Design of an LED Spot Light System with a Projection Distance of 10 Km

crystals

Article Design of an LED Spot Light System with a Projection Distance of 10 km

Chi-Shou Wu 1, Kuan-Yu Chen 1, Xuan-Hao Lee 1,*, Shih-Kang Lin 1, Ching-Cherng Sun 1,2, Jhih-You Cai 1, Tsung-Hsun Yang 1 and Yeh-Wei Yu 1

1 Department of Optics and Photonics, National Central University, Chung-Li 32001, Taiwan; [email protected] (C.-S.W.); [email protected] (K.-Y.C.); [email protected] (S.-K.L.); [email protected] (C.-C.S.); [email protected] (J.-Y.C.); [email protected] (T.-H.Y.); [email protected] (Y.-W.Y.) 2 Department of Electrophysics, National Chiao Tung University, Hsin-Chu 30010, Taiwan * Correspondence: [email protected] or [email protected]; Tel.: +886-3-422-7151

Received: 9 September 2019; Accepted: 12 October 2019; Published: 13 October 2019

Abstract: We designed a spot light system with an illumination range of 10 km. In the designed system, an appropriate white light-emitting diode (LED) was selected according to the exitance and injection power required. Subsequently, through a first-order optical design, the geometry of the lens and reflector was determined using geometrical calculation. Because the central illuminance of the projection spot of the reflector was 2.5 times that of the cover lens, we first considered the fabrication error of the reflector. According to the adjustment of the optimized distance between the white LED and reflector, we modified the design of the cover lens to fit the new location of the white LED. An LED spot light module containing 16 spot light units was used. The module’s power injection was only 68.2 W. Because of the excellent performance of the designed system in terms of the divergence angle of the projection beam and maximum luminous intensity, which were 1.6◦ and 2,840,000 cd, respectively, the projection distance of the LED spot light module was 3.37 Km, according to the ANSI regulation. Finally, a spot light system with nine modules and capable of achieving a projection distance of 10 km was successfully fabricated.

Keywords: white LED; long-distance projection; exitance; spot light

1. Introduction White light-emitting diodes (LEDs) have been extensively applied in general lighting and special lighting [1–5], such as stadium [6], museum [7], and workplace lighting [8], thanks to their superior characteristics, including long life, compact size, vivid color, fast response, and environmental benefits [9–16]. A white LED is mostly fabricated by covering a blue die with yellow phosphor; such a white LED is also called a phosphor-converted white LED (pcW-LED) [17]. The die refers to the LED chip, which is the core device of LED illumination. Yellow phosphor can transform blue light into yellow light and balance the leakage blue light to provide white light. Therefore, achieving a stable transformation from blue light to yellow light and fixing the color coordinate of the targeted white light are important tasks. However, fixing the color coordinate is difficult because the heat generated by the blue die is transferred to the phosphor layer and thus causes thermal quenching, which reduces the quantum efficiency of the phosphor layer and changes the color coordinate and color temperature [18]. Therefore, the thermal problem is critical in LED lighting, especially when the driving power is large. Traditional plasma-type light sources are applied in projection lamps, including automotive headlamps and long-distance projection lamps, due to their large exitance [19–29]. By contrast, the exitance of a pcW-LED is limited because the power density of a pcW-LED is strictly limited.

Crystals 2019, 9, 524; doi:10.3390/cryst9100524 www.mdpi.com/journal/crystals Crystals 2018, 8, x FOR PEER REVIEW 2 of 11 thermoelectricCrystals 2019 cooler, 9, 524 to control the board temperature at 100 °C; moreover, the flux was2 of 12measured using pulse driving current. We found that compared with the other pcW-LEDs, Osram HWQP had a larger Althoughexitance higher and powerhigher density current can beinjection achieved density under a special at a power design for density heat dissipation, of 5 W/mm the system2 [30]. Thus, Osram HWQPcost increases. was determined In this study, to we be measured more suitable the exitance than and the injection other pcW-LEDs current density for designing of six types ofa compact long-distancecommercial spot pcW-LEDs.light system. The aforementionedThis paper presents measurements the design were conducted details of using the aspot thermoelectric light system and cooler to control the board temperature at 100 ◦C; moreover, the flux was measured using pulse driving results obtainedcurrent. We through found thatan experimental compared with theevaluation other pcW-LEDs, of the system. Osram HWQP We used had the a larger ANSI exitance regulation to calculateand the higher beam current distance, injection defin densityed as at athe power distance density offrom 5 W /themm 2lu[30minaire]. Thus, Osramat which HWQP the was measured illuminancedetermined is equivalent to be more to suitable0.25 lux than [31]. the Finally, other pcW-LEDs we evaluated for designing the final a compact version long-distance of the LED spot projection system andlight verified system. Thisthat paperthe projection presents the distance design details could of reach the spot 10 km. light system and results obtained through an experimental evaluation of the system. We used the ANSI regulation to calculate the beam 2. Opticaldistance, Modeling defined as the distance from the luminaire at which the measured illuminance is equivalent to 0.25 lux [31]. Finally, we evaluated the final version of the LED projection system and verified that Duethe to projection its superior distance exitance could reachand 10power km. density, Osram HWQP was selected as the light source of the 10-km2. Optical projection Modeling lamp. The pcW-LED of Osram HWQP is displayed in Figure 1, in which the yellow region is the active emitting area. The lateral dimensions of the pcW-LED of Osram HWQP Due to its superior exitance and power density, Osram HWQP was selected as the light source were measuredof the 10-km to be projection 1 mm × lamp. 1 mm. The To pcW-LED achieve of Osrama 10 km HWQP projection is displayed distance, in Figure the1, divergence in which the angle of the lampyellow must region be extremely is the active small, emitting and area. the The optical lateral model dimensions should of the be pcW-LED sufficiently of Osram precise. HWQP To build a precise lightwere measuredsource model, to be 1 mmwe must1 mm. obtain To achieve accura a 10tekm light projection source distance, dimensions. the divergence There exists angle ofan empty × area at thethe upper lamp must corner. be extremely The datasheet small, and of theOsram optical HWQP model shouldindicates be su thatfficiently its angular precise. Tofull build width a at half maximumprecise (FWHM) light source is 120° model, [30], we which must obtaincoincides accurate with light the source feature dimensions. of a Lambertian There exists light an empty source. This area at the upper corner. The datasheet of Osram HWQP indicates that its angular full width at property can thus facilitate the modeling process. The yellow part in Figure 1 represents the half maximum (FWHM) is 120◦ [30], which coincides with the feature of a Lambertian light source. LambertianThis emitting property can area. thus To facilitate ensure the the modeling accuracy process. of theThe light yellow source part model in Figure for1 representspractical application, the we mustLambertian measure emittingthe light area. patterns To ensure in the the accuracymidfield of regime the light [32–34]. source model Figure for practical2 displays application, the simulation algorithmwe and must the measure corresponding the light patterns experimental in the midfield setu regimep; in the [32 measurement,–34]. Figure2 displays the pcW-LED the simulation was located at the centeralgorithm of the and rotational the corresponding axis of experimentalthe rotational setup; stage. in the By measurement, controlling the the pcW-LED iris size, was we located could control at the center of the rotational axis of the rotational stage. By controlling the iris size, we could control the angularthe angular resolution resolution in the in themeasurement. measurement. FigureFigure3 illustrates 3 illustrates a comparison a comparison of the simulation of the simulation and the and the measurementmeasurement results results inin thethe midfield midfield regime. regime. Nearly Nearly all normalized all normalized cross-correlation cross-correlation (NCC) values (NCC) values werewere higher higher thanthan 99.9% 99.9% in bothin both the horizontalthe horizontal and vertical and vertical directions directions [35]. This [35]. indicates This that indicates the that the OsramOsram HWQP HWQP optical optical model model isis su suffifficientlyciently precise precise in our in design. our design.

