Solid State Lighting ECE 198 Lab Manual

Solid State Lighting ECE 198 Lab Manual Introduction/Overview The purpose of this lab is to investigate the performance of four different types of lig...
Author: Derrick Edwards
5 downloads 0 Views 602KB Size
Solid State Lighting ECE 198 Lab Manual Introduction/Overview The purpose of this lab is to investigate the performance of four different types of lighting. You will do this by studying four different commonly available light bulbs: halogen, incandescent, compact fluorescent, and light-emitting diode (LED). Over the course of the lab, you will perform variety of tests on each of the light bulbs in order to characterize their performance. You will first look at the power required to operate each light bulb. Next, you will look at the spectrum of the light emitted from the each light bulb, using the spectrometers you are already familiar with from the nanotechnology lab. Next you will look at the power emitted from the light bulb, by measuring the emission at fixed distances, but different angles from the bulb. Finally, you will use an IR camera to measure the temperature of the bulb by detecting the greybody emission given off by each bulb. Ultimately, the goal of the lab is to gain an understanding of the pros and cons of the various forms of lighting commonly used in our society today. In doing so, you will want to consider not only the technical properties of the bulbs that you measure in lab, but also the cost of the lightbulb, the cost to operate the lightbulb, and if you are feeling especially ambitious, other important factors such as the environmental impact of the bulbs’ manufacturing process, the scarcity of the bulb materials, and perhaps even political considerations. The final report for this lab will consist of a 2-3 page position paper. You will take the role of a consultant to either a large company, a government institution, or an academic institution. You are to imagine that the institution has hired you to do a study of the lighting they are using. Your report should advise your employer as to their future approach to lighting.

1

Lighting A significant fraction of the electricity used in this country is used for lighting applications, whether this is residential, industrial, or the lighting of public places. The ability to keep our homes, places of business, and public spaces well-lit impacts our societies’ productivity, comfort and safety, and can have a significant impact on a country’s GDP. We don’t often think about the effect light has on economic productivity or personal happiness or comfort, but it is, in fact, the case that lighting can significantly affect all of these and more. As the world population grows, and as more and more people from developing countries require sources of light, there has been a significant increase in light consumption globally. With that increase of consumption, of course, comes a parallel increase in the energy required to provide this lighting. This energy consumption is significant (and should probably be a part of your lab report), and places a strain on energy production, and contributes to the universally accepted, within the large majority of the scientific community, phenomenon of global warming.

Figure 1. (a) 1 steradian “cuts out” an area of a sphere equal to r2 (http://www.mathsisfun.com/geometry/steradian.html ). (b) Luminosity function for the human eye at moderate light levels (http://climatesanity.wordpress.com/2009/07/18/ more-on-compact-fluorescent-lights/ ). Light is typically measured in units of lumens, which is a measure of the total visible light emitted by a source. 1 lumen is equal to emission of 1 candela of light across a solid angle of 1 steradian. 1 Candela is defined as the intensity of a light source emitting 1/683 Watts of light per steradian at 550nm (in the green). So therefore, a lumen is equivalent to 1/683 Watts at 550nm. The wavelength 550nm is chose as it is close to the peak of the eye’s response to light. If the light is at some other wavelength, then the same amount of intensity is actually less than 1 candela, determined by the “luminosity function”, shown in Figure 1(b). This function accounts for the fact that the same amount

2

Figure 2. Global lighting electricity consumption 1995-2030 from 2008 scenarios. Dashed line shows expected consumption assuming no policy changes are made. From (http://www.iea.org/Textbase/npsum/lll.pdf ) of power at green and red wavelengths, will be perceived differently by the human eye. The energy consumption for lighting (or any other energy usage) is typically measured in Watt-hrs. A Watt is the measure of Power, which is energy per unit time. Thus, a Watthr is the energy consumed when a power of 1 Watt is expended for an hour. Typically, energy consumption is measured in kilo-, mega-, giga-, or tera-watt hours, depending on the scale of energy consumption. Currently, light constitutes approximately 19% of all electricity usage in the world. As can be seen in Figure 2, electricity consumption for lighting is projected to increase significantly in the coming years. Figure 3 shows the total light consumption in the United Kingdom, as an example, showing the various sources of lighting, and total Tera-lumen-hrs/year of light consumption throughout history.

