. By employing the use of several phosphors their spectral bandwidth is significantly increased improving both spectral power distribution and color rendering.
Hobbyist may find it interesting to know that white phosphor based LED’s operate on the same principle as fluorescent lamps in that they use shorter wave lengths of light to excite phosphors in order to produce longer wavelengths of light. While LED’s offer a promising future for aquarium lighting, fluorescent, and metal halide technology, is improving as well in terms of energy efficiency and light output. Combined with the recent emergence of Plasma lighting it will be exciting to see what the future holds for aquarium lighting technologies.
The post comparing the Radiometric output of a light source with its Photo-metric output is completed and will be posted shortly. I would like to give a special thanks to Julian Sprung for his help in editing this piece. I find it fitting considering he was one of the first people to offer encouragement when I first started my lighting career many years ago. I would also like to thank Dana Riddle for his help in reviewing this material.