Figure 1. Geometry of Osram HWQP LED. Figure 1. Geometry of Osram HWQP LED.

Crystals 2018, 8, x FOR PEER REVIEW 2 of 11 thermoelectric cooler to control the board temperature at 100 °C; moreover, the flux was measured using pulse driving current. We found that compared with the other pcW-LEDs, Osram HWQP had a larger exitance and higher current injection density at a power density of 5 W/mm2 [30]. Thus, Osram HWQP was determined to be more suitable than the other pcW-LEDs for designing a compact long-distance spot light system. This paper presents the design details of the spot light system and results obtained through an experimental evaluation of the system. We used the ANSI regulation to calculate the beam distance, defined as the distance from the luminaire at which the measured illuminance is equivalent to 0.25 lux [31]. Finally, we evaluated the final version of the LED projection system and verified that the projection distance could reach 10 km.

2. Optical Modeling Due to its superior exitance and power density, Osram HWQP was selected as the light source of the 10-km projection lamp. The pcW-LED of Osram HWQP is displayed in Figure 1, in which the yellow region is the active emitting area. The lateral dimensions of the pcW-LED of Osram HWQP were measured to be 1 mm × 1 mm. To achieve a 10 km projection distance, the divergence angle of the lamp must be extremely small, and the optical model should be sufficiently precise. To build a precise light source model, we must obtain accurate light source dimensions. There exists an empty area at the upper corner. The datasheet of Osram HWQP indicates that its angular full width at half maximum (FWHM) is 120° [30], which coincides with the feature of a Lambertian light source. This property can thus facilitate the modeling process. The yellow part in Figure 1 represents the Lambertian emitting area. To ensure the accuracy of the light source model for practical application, we must measure the light patterns in the midfield regime [32–34]. Figure 2 displays the simulation algorithm and the corresponding experimental setup; in the measurement, the pcW-LED was located at the center of the rotational axis of the rotational stage. By controlling the iris size, we could control the angular resolution in the measurement. Figure 3 illustrates a comparison of the simulation and the measurement results in the midfield regime. Nearly all normalized cross-correlation (NCC) values were higher than 99.9% in both the horizontal and vertical directions [35]. This indicates that the Osram HWQP optical model is sufficiently precise in our design.

Crystals 2019, 9, 524 Figure 1. Geometry of Osram HWQP LED. 3 of 12

Crystals 2018, 8, x FOR PEER REVIEW 3 of 11

Figure 2. (a) Algorithm used in the simulation along the (b) horizontal and vertical directions. (c) Figure 2. (a) Algorithm used in the simulation along the (b) horizontal and vertical directions. (c) OpticalFigure 2.setup (a) Algorithm in the experimental used in the measurem simulationent, along with the a controllable(b) horizontal iris and shown vertical in (directions.d), along the(c) Optical setup in the experimental measurement, with a controllable iris shown in (d), along the (e) horizontalOptical setup and inthe the vertical experimental directions. measurem ent, with a controllable iris shown in (d), along the horizontal and the vertical directions. horizontal and the vertical directions.

Figure 3. Comparison of the simulated and measured normalized cross-correlation values in the Figure 3. Comparison of the simulated and measured normalized cross-correlation values in the midfield regime and beyond along the horizontal direction at the distance of (a) 1.5, (b) 3, and (c) 5 cm midfield regime and beyond along the horizontal direction at the distance of (a) 1.5, (b) 3, and (c) 5 andFigure along 3. Comparison the vertical direction of the simulated at the distance and measured of (d) 1.5, (noe)rmalized 3, and (f) 5cross-correlation cm. values in the cmmidfield and along regime the and vertical beyond direction along atthe the hori distancezontal ofdirection (d) 1.5, (ate) the 3, and distance (f) 5 cm. of ( a) 1.5, (b) 3, and (c) 5 3. Opticalcm and Design along the and vertical Optimization direction at the distance of (d) 1.5, (e) 3, and (f) 5 cm. 3. Optical Design and Optimization The first-order optical design objective was to determine a rough geometry for achieving a 3. Optical Design and Optimization projectionThe first-order spot with aoptical sufficiently design narrow objective divergence was to angle.determine In addition, a rough we geometry endeavored for toachieving collect all a theprojection fluxThe to first-order spot increase with theaoptical sufficiently illuminance design narrow objective on the divergence target was atto a angle.determine distance In addition, of a 10rough km. we geometry An endeavored appropriate for toachieving approachcollect all a theprojection flux to spotincrease with the a sufficiently illuminance narrow on the divergence target at a distanceangle. In of addition, 10 km. Anwe appropriateendeavored approachto collect allto thisthe fluxgoal to would increase be theto useilluminance a TIR lens on thefor targetcollecti atng a distancethe light ofemitted 10 km. alongAn appropriate all directions approach [36–39]. to However,this goal would a volume-limited be to use aTIR TIR lens lens would for collecti be unsuitng ablethe lightfor achieving emitted aalong high allflux directions and an extremely [36–39]. smallHowever, divergence a volume-limited angle; instead, TIR lensa hybrid would system be unsuit comprisingable for achievinga lens with a ahigh parabolic flux and reflector an extremely would besmall more divergence suitable. Accordingangle; instead, to the a hybridANSI regulation, system comprising the farthest a lens distance with isa parabolicthe maximum reflector distance would at whichbe more the suitable. illuminance According is at leastto the 0.25 ANSI lux regulation, [31]. To reach the farthest this target, distance we used is the LED maximum clusters distance to emit ata sufficientlywhich the illuminance high flux. Our is at data least indicated 0.25 lux that [31]. th Toe pcW-LED reach this could target, be we driven used at LED a maximum clusters powerto emit of a 5.2sufficiently W, where high the flux.current Our injection data indicated was determined that the pcW-LED to be 1.5 couldA. The be objective driven at of a the maximum proposed power design of was5.2 W, not where to drive the thecurrent power injection to this waslevel. determined In the experimental to be 1.5 A. measurement, The objective the of thepcW-LED proposed emitted design a stablewas not flux to driveof 340 the lm power under to thisa designed level. In theheat experimental dissipation measurement,structure with the a pcW-LEDpower injection emitted of a stable flux of 340 lm under a designed heat dissipation structure with a power injection of