As can be

clearly seen, the sources of lighting of lighting

1 2

JY Tsao, et al, J. Phys. D: Appl. Phys. 43 (2010). R Fouquet and PJG Pearson, Energy J., 27, (2006).

3

Figure 3. Lighting consumption, in Tlmh/yr, in the UK from 1700-2000, showing sources of lighting energy. From [1], adapted from [2].

used over history have changed, from candles to gas to kerosene and now, electricity. Nonetheless, the consumption trend is steadily upward.

For this reason, there is

significant interest in developing more energy-efficient lighting, in order to lower global energy consumption and mitigate global warming. However, energy-efficient lighting must be developed with certain constraints in mind. Clearly, using a candle instead of fifteen 100W incandescent light bulbs would save electricity, but it unlikely that such an approach to lighting would be generally accepted by anyone. Thus, any new lighting technology must address a number of critical specifications. Before we discuss the specifications you will be studying, lets get to know our light bulbs. Light Sources In this lab you will measure the emission spectrum, heating, power consumption, and power emission from each of four light sources. You will see that certain sources show highly blackbody-like emission, whereas other sources give off white light by combining emission of multiple different wavelengths. The different spectra from each of the bulbs results from the different light emission mechanisms used to generate the white light. As discussed in class, the incandescent light bulb uses a heated Tungsten filament (Figure 4(a)) which emits, essentially, a blackbody spectrum (Figure 4(b)). While most of the emission from such a bulb lies in the IR and UV portions of the electromagnetic spectrum (where we cannot see it), these bulb also emit significant visible light. However, when thinking about the energy efficiency of a light source, it is important to consider how much light is ‘wasted’ by being at wavelengths we cannot see!

Figure 4: (a) Schematic of incandescent light bulb and (b) Emission spectrum of incandescent bulb (300K) the sun (6000K). Grey bar shows visible wavelengths. From (a) http://electrical-engineering-portal.com/lights-up-the-facts-about-lighting , and (b) http://www.sciencebuddies.org/science-fair-projects/project_ideas/Phys_p030.shtml

4

A halogen bulb is really nothing more than an improved incandescent. The reason incandescent bulbs have such short lifetimes (750-1,000 hours) is because at very high temperatures, the Tungsten filament actually evaporates, depositing Tungsten on the walls of the bulb.

As the filament evaporates, it gets thinner, and this means the

resistance of the bulb increases, and the bulb gets hotter, until finally, the process ‘runs away’, namely more evaporation leads to more heating leads to more evaporation, etc., and the filament breaks. A halogen bulb somewhat rectifies this situation. The halogen bulb is much smaller, and contains a halogen gas within the bulb. This gas is designed such that when the Tungsten evaporates, it mixes with the gas, and then actually redeposit on the filament, essentially offsetting the evaporation process that causes traditional incandescents to die so quickly. As a result, halogens can last longer, and can operate at higher temperatures. However, the bulb itself is still very hot, and much more heat than light is generated, just as in a traditional incandescent. This is also why halogens are such fire hazards, and are often banned from dorm rooms!

Figure 5: Light emission process for a fluorescent light bulb, as well as image of a compact fluorescent bulb (from http://getleducated.com/tag/compact-fluorescent-lamp/). Fluorescent bulbs operate differently than incandescents. In a fluorescent bulb, light emission occurs in three steps, shown schematically in Figure 4. First, electrons are generated in either one of two ways: 1) by applying a voltage between two metal ‘plates’ where this voltage is strong enough to pull electrons off of one of the plates and accelerate them to the other (cold cathode emission). Alternatively, a filament (thin coil