Crystals 2019, 9, 524 4 of 12 toCrystals this 2018 goal, 8 would, x FOR PEER be to REVIEW use a TIR lens for collecting the light emitted along all directions [364 –of39 11]. However, a volume-limited TIR lens would be unsuitable for achieving a high flux and an extremely approximately 4.3 W. Therefore, the pcW-LED was driven at this power when the entire system was small divergence angle; instead, a hybrid system comprising a lens with a parabolic reflector would assembled. Our aim was to maintain the total power of the pcW-LEDs below 700 W. Therefore, we be more suitable. According to the ANSI regulation, the farthest distance is the maximum distance used 144 pcW-LEDs. In the first step of calculation, we obtained a total flux of 48960 lm and targeted at which the illuminance is at least 0.25 lux [31]. To reach this target, we used LED clusters to emit a illuminance of 0.25 lux. Projecting the entire portion of light onto a target is impractical. In our sufficiently high flux. Our data indicated that the pcW-LED could be driven at a maximum power of experience, 40–50% of light may be projected to the target area. This thus signifies that the optical 5.2 W, where the current injection was determined to be 1.5 A. The objective of the proposed design was utilization factor (OUF) is 40–50%. To simplify the calculation, if the target illumination area is t2, t not to drive the power to this level. In the experimental measurement, the pcW-LED emitted a stable would be 313 m. The expected projection distance reaches 10 km. Therefore, the full divergence angle flux of 340 lm under a designed heat dissipation structure with a power injection of approximately 4.3 of the spot light was determined to be 1.79°. Regarding the effective area (1 mm2) of the pcW-LED, W. Therefore, the pcW-LED was driven at this power when the entire system was assembled. Our aim the projection lens should be located at a distance of 33.8 mm from the top surface of the pcW-LED. was to maintain the total power of the pcW-LEDs below 700 W. Therefore, we used 144 pcW-LEDs. We could then determine the vertical dimension of the optical system. To determine the diameter of In the first step of calculation, we obtained a total flux of 48960 lm and targeted illuminance of 0.25 the reflector, we set the divergence angle to 1.79° and the reflector length (L) to 33.8 mm. As displayed lux. Projecting the entire portion of light onto a target is impractical. In our experience, 40–50% of in Figure 4, a parabolic reflector was used to collimate the light at larger angles. The nonzero light may be projected to the target area. This thus signifies that the optical utilization factor (OUF) is divergence angle was caused by the lateral extension of the emitting area of the pcW-LED. When the 40–50%. To simplify the calculation, if the target illumination area is t2, t would be 313 m. The expected length of the reflector was 33.8 mm, the light primarily contributing to the divergence angle projection distance reaches 10 km. Therefore, the full divergence angle of the spot light was determined originated from the reflector nearly halfway2 through the length of the reflector (i.e., 16.9 mm), as to be 1.79◦. Regarding the effective area (1 mm ) of the pcW-LED, the projection lens should be located depicted in Figure 5a. If the radius of the open top of the reflector is r, the lateral extension of the light at a distance of 33.8 mm from the top surface of the pcW-LED. We could then determine the vertical source can be Δr and the half divergence angle (Δθ) can be expressed as follows: dimension of the optical system. To determine the diameter of the reflector, we set the divergence angle to 1.79 and the reflector length (L) to 33.8 mm. As displayed in Figure4, a parabolic reflector was ◦ 𝑟∆𝑟 𝑡 used to collimate the light at larger∆𝜃 angles. tan The nonzero tan divergence angle was caused by the lateral(1) extension of the emitting area of the pcW-LED. When𝐿 the length of the𝐿 reflector was 33.8 mm, the light primarilyFigure contributing 5b illustrates to a the plot divergence of Δθ versus angle r at originatedL values offrom 33.8 and the reflector16.9 mm. nearlyTo achieve halfway a divergence through theangle length smaller of the than reflector 1.79°, the (i.e., radius 16.9 mm), of the as parabolic depicted reflector in Figure should5a. If thebe larger radius than of the 16 open mm. topHowever, of the θ reflectorFigure 4 isillustratesr, the lateral that extension the cover of lens the occupies light source the cancentral be ∆ arear and of the the half projection divergence system. angle Therefore, (∆ ) can be expressed as follows: the reflector surface should be moved outward. Consequently, the diameter of the lens would also 1 r + ∆r 1 t be limited. The diameter of the lens∆ θwas= tanset −32 mm, and thetan −diameter of the open top of the reflector(1) L − L was set to 65 mm.

Figure 4. Basic optical structure of the projection light. Figure 4. Basic optical structure of the projection light.

Crystals 2019, 9, 524 5 of 12 Crystals 2018, 8, x FOR PEER REVIEW 5 of 11

Figure 5. (a) Schematic of the reflector when controlling the divergence angle and (b) plot of the half Figure 5. (a) Schematic of the reflector when controlling the divergence angle and (b) plot of the half divergence angle versus the radius of the reflector. divergence angle versus the radius of the reflector. Figure5b illustrates a plot of ∆θ versus r at L values of 33.8 and 16.9 mm. To achieve a divergence Figure 6 presents the layout of the optical components, including the parabolic reflector and angle smaller than 1.79◦, the radius of the parabolic reflector should be larger than 16 mm. However, cover lens. The figure reveals that the emitting area of the pcW-LED is located at the focus of the Figure4 illustrates that the cover lens occupies the central area of the projection system. Therefore, parabolic reflector. The cover lens is placed on the open top of the reflector, and the lens is attached the reflector surface should be moved outward. Consequently, the diameter of the lens would also be to the center of the cover. Because of the 32-mm diameter of the lens, a clear ring aperture can be limited. The diameter of the lens was set 32 mm, and the diameter of the open top of the reflector was obtained for the light reflected by the parabolic reflector. Our simulation revealed the central set to 65 mm. illuminance levels (full divergence angles) to be approximately 2515 lux (1.38°), 1034 lux (1.2°), and Figure6 presents the layout of the optical components, including the parabolic reflector and cover 3549 lux (1.33°) for the parabolic reflector, cover lens, and entire system, respectively. Accordingly, lens. The figure reveals that the emitting area of the pcW-LED is located at the focus of the parabolic the projection spot of the entire system fits the first-order design objective. reflector. The cover lens is placed on the open top of the reflector, and the lens is attached to the center of the cover. Because of the 32-mm diameter of the lens, a clear ring aperture can be obtained for the light reflected by the parabolic reflector. Our simulation revealed the central illuminance levels (full divergence angles) to be approximately 2515 lux (1.38◦), 1034 lux (1.2◦), and 3549 lux (1.33◦) for the parabolic reflector, cover lens, and entire system, respectively. Accordingly, the projection spot of the entire system fits the first-order design objective. In practice, a reflector and cover lens incur a certain amount of distortion during mass production, in which optical components are formed using the plastic medium and an injection molding machine. One drawback is that the shrinkage or distortion rates of the reflector and lens may be different, which could cause defocusing of the optical components. To avoid this situation, a component must be fabricated first, after which other components are adjusted by considering the distortion of the fabricated component. The central illuminance of the projection spot of the reflector was determined to be approximately 2.5 times that of the cover lens. Moreover, controlling reflector distortion would be more difficult than controlling cover lens distortion because of the larger size and complicated shape of the reflector. Therefore, we decided to fabricate the reflector first. The reflector was fabricated using an injection molding machine and coated with aluminum. To ensure a suitable fabrication tolerance, we examined the defocus level by adjusting the longitudinal distance of the pcW-LED attached to an MCPCB (Figure7). In the original design, the emitting surface of the pcW-LED was located at +0.5 mm from the bottom of the reflector; however, in the experiment, the best position was found to be approximately 0.1 mm from the bottom of the reflector (Figure8a). Figure8 b − illustratesFigure the6. Structure illuminance of the at a(a distance) parabolic of reflector, 10 m for (b nine) cover reflectors. lens, and Although (c) entire the system. best positionsSimulated were different for nine reflectors, the position at 0.1 mm was the optimal one. Thus, the distortion of the projection spot at a distance of 1 km and divergence− angle of the light coming from the (d) parabolic reflectorreflector, caused (e) cover a defocus lens, and of 0.6 (f) mm.entire The system. measured central illuminance varied from 2403 to 2569 lux, representing 96–102% of the simulation value. In practice, a reflector and cover lens incur a certain amount of distortion during mass production, in which optical components are formed using the plastic medium and an injection molding machine. One drawback is that the shrinkage or distortion rates of the reflector and lens may

Crystals 2018, 8, x FOR PEER REVIEW 5 of 11

Figure 5. (a) Schematic of the reflector when controlling the divergence angle and (b) plot of the half divergence angle versus the radius of the reflector.

Figure 6 presents the layout of the optical components, including the parabolic reflector and cover lens. The figure reveals that the emitting area of the pcW-LED is located at the focus of the parabolic reflector. The cover lens is placed on the open top of the reflector, and the lens is attached to the center of the cover. Because of the 32-mm diameter of the lens, a clear ring aperture can be obtained for the light reflected by the parabolic reflector. Our simulation revealed the central illuminance levels (full divergence angles) to be approximately 2515 lux (1.38°), 1034 lux (1.2°), and Crystals3549 lux2019 (1.33°), 9, 524 for the parabolic reflector, cover lens, and entire system, respectively. Accordingly,6 of 12 the projection spot of the entire system fits the first-order design objective.

Crystals 2018, 8, x FOR PEER REVIEW 6 of 11 be different, which could cause defocusing of the optical components. To avoid this situation, a component must be fabricated first, after which other components are adjusted by considering the distortion of the fabricated component. The central illuminance of the projection spot of the reflector was determined to be approximately 2.5 times that of the cover lens. Moreover, controlling reflector distortion would be more difficult than controlling cover lens distortion because of the larger size and complicated shape of the reflector. Therefore, we decided to fabricate the reflector first. The reflector was fabricated using an injection molding machine and coated with aluminum. To ensure a suitable fabrication tolerance, we examined the defocus level by adjusting the longitudinal distance of the pcW-LED attached to an MCPCB (Figure 7). In the original design, the emitting surface of the pcW-LED was located at +0.5 mm from the bottom of the reflector; however, in the experiment, the best position was found to be approximately −0.1 mm from the bottom of the reflector (Figure 8a). Figure 8b illustrates the illuminance at a distance of 10 m for nine reflectors. Although the best positionsFigure were 6. Structure different of the for (a nine) parabolic reflectors, reﬂector, the (b )position cover lens, at and −0.1 (c )mm entire was system. the Simulatedoptimal one. projection Thus, the Figure 6. Structure of the (a) parabolic reflector, (b) cover lens, and (c) entire system. Simulated distortionspot at of a distancethe reflector of 1 km caused and divergence a defocus angleof 0.6 of mm. the lightThe comingmeasured from central the (d) illuminance parabolic reﬂector, varied (e )from projection spot at a distance of 1 km and divergence angle of the light coming from the (d) parabolic 2403 coverto 2569 lens, lux, and representing (f) entire system. 96–102% of the simulation value. reflector, (e) cover lens, and (f) entire system.

In practice, a reflector and cover lens incur a certain amount of distortion during mass production, in which optical components are formed using the plastic medium and an injection molding machine. One drawback is that the shrinkage or distortion rates of the reflector and lens may

Figure 7. Distance between the pcW-LED and reflector was adjusted when (a) the pcW-LED was at the designedFigure 7. positionDistance andbetween (b) departed the pcW-LED from the and position reflector in (wasa). adjusted when (a) the pcW-LED was at the designed position and (b) departed from the position in (a). We adjusted the location of the pcW-LED to the optimized position by considering the fabricated reflector. We then modified the lens shape to shorten the focal length. The lens diameter was adjusted to 31.84 mm to fit the optimized position of the pcW-LED (Figure9). The central illuminance (divergence angle) varied from 1035 lux (1.2◦) to 1032 lux (1.19◦) in the simulation. The cover lens was made with PMMA using the injection molding machine. Figure 10 displays the projection spots for the nine samples. The measured central illuminance ranged from 791 to 915 lux, representing 77% to 89% of the simulation value. The degradation of illuminance may have been caused by the deformation of the lens during its manufacturing, where defocus, distortion, and dispersion may have been induced. Finally, the measured central illuminance of the spot light unit was 2895–3298 lux, representing 86–98% of the simulation value (Figure 11).