5

of wire) is heated to temperatures high enough that electrons in the filament have enough thermal energy to escape the filament (thermionic emission). Either way, as the electrons travels through the bulb, they scatter off of mercury gas atoms, which are present in the tube, along with other gases (Xenon, Neon, Argon, etc). These electrons have enough energy to excite electrons in the mercury atoms to higher energy states. The electrons then drop down to their ground state by emission of UV photons. In a way, this process is very similar to the photoluminescence you studied in the nanotechnology lab. Of course, the UV photons are not visible to the human eye, so they need to be converted to visible light. This is done by means of phosphurs, materials which coat the inner walls of the fluorescent tube. When a UV photon strikes the phosphor, it is absorbed and excites an electron to a higher energy state. The electron in the phosphor then relaxes to the ground state by emitting lower energy (visible) light. In order to generate white light, different phosphurs with different fluorescence energies must be used. Take a second to think about the energy losses in this system, using the example of a thermionic emission process: 1) Not all of the heat energy generates electrons escaping the filament, some just goes to heat, 2) the UV photon generated by the emitted electron must, by definition have a lower energy than the electron, so there is energy lost here (usually in heat, via collisions with other atoms in the gas), also the electron can scatter off of non-Mercury atoms, which will cause energy loss, but not generate photons. 3) The UV photon is used to generate a visible photon, which is a lower energy photon, so even if every UV photon generates a visible photon (100% quantum efficiency), there will be an energy loss in this process, since the photons do not have equal energies. 4) The voltages and currents entering the bulb must be controlled carefully by a “ballast”, which prevent excess current from running through the bulb and overheating (and blowing up) the bulb. These ballasts are not 100% efficient, so there is also energy lost before you even get to the bulb itself. All told, only a fraction of the electrical energy used for a fluorescent lamp ever makes it to light (maybe 40-45%). The last type of light bulb you will be studying is the LED light. We have already discussed the operation of LEDs (light emitting diodes) in class. AS you know, LEDs are made from semiconductors, specifically, from the junction of n-type (current carried by electrons) and p-type (current carried by holes) semiconductor material. At the junction

6

Figure 6. (a) Basic bandstructure diagram for LED (b) Comparison of three different LED designs: LED+ phosphor, RGB LED, and a stacked LED. (c) Spectra for a typical LED + Phosphur and (d) for a white light LED made from three different (RGB) LEDs. Images from (a) http://www2.warwick.ac.uk/fac/sci/physics/current/postgraduate/regs/ mpags/ex5/devices/led/ , (b) http://spie.org/x24065.xml , (c) http://www.sciencedirect. com/science/article/pii/S1350946211000267, and (d) http://www.digikey.com/us/en /techzone/lighting/resources/articles/color-management-of-a-red-green-and-blue-led.html between the n- and p- type material, electrons can recombine with holes, and in doing so, can emit a photon. The photon emitted is at an energy determined by the bandgap of the semiconductor. There are a number of ways to generate white light from LEDs. The first is to group together 3 or more LEDs on a single mount. If you have a red, blue, and green LED grouped together, then you can control their relative emission to create white, or at least ‘whitish’, light (see Characterization section for details on how we define white light). Another form of LED light bulbs use phosphurs, just like our fluorescent bulbs. Here, instead of generating UV light from the collision of energetic electrons with a Mercury gas, UV light is generated by the diode, and is then absorbed by phosphurs on the LED walls, which then emit red, green, and blue light. LED light bulbs can be highly efficient and can also last for much longer than tradition incandescent light bulbs. Unlike fluorescents, LEDs do not contain Mercury, which is a hazardous material. However, LEDs are much more expensive than traditional incandescents, fluorescents, or halogen bulbs, and cannot work directly off of wall voltages. Because an LED requires a DC

7

voltage in the range of 3-5V, the AC wall power of 120V must be transformed down to a DC voltage in the 3-5V range. Now that you have a basic understanding of the operational principles of the bulbs you will be studying, we should discuss exactly what you will be looking at while you are investigating these light sources.