Figure 8. (a) Projection spots and variation of the central illuminance when the pcW-LED was shifted longitudinally. (b) Plot of the measured central illuminance versus the defocus for nine samples. The dark green line represents the results for the original design. The measurement was conducted for an injection current of 1.5 A with a heavy heat sink.

Crystals 2018, 8, x FOR PEER REVIEW 6 of 11

be different, which could cause defocusing of the optical components. To avoid this situation, a component must be fabricated first, after which other components are adjusted by considering the distortion of the fabricated component. The central illuminance of the projection spot of the reflector was determined to be approximately 2.5 times that of the cover lens. Moreover, controlling reflector distortion would be more difficult than controlling cover lens distortion because of the larger size and complicated shape of the reflector. Therefore, we decided to fabricate the reflector first. The reflector was fabricated using an injection molding machine and coated with aluminum. To ensure a suitable fabrication tolerance, we examined the defocus level by adjusting the longitudinal distance of the pcW-LED attached to an MCPCB (Figure 7). In the original design, the emitting surface of the pcW-LED was located at +0.5 mm from the bottom of the reflector; however, in the experiment, the best position was found to be approximately −0.1 mm from the bottom of the reflector (Figure 8a). Figure 8b illustrates the illuminance at a distance of 10 m for nine reflectors. Although the best positions were different for nine reflectors, the position at −0.1 mm was the optimal one. Thus, the distortion of the reflector caused a defocus of 0.6 mm. The measured central illuminance varied from 2403 to 2569 lux, representing 96–102% of the simulation value.

Figure 7. Distance between the pcW-LED and reflector was adjusted when (a) the pcW-LED was at the designed position and (b) departed from the position in (a). Crystals 2019, 9, 524 7 of 12

Crystals 2018, 8, x FOR PEER REVIEW 7 of 11 Crystals 2018, 8, x FOR PEER REVIEW 7 of 11 We adjusted the location of the pcW-LED to the optimized position by considering the fabricated reflector.We adjusted We then themodified location the of lens the pcW-LEDshape to shor to theten optimized the focal length. position The by lens considering diameter thewas fabricated adjusted toreflector. 31.84 mmWe thento fit modified the optimized the lens shapeposition to shorof theten pcW-LEDthe focal length. (Figure The 9). lens The diameter central wasilluminance adjusted (divergenceto 31.84 mm angle) to fit varied the optimized from 1035 luxposition (1.2°) ofto 1the032 pcW-LEDlux (1.19°) in(Figure the simulation. 9). The central The cover illuminance lens was made(divergence with PMMA angle) variedusing thefrom injection 1035 lux molding (1.2°) to machine.1032 lux (1.19°) Figure in 10 the displays simulation. the projection The cover spotslens was for themade nine with samples. PMMA The using measured the injection central molding illuminance machine. ranged Figure from 10791 displays to 915 lu thex, representingprojection spots 77% for to the nine samples. The measured central illuminance ranged from 791 to 915 lux, representing 77% to 89% of the simulation value. The degradation of illuminance may have been caused by the deformation89% Figureof the 8. simulation(ofa) the Projection lens duringvalue. spots and Theits variation manufacturing, degradation of the central of where illuminance illuminance defocus, whenmay distortion, thehave pcW-LED beenand dispersion wascaused shifted by may the havedeformationlongitudinally. beenFigure induced. 8.of ( athe) Projection ( blensFinally,) Plot during ofspots the the andmeasuredits measured manufacturing,variation centralcentral of the illuminance centralilluminance where illuminance defocus, versus of the thewhendistortion, spot defocus the light pcW-LED for andunit nine dispersionwas was samples. 2895–3298 shifted may lux,have representingThe beenlongitudinally. dark induced. green 86–98% line (Finally,b) represents Plot of of thethe the thesimulation measured results for cent valuecentral theral original illuminance(Figure illuminance design. 11). versus The of measurementthethe despotfocus light for was unitnine conducted samples.was 2895–3298 for The lux, representingandark injection green current line 86–98% represents of 1.5 of Athe withthe simulation resu a heavylts for heat thevalue sink.original (Figure design. 11). The measurement was conducted for an injection current of 1.5 A with a heavy heat sink.

Figure 9. (a) Structure of the modified cover lens; (b) ray fans in the simulation; and (c) divergence Figure 9. (a) Structure of the modiﬁed cover lens; (b) ray fans in the simulation; and (c) divergence angleFigure of 9. the (a )spot Structure light. of the modified cover lens; (b) ray fans in the simulation; and (c) divergence angleangle ofof thethe spotspot light.light.

FigureFigure 10.10. ((a)) SimulatedSimulated projectionprojection spotspot ofof thethe covercover lenslens atat aa distancedistance ofof 1010 mm andand ((bb)) correspondingcorresponding measurementmeasurementFigure 10. (a) forSimulatedfor thethe ninenine projection samples. samples. The spotThe measurements ofmeasurements the cover lens were were at conducteda conducteddistance of at at10 an an m injection injectionand (b) current corresponding current of of 1.5 1.5 A withAmeasurement with a heavy a heavy heat for heat sink.the sink. nine samples. The measurements were conducted at an injection current of 1.5 A with a heavy heat sink.

Crystals 2018, 8, x FOR PEER REVIEW 8 of 11

Crystals 2019, 9, 524 8 of 12 Crystals 2018, 8, x FOR PEER REVIEW 8 of 11

Figure 11. (a) Simulated projection spot of the spot light unit at a distance of 10 m and (b) corresponding measurements for the nine samples. The measurements were conducted at an injection current of 1.5 A with a heavy heat sink.

4. PerformanceFigure 11. (a) of Simulated the Spot projection Light System spot of the spot light unit at a distance of 10 m and (b) corresponding Figure 11. (a) Simulated projection spot of the spot light unit at a distance of 10 m and (b) measurements for the nine samples. The measurements were conducted at an injection current of 1.5 A correspondingTo achieve a measurementsprojection distance for the nine of 10 samples. km, we The constructed measurements a spot were light conducted module at an comprising injection 16 with a heavy heat sink. spotcurrent light units. of 1.5 TheA with final a heavy system heat contained sink. nine modules; thus, the entire system contained 144 spot 4.light Performance units. Figure of the12 displays Spot Light one Systemof the modules; the total power, injection current, divergence angle, 4.and Performance maximum of luminous the Spot intensityLight System of the module were 68.2 W, 1.25 A, 1.6°, and 2,840,000 cd, To achieve a projection distance of 10 km, we constructed a spot light module comprising 16 spot respectively. The projection distance was 3.37 km, according to the ANSI regulation [31]. Figure 13 lightTo units. achieve The finala projection system contained distance of nine 10 modules;km, we constructed thus, the entire a spot system light contained module comprising 144 spot light 16 illustrates the entire system. The projected light split at a short distance and then merged toward a units.spot light Figure units. 12 The displays final system one of contained the modules; nine the mo totaldules; power, thus, the injection entire current,system contained divergence 144 angle, spot lightround units. spot Figure at a farther 12 displays distance one [40]. of the Finally, modules; the theentire total system power, projected injection light current, to a divergencedistance of angle,10 km and maximum luminous intensity of the module were 68.2 W, 1.25 A, 1.6◦, and 2,840,000 cd, respectively. at steady state. The total electric power of the entire system was less than 700 W. Compared with Theand projectionmaximum distance luminous was intensity 3.37 km, of according the module to the were ANSI 68.2 regulation W, 1.25 [31 A,]. Figure1.6°, and 13 illustrates2,840,000 thecd, traditional xenon projection lamps, whose electric power is in kilowatts, the proposed system entirerespectively. system. The The projection projected distance light split was at 3.37 a short km, distanceaccording and to the then ANSI merged regulation toward [31]. a round Figure spot 13 exhibited greater energy-saving effects. Heat considerably affects LEDs, and the power density of an atillustrates a farther the distance entire [system.40]. Finally, The projected the entire light system split projected at a short light distance to adistance and then of merged 10 km attoward steady a LED is limited. Therefore, an adequate heat dissipation design is required. Moreover, due to the high state.round The spot total at a electricfarther distance power of [40]. the entireFinally, system the en wastire system less than projected 700 W. Comparedlight to a distance with traditional of 10 km luminous intensity of the proposed system, attention should be paid to photobiological safety to xenonat steady projection state. The lamps, total whose electric electric power power of the is en intire kilowatts, system thewas proposed less thansystem 700 W. exhibited Compared greater with prevent strong light from causing damage to the human eye [41,42]. energy-savingtraditional xenon effects. projection Heat considerably lamps, whose affects electric LEDs, power and theis powerin kilowatts, density the of anproposed LED is limited.system The proposed system is suitable for several applications due to its advantages such as compact Therefore,exhibited greater an adequate energy-saving heat dissipation effects. Heat design considerably is required. affects Moreover, LEDs, and due the to power the high density luminous of an size and low energy consumption. It is suitable for use in ships and fishing boats. Large-volume intensityLED is limited. of the proposedTherefore, system, an adequate attention heat should dissipation be paid design to photobiological is required. Moreover, safety to due prevent to the strong high multicombination projection systems are suitable for use in lighthouse and bird towers. lightluminous from causingintensity damage of the proposed to the human system, eye [attention41,42]. should be paid to photobiological safety to prevent strong light from causing damage to the human eye [41,42]. The proposed system is suitable for several applications due to its advantages such as compact size and low energy consumption. It is suitable for use in ships and fishing boats. Large-volume multicombination projection systems are suitable for use in lighthouse and bird towers.

Figure 12. (a) Spot light module and (b) measured intensity distribution. Figure 12. (a) Spot light module and (b) measured intensity distribution.

Figure 12. (a) Spot light module and (b) measured intensity distribution.

Crystals 2018, 8, x FOR PEER REVIEW 9 of 11 Crystals 2019, 9, 524 9 of 12

Figure 13. (a,b) Entire spot light system for 10-km projection; (c) light paths from the system; and (d) Figurelight 13. path (a,b observed) Entire spot at a distancelight system far away for the10-km system. projection; (c) light paths from the system; and (d) light path observed at a distance far away the system. The proposed system is suitable for several applications due to its advantages such as compact size5. Conclusion and low energy consumption. It is suitable for use in ships and fishing boats. Large-volume multicombination projection systems are suitable for use in lighthouse and bird towers. In this study, we designed a white LED spot light system to achieve a 10-km projection distance. 5.We Conclusions created the design after comparing six commercial high-power white LEDs on the basis of their exitance and injection power. Subsequently, we selected Osram HWQP as the light source. Through preciseIn thisoptical study, modeling we designed of the alight white source, LED spotwe dec lightided system to use to aachieve hybrid aoptical 10-km system projection containing distance. a Wecover created lens and the parabolic design afterreflector comparing rather than six a commercial TIR lens. The high-power expected projection white LEDs distance on the was basis 10 km. of theirTherefore, exitance the andfull divergence injection power. angle of Subsequently, the spot light we system selected was Osram1.79°. Rega HWQPrding as the the effective light source. area Through(1 mm2) of precise Osram optical HWQP, modeling the projection of the lens light should source, bewe located decided at a todistance use a hybridof more optical than 33.8 system mm containingfrom the top a coversurface lens of andthe pcW-LED. parabolic reflector The minimum rather than radius a TIR of the lens. reflector The expected was determined projection through distance wassimple 10 geometrical km. Therefore, calculation. the full divergence angle of the spot light system was 1.79◦. Regarding the 2 effectiveConsidering area (1 mm practical) of Osram fabrication HWQP, errors, the projection we fabricated lens should the bereflector located first at a distancethrough ofan more injection than 33.8molding mm frommachine. the topThe surface reflector’s of the distortion pcW-LED. changed The minimum the best radiuslocation of of the the reflector pcW-LED was from determined +0.5 to through−0.1 mm simpleon a reference geometrical plane. calculation. The cover lens was then modified to fit the new location. The measured centralConsidering illuminance practical of the reflector fabrication was 2403–2569 errors, we lux, fabricated representing the reflector 96–102% first of the through simulation an injection value, molding machine. The reflector’s distortion changed the best location of the pcW-LED from +0.5 to 0.1 and that of the cover lens was 791–915 lux, representing 77–89% of the simulation value. −The mmdegradation on a reference of illuminance plane. The was cover poss lensibly was due then to modified the deformation to fit the newof the location. lens during The measured manufacturing, central illuminancewhere defocus, of the distortion, reflector wasand 2403–2569dispersion lux, may representing have been 96–102% induced. of Finally, the simulation the measured value, and central that ofilluminance the cover of lens the was spot 791–915 light unit lux, was representing 2895–3298 77–89% lux, representing of the simulation 86–98% value.of the simulation The degradation value. of illuminanceThe entire was system possibly contained due to the nine deformation modules, each of the of lens which during comprised manufacturing, 16 LED spot where light defocus, units. distortion,The power andinjection dispersion of each may module have beenwas only induced. 68.2 Finally,W. Because the measuredof the excellent central performance illuminance of the spotLED lightspot unitlight was module 2895–3298 in terms lux, representingof the divergen 86–98%ce angle of the of simulation the projection value. beam and maximum luminousThe entire intensity, system which contained were 1. nine6° and modules, 2,840,000 each cd, of respectively, which comprised the projection 16 LED spotdistance light of units. the Themodule power could injection reach of3.37 each km, module according was onlyto the 68.2 ANSI W. Becauseregulation of the[31]. excellent Finally,performance the projection of distance the LED spotof the light entire module system in was terms demonstrated of the divergence to reach angle appr ofoximately the projection 10 km beamwith a and divergence maximum angle luminous of 1.6°. intensity, which were 1.6◦ and 2,840,000 cd, respectively, the projection distance of the module could reachAuthor 3.37 Contributions: km, according Conceptualization to the ANSI regulationand writing [work,31]. Finally, C.C.S.; Simulation, the projection K.Y.C.; distance Experiment, of the C.S.W., entire X.H.L., K.Y.C., S.K.L., and S.Y.C.; part of calculation, T.H.Y., C. C. S., C.S.W., X.H.L.; discussion, Y.W.Y. system was demonstrated to reach approximately 10 km with a divergence angle of 1.6◦. Funding: This research has been funded by the Ministry of Science and Technology of the Republic of China, grant numbers MOST 106-2221-E-008-065-MY3 and 108-2622-E-008-010-CC2.