Ultimately, because you are trying to make a

quantitative determination about lighting use, you will need to investigate a number of factors, including power consumption (which, along with the initial bulb cost should give you an idea about the cost/time of a given bulb), light quality (you will look at the spectral emission from each bulb), light power (how much energy is radiated into the visible wavelength range), emission directionality (in what direction is the light emitted from the diode), and finally, how much energy is lost to heat? Characterization of Light Bulbs In this lab, you will carefully analyze a number of light bulb properties. When deciding on a lighting technology, or really, any technology at all, the potential consumer must consider a wide range of factors, ranging from purely economic, to societal, to personal. If you are running a business, then economics are going to be very important, obviously. Thus, it may well be that your first consideration might be price. In order for any new energy efficient technology to gain market share, it must demonstrate that it will not be more expensive, or at least not substantially more expensive, than the existing technology. However, there are two prices we must consider. There is the up-front cost to purchase the technology, and there is the cost to operate the technology. For a light bulb, these costs could be quantified as follows:

where Xi is the up-front cost for light bulb i, Pi is the power consumption of the bulb, Ri is the cost of electricity per kWatt-hr, and Ti is the lifetime of the bulb, in hours. Of course, different light bulbs can last for different amounts of time, so the best way to describe the cost is to normalize cost per unit time.

So if we compare a $1 light bulb (A) that draws 100W of power and lasts for 1,000 hours to a $30 light bulb that draws 23W and lasts for 10,000 hours, there is one factor that then

8

determines the relative cost-effectiveness of the two bulbs: the cost of electricity per kWhr. In determining the cost-effectiveness of the various bulbs, you will need to make estimates about the current and future price of electricity (this does not mean guessing, it means using sources) and you will need to determine the lifetime of the bulbs (this can usually be found on the bulb packaging). The power consumption of the light bulbs is something you will determine experimentally in this lab! But price is not the only consideration. One must also consider the quality of the light output by the bulb. Visible light is most often described used the CIE 1931 color space, which is shown in Figure 7. Here, all of the colors visible to the human eye are plotted. The plot axes are x and y, which are related, somewhat non-trivially, to the response of the light receptors in the human eye. The outer edge of the curve is the line consisting of all of the monochromatic frequencies. Any color visible to the human eye can be described by a position on the CIE 1931 plot. Given two sources with emission corresponding to two points on the CIE plot, any color on the line connecting those two points can be achieved by controlling the relative intensity of the two sources. Likewise, any 3 sources on the CIE plot correspond to three points, making a triangle, and can be used to generate any color within the bounds of the triangle. When making displays (for cell phones, TVs, computer screens, etc.) typically 3 light sources are used: red, green, and blue.

Depending on the emission of these sources, different triangles can be

achieved, but due to the irregular shape of the CIE plot, one can clearly see that any three emitters cannot cover the entire visible range! White light, on the other hand, is most usually described by a Planck temperature. If you remember, the Planck emission for a blackbody gives the thermal emission as a function of wavelength for a given Blackbody temperature. As temperature increases, the peak of the blackbody emission moves from longer wavelengths (in the infrared) to shorter wavelengths where there can be a significant amount of emission in the visible wavelength range. This is shown in Figure 7, which plots the points corresponding to blackbody emission for a range of temperatures over the CIE 1931 curve, this collection of points is known as the Planck curve. For a white light source, the emission is often quantified by the correlated color temperature (CCT), which tells one the “warmth” of a lighting source. Referring to a light source as either “warm” or “cold” can be a little

9

misleading. Be careful here, it turns out that “warm” light sources actually correspond to lower temperature Blackbody emission and “cooler” white light sources actually have more emission at shorter (blue) wavelengths, and therefore correspond to higher temperature Blackbody emission. Using temperature as a measure of white light emission works well for incandescent light bulbs, which emit light from heated Tungsten filaments, and whose emission closely follows that of a blackbody source. Other white light sources (such as LEDs or fluorescent light bulbs) have sharp emission peaks at certain frequencies. These peaks correspond to positions on the CIE 1931 plot, and by control of their relative intensity, the “warmth” of the light bulb can be controlled, and total combined emission at points along or near the Planck curve can be obtained. Often, you will hear the term color rendering index used to describe a white light source. This basically scales an artificial light source from 0-100 in terms of how closely it mimics a perfect blackbody emitter, in other words, how close it is on the CIE 1931 plot to the blackbody curve.