Acknowledgments: The author would like to thank BRO for sponsoring ASAP, Benex for supporting in fabrication of the optical components, and all team members led by Ching-Cherng Sun and Tsung-Hsun Yang for assembling the whole system.

Crystals 2019, 9, 524 10 of 12

Author Contributions: Conceptualization and writing work, C.-C.S.; Simulation, K.-Y.C.; Experiment, C.-S.W., X.-H.L., K.-Y.C., S.-K.L., and J.-Y.C.; part of calculation, T.-H.Y., C.-C.S., C.-S.W., X.-H.L.; discussion, Y.-W.Y. Funding: This research has been funded by the Ministry of Science and Technology of the Republic of China, grant numbers MOST 106-2221-E-008-065-MY3 and 108-2622-E-008-010-CC2. Acknowledgments: The author would like to thank BRO for sponsoring ASAP, Benex for supporting in fabrication of the optical components, and all team members led by Ching-Cherng Sun and Tsung-Hsun Yang for assembling the whole system. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Wang, S.; Wang, K.; Chen, F.; Liu, S. Design of primary optics for LED chip array in road lighting application. Opt. Express 2011, 19, A716–A724. [CrossRef] 2. Lee, X.H.; Moreno, I.; Sun, C.C. High-performance LED street lighting using microlens arrays. Opt. Express 2013, 21, 10612–10621. [CrossRef][PubMed] 3. Lee, X.H.; Chang, Y.Y.; Sun, C.C. Highly energy-efficient agricultural lighting by B+R LEDs with beam shaping using micro-lens diffuser. Opt. Commun. 2013, 291, 7–114. [CrossRef] 4. Jafrancesco, D.; Mercatelli, L.; Sansoni, P.; Fontani, D.; Sani, E.; Coraggia, S.; Meucci, M.; Francini, F. Optical design of a light-emitting diode lamp for a maritime lighthouse. Appl. Opt. 2015, 54, 3252–3262. [CrossRef] 5. Lee, X.H.; Yang, J.T.; Chang, J.H.; Chien, W.T.; Lo, Y.C.; Lin, C.C.; Sun, C.C. An LED-based luminaire for badminton court illumination. Lighting Res. Technol. 2017, 49, 396–406. [CrossRef] 6. Fantozzi, F.; Leccese, F.; Salvadori, G.; Rocca, M.; Garofalo, M. LED lighting for indoor sports facilities: Can its use be considered as sustainable solution from a techno-economic standpoint? Sustainability 2016, 8, 618. [CrossRef] 7. Piccablotto, G.; Aghemo, C.; Pellegrino, A.; Iacomussi, P.; Radis, M. Study on conservation aspects using LED technology for museum lighting. Energy Procedia 2015, 78, 1347–1352. [CrossRef] 8. Kim, A.; Wang, S.; McCunn, L.J. Building value proposition for interactive lighting systems in the workplace: Combining energy and occupant perspectives. J. Build. Eng. 2019, 24, 100752. [CrossRef] 9. Narendran, N.; Maliyagoda, N.; Bierman, A.; Pysar, R.; Overington, M. Characterizing white LEDs for general illumination applications. Proc. SPIE 2000, 3938, 240–248. [CrossRef] 10. Steigerwald, D.A.; Bhat, J.C.; Collins, D.; Fletcher, R.M.; Holcomb, M.O.; Ludowise, M.J.; Martin, P.S.; Rudaz, S.L. Illumination with solid state lighting technology. IEEE J. Sel. Top. Quantum Electron. 2002, 8, 310–320. [CrossRef] 11. Zukauskas, A.; Shur, M.S.; Caska, R. Introduction to Solid-State Lighting; John Wiley & Sons: New York, NY, USA, 2002; ISBN 978-0-471-21574-5. 12. Schubert, E.F.; Kim, J.K. Solid-state light sources getting smart. Science 2005, 308, 1274–1278. [CrossRef] [PubMed] 13. Schubert, E.F. Light-Emitting Diods, 2nd ed.; Cambridge University Press: New York, NY, USA, 2006; ISBN 9780521865388. 14. Krames, M.R.; Shchekin, O.B.; Mueller-Mach, R.; Mueller, G.O.; Zhou, L.; Harbers, G.; Craford, M.G. Status and future of high-power light-emitting diodes for solid-state lighting. J. Disp. Technol. 2007, 3, 160–175. [CrossRef] 15. Pimputkar, S.; Speck, J.S.; DenBaars, S.P.; Nakamura, S. Prospects for LED lighting. Nat. Photonics 2009, 3, 180–182. [CrossRef] 16. Karlicek, B.; Sun, C.C.; Zissis, G.; Ma, R. Handbook of Advanced Lighting Technology; Springer: Cham, Switzerland, 2017; ISBN 978-3-319-00175-3. 17. Sun, C.C.; Chang, Y.Y.; Yang, T.H.; Chung, T.Y.; Chen, C.C.; Lee, T.X.; Li, D.R.; Lu, C.Y.; Ting, Z.Y.; Glorieux, B.; et al. Packaging efficiency in phosphor-converted white LEDs and its impact to the limit of luminous efficacy. J. Soild State Lighting 2014, 1, 19. [CrossRef] Crystals 2019, 9, 524 11 of 12

18. Yang, T.H.; Wu, S.M.; Sun, C.C.; Glorieux, B.; Chen, C.Y.; Chang, Y.Y.; Lee, X.H.; Yu, Y.W.; Chung, T.Y.; Lai, K.Y. Stabilizing CCT in pcW-LEDs by self-compensation between excitation efficiency and conversion efficiency of phosphors. Opt. Express 2017, 25, 29287–29295. [CrossRef] 19. Derlofske, J.F.V.; McClogan, M.W. White LED sources for vehicle forward lighting. Proc. SPIE 2002, 4776, 195–205. [CrossRef] 20. Albou, P. LED modules for front lighting applications. In Proceedings of the 5th International Symposium on Progress in Automobile Lighting, München, Germany, 23–24 September 2003; Herbert Utz Verlag: München, Germany, 2003; Volume 10, pp. 37–51. 21. Luce, T. Customized LED light sources for automotive applications. In Proceedings of the 5th International Symposium on Progress in Automobile Lighting, München, Germany, 23–24 September 2003; Herbert Utz Verlag: München, Germany, 2003; Volume 10, pp. 614–620. 22. Moeller, N. Design aspects and headlamp performance—The left and right headlamp are not the same. In Proceedings of the 5th International Symposium on Progress in Automobile Lighting, München, Germany, 23–24 September 2003; Herbert Utz Verlag: München, Germany, 2003; Volume 10, pp. 686–697. 23. Tatsukawa, M.; Ishida, H.; Sazuka, K. Vehicle Headlamp. U.S. Patent 6,891,333 B2, 10 May 2005. 24. Cvetkovic, A.; Dross, O.; Chaves, J.; Benítez, P.; Miñano, J.C.; Mohedano, R. Etendue-preserving mixing and projection optics for high-luminance LEDs, applied to automotive headlamps. Opt. Express 2006, 14, 13014–13020. [CrossRef] 25. Chen, F.; Wang, K.; Qin, Z.; Wu, D.; Luo, X.; Liu, S. Design method of high-efficient LED headlamp lens. Opt. Express 2010, 18, 20926–20938. [CrossRef] 26. Brick, P.; Schmid, T. Automotive headlamp concepts with low-beam and high-beam out of a single LED. Proc. SPIE 2011, 8170, 817008. [CrossRef] 27. Hsieh, C.C.; Li, Y.H.; Hung, C.C. Modular design of the LED vehicle projector headlamp system. Appl. Opt. 2013, 52, 5221–5229. [CrossRef] 28. Sun, W.S.; Tien, C.L.; Lo, W.C.; Chu, P.Y. Optical design of an LED motorcycle headlamp with compound reflectors and a toric lens. Appl. Opt. 2015, 54, E102–E108. [CrossRef][PubMed] 29. Chen, H.C.; Zhou, J.H.; Zhou, Y. Stacking illumination of a confocal reflector light emitting diode automobile headlamp with an asymmetric triangular prism. Appl. Opt. 2017, 56, 1087–1093. [CrossRef][PubMed] 30. The Datasheet of OSRAM HWQP LED, OSRAM Opto Semiconductors GmbH. Available online: https: //dammedia.osram.info/media/resource/hires/osram-dam-5710564/LUW%20HWQP_EN.pdf (accessed on 12 October 2019). 31. American National Standards Institute (ANSI). ANSI/NEMA FL 1-2009: Flashlight Basic Performance Standard; National Electrical Manufacturers Association: Virginia, VA, USA, 2009. 32. Sun, C.C.; Lee, T.X.; Ma, S.H.; Lee, Y.L.; Huang, S.M. Precise optical modeling for LED lighting verified by cross correlation in the midfield region. Opt. Lett. 2006, 31, 2193–2195. [CrossRef][PubMed] 33. Chien, W.T.; Sun, C.C.; Moreno, I. Precise optical model of multi-chip white LEDs. Opt. Express 2007, 15, 7572–7577. [CrossRef][PubMed] 34. Sun, C.C.; Chien, W.T.; Moreno, I.; Hsieh, C.C.; Lo, Y.C. Analysis of the far-field region of LEDs. Opt. Express 2009, 17, 13918–13927. [CrossRef][PubMed] 35. Lewis, J.P. Fast Template Matching, Vision Interface 95. In Proceedings of the Canadian Image Processing and Pattern Recognition Society, Quebec City, QC, Canada, 15–19 May 1995; pp. 120–123. 36. Chen, J.J.; Lin, C.T. Freeform surface design for a light-emitting diode–based collimating lens. Opt. Eng. 2010, 49, 093001. [CrossRef] 37. Chen, J.J.; Wang, T.Y.; Huang, K.L.; Liu, T.S.; Tsai, M.D.; Lin, C.T. Freeform lens design for LED collimating illumination. Opt. Express 2012, 20, 10984–10995. [CrossRef][PubMed] 38. Parkyn, W.A.; Pelka, D.G. New TIR lens applications for light-emitting diodes. Proc. SPIE 1997, 3139, 135–140. [CrossRef] 39. Welford, W.T.; Winston, R. High Collection Nonimaging Optics; Elsevier Inc.: Amsterdam, The Netherlands, 1989; ISBN 978-0-12-742885-7. 40. Moreno, I.; Sun, C.C.; Ivanov, R. Far-field condition for light-emitting diode arrays. Appl. Opt. 2009, 48, 1190–1197. [CrossRef] Crystals 2019, 9, 524 12 of 12

41. Leccese, F.; Vandelanotte, V.; Salvadori, G.; Rocca, M. Blue light hazard and risk group classiﬁcation of 8 W LED tubes, replacing ﬂuorescent tubes, through optical radiation measurements. Sustainability 2015, 7, 13454–13468. [CrossRef] 42. Jung, M.H.; Yang, S.J.; Yuk, J.S.; Oh, S.Y.; Kim, C.J.; Lyu, J.; Choi, E.J. Evaluation of Blue Light Hazards in LED Lightings. J. Korean Ophthalmic Opt. Soc. 2015, 20, 293–300. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).