Figure 7: CIE 1931 plot with overlaid triangle showing range of colors achievable with three different colored light sources. In addition, the Planck blackbody curve is plotted in black showing the points on the CIE 1931 corresponding to white light emission (http://notebook.pconline.com.cn/testing/contrast/0706/1033209_3.html). The spectrum emitted by your light source will tell you something about the light quality. But you will also want to look at the efficiency of the light bulb. You already

10

know the power emitted by the bulb, you now want to be able to estimate how much of that power is going into visible light. Your spectral measurements tell you this, to a certain extent, since you can see the emission from 400-1100nm on the spectrometer. But this is still a very limited range. In order to try to understand more fully the fraction of light going into the visible, you will also take power measurements on the light emitted from the source. This will give a more direct measurement of the total visible power, and the power emitted into the non-visible portion of the 400-1100nm spectrum, it will also allow

you

to

determine

the

directionality of light emission from your source. However, a large amount of energy can also be dissipated as heat. This energy is harder to detect optically, but not impossible. We will use a thermal imaging camera, which measure the thermal (blackbody) radiation from objects in the wavelength range Figure 8. Thermal image of a home. Often, thermal imaging is used to measure heat loss from buildings in order to improve insulation and decrease home heating (or cooling) costs. From http://www.coasttocoasthomeservices.com /infrared.htm

between 7-13µm. Again, this does not give a full picture of the energy wasted for each bulb, but it can tell you about the approximate temperature of each bulb. The higher the temperature of the bulb, the more

likely it is that significant energy is going into heat, and not light. The next portion of the lab manual describes the experiments you will be undertaking in greater detail. You will have a range of characterization tools available to you in this lab, feel free to use them in any way you think might help you best characterize the performance of the light sources under investigation.

11

Lab Materials and Equipment 

    

Light bulbs (60W equivalent) o Halogen ($6.99 each) o Incandescent ($1.44 each) o Compact Fluorescent ($11.47 each) o LED (19.91 each) Lighting Fixture Avantes Educational Spectrometer Multimeter Optical power meter Infrared Camera

Tests 1. Power consumption The first test we will perform concerns how much energy the light bulbs draw. This is an important consideration when evaluating light bulbs because the amount of energy consumed has a direct correlation with the long term cost of a light bulb. The box you are using to power the light bulb has 3 sets of inputs. One goes to the wall outlet, one goes to the light bulb, and one is a series connect that will allow you to measure current. Use the multimeter to measure the current by connecting “+” (red) terminal on the meter to the “+” MEAS terminal on the box, and likewise with the “-“ (black) terminals. When the bulb is turned on, the meter will display the root mean square (RMS) current passing through the bulb, which will allow you to calculate the amount of power consumed. (Note: you will have to choose the correct multimeter setting, which will depend on the specific multimeter) In the United States, wall outlets are powered by 120V of RMS AC power. This means that the effective average power of the wall outlet is 120V, though the peak voltage is actually 172V. Since we want to know the power consumed over a long period of time, we care about average power, PAVG, which is given by VRMS multiplied with IRMS. Calculate the average power consumed by each light bulb. You can then use the power consumption to calculate the cost of running the light bulb. The amount of energy consumed is given by multiplying the power (in kW) by the time the light bulb is on (in hours). The electric company bills the user by the number of kW-hours (or kWh) consumed, at a rate of something like $0.15 per kWh. You can use this information to calculate the cost, per unit time, of running each light bulb. 2. Spectrum The next aspect of the light bulbs we would like to test is the quality of light. As we saw with the ThorLabs white light source in the first lab, the spectrum of a 12

generic white light source is a bell-shaped curve that covers the entire visible spectrum. In general, this mix of white light creates a pleasant tone because it imitates the spectrum of sunlight, which contains a broad mix of wavelengths. Since humans evolved outdoors, the most comfortable form of indoor lighting has a broad mix of wavelengths. Using the Avantes spectrometer we used in the nanotechnology lab, measure (and save) the spectrum of each light source. Which does the best job of imitating sunlight? Which is the worst? Do these conclusions match your personal preference amongst the lighting sources? Also, think about why different types of bulbs produce different spectra. 3. Power output Next, we will analyze the optical power output of the light bulbs. Along with the quality of light, the amount of light generated is important to fully understand how much bang you get for your electricity bill buck. This is the trickiest measurement to make, and will be left open to your interpretation to some extent. We will have a ThorLabs S121C power meter that will measure the optical intensity at a given position. It has a wavelength range from 400nm-1100nm. Since we only care about visible light, use the FGS900 filter, which is a 315700nm bandpass filter. This will filter out the infrared light, and allow us to measure strictly the visible light. Measure the power at various points around the light bulb in order to determine the spatial emission profile (i.e. the directions in which the bulb emits the strongest). Note that the bulb is close to radially symmetric, so you only need a vertical profile in order to get all of the requisite information about the bulb. Therefore, measure the optical power at a fixed distance while varying angle relative to the table. Estimate how much total light is output by using the detector size (9.5mm diameter), your spatial profile, and your knowledge of geometry. Try measuring power with and without the filter. What does this tell you? 4. Temperature The last factor we will look at is temperature. Generally, this will be correlated with power consumption and power output because heat is the other way power gets dissipated (besides as light). Essentially, the power that is consumed but not emitted as light is released as heat. The temperature won’t affect the quality of the light, but with hot temperatures comes risk of skin burns and fire/combustion so there is a safety aspect to consider. To measure the temperature, we will use a FLIR ThermaCAM BX320 thermal camera provided by Professor Wasserman. Please be careful, as this equipment does not belong to the class. We only have one, so you will have to coordinate with other lab groups. To use the camera, point the camera so that the “+” on the screen is on the object you want to measure. In the top-right corner, there is an approximate temperature reading for the object. Note: Light bulbs don’t get hot

13

immediately, so to get an accurate idea of how hot your light bulb gets, make sure it has been continuously on for at least 5-10 minutes.

Evaluation Factors Each light bulb will have both redeeming qualities and shortcomings. In order to properly evaluate which bulb is the best for a given situation, you must consider a variety of factors. These include, but are not limited to:       

Cost (up front and long term) Environmental Impact? Quality of light o Strength o Color Safety Performance change over time Light emission profile Fraction of total light that is visible

Questions to Ask Yourself   

What bulbs are best, and why? What factors would lead you to choose a different bulb? Which bulb is worst?

Lab Write-Up The lab report for the Solid State Lighting lab will be in the form of a position paper. Imagine you are a consultant working for either a large business, local, state, or federal government, or some other large consumer of electricity. Your boss has asked you to look into lighting costs for the institution, and to suggest potential methods for cutting costs. If you are going to write from the point of view of a government agency, you should also take into account other important issues that might not show up on a short term balance sheet, such as the geopolitical and environmental considerations associated with energy use. Even if you are writing from the point of view of a private company, you cannot ignore political considerations. Set the framework of the problem out to your employer using sources from the literature.

You must consider energy prices (present

and future), electricity sources (present and future), global demand for electricity and potential changes in the political and regulatory framework that affect your choice of

14

lighting. Lay out the problem and the costs associated with this problem clearly, and substantiate your claims using reputable sources (NOT the lab report and NOT Wikipedia). Next, use the data you have collected in this lab to make a case for the future lighting practices of your institution. Show and use data from lab, and relate this data to the framework set out in the first part of the lab report. Conclude the report by making a suggestion to your employer, with quantitative reasoning clearly substantiating your recommendation with data from the lab and data collected from the literature. The total length of the memo/position paper should be approximately 3 pages, single spaced, with figures, but if you need to go longer to make your point, you may go up to 5 pages.

15

Suggest Documents