Application Brief AB05
`
`Thermal Design Using
`LUXEON® Power Light Sources
`
`Introduction
`LUXEON® Power Light Sources provide the highest light output with the smallest footprint of any Light
`Emitting Diodes (LEDs) in the world. This is due, in part, to LUXEON's ground breaking thermal design.
`LUXEON is the first LED solution to separate thermal and electrical paths, drawing more heat away from
`the emitter core and significantly reducing thermal resistance. As a result, LUXEON packages handle
`significantly more power than competing LEDs. LUXEON's larger, brighter emitters together with its
`unique high(cid:2)power capabilities provide a tremendous amount of light in a small, durable package. This, in
`turn, provides lighting designers with a unique opportunity to explore new designs and product ideas and
`to improve the quality, energy(cid:2)efficiency, safety and longevity
`of existing products.
`
`Lighting designers working with LUXEON Power Light Sources do need to consider some potentially
`unfamiliar factors, such as the impact of temperature rise on optical performance. Proper thermal design
`is imperative to keep the LED emitter package below its rated operating temperature. This application
`note will assist design engineers with thermal management strategies.
`
`We recommend taking the time to develop a thermal model for your application before finalizing your
`design. The LUXEON Custom Design Guide provides important details about operating temperatures for
`each LED emitter package. Once you determine your target temperature, a thermal model will allow you
`to consider the impact of factors such as size, type of heat sink, and
`airflow requirements.
`
`Index
`
`Lighting designers needing additional development support for thermal
`management issues will find ample resources. The thermal management
`industry has grown along side advances in electronics design. The
`thermal analysis resources section at the end of this document provides
`a useful introduction to some industry resources.
`
`Introduction . . . . . . . . . . . . . . . . . .1
`Minimum Heat Sink Requirements .2
`Thermal Modeling . . . . . . . . . . . . . .2
`Inputs/Output of the Thermal Model 4
`Heat Sink Characterization . . . . . .4
`Attachment to Heat Sinks . . . . . . .7
`Best Practices for Thermal Design .8
`Evaluating Your Design . . . . . . . . .8
`Validation of Method . . . . . . . . . . .11
`
`SEC et al. v. MRI
`SEC Exhibit 1024.001
`IPR 2023-00199
`
`

`

`Minimum Heat Sink Requirements
`All LUXEON products mounted on an aluminum, metal(cid:2)core
`printed circuit board (MCPCB, also called Level 2 products)
`can be lit out of the box, though we do not recommend
`lighting the Flood for more than a few seconds without an
`additional heat sink.
`
`As a rule, product applications using LUXEON Power Light
`Sources require mounting to a heat sink for proper thermal
`management in all operating conditions. Depending on the
`application, this heat sink can be as simple as a flat,
`aluminum plate.
`
`The LUXEON Star, Line and Ring products consist of LEDs
`mounted on MCPCB in various configurations (see the
`LUXEON Product Guide). These products have 1 in2 of
`MCPCB per emitter. The MCPCB acts as an electrical inter(cid:2)
`connect, as well as a thermal heat sink interface. While we
`recommend using an additional heat sink, these products
`can be operated at 25°C without one. The MCPCB can get
`very hot (~70°C) without a heat sink. Use appropriate
`precautions.
`
`A LUXEON Flood should be mounted to a heat sink before
`being illuminated for more than a few seconds. A flat
`aluminum plate with an area of about 36 in2 (6" x 6" x
`0.0625" thick) is adequate when operating at 25°C.
`Thermal Modeling
`The purpose of thermal modeling is to predict the junction
`temperature (Tjunction). The word "junction" refers to the p(cid:2)n
`junction within the semiconductor die. This is the region of
`the chip where the photons are created and emitted. You
`can find the maximum recommended value for each
`LUXEON product in your data sheet. This section describes
`how to determine the junction temperature for a given appli(cid:2)
`cation using a thermal model.
`
`A. Thermal Resistance Model
`One of the primary mathematical tools used
`in thermal management design is thermal resistance (RΘ).
`Thermal resistance is defined as the ratio of temperature
`difference to the corresponding power dissipation. The
`overall RΘJunction(cid:2)Ambient (J(cid:2)A) of a LUXEON Power Light Source
`plus a heat sink is defined in Equation 1:
`
`Equation 1. Definition of Thermal Resistance
`Δ
`T
`
`Ambient
`
`−
`
`Junction
`P
`d
`
`Θ
`
`R
`
`−
`Junction Ambient
`
`=
`
`Where:
`ΔT = TJunction (cid:2) TAmbient (°C)
`Pd = Power dissipated (W)
`Pd = Forward current (If) * Forward voltage (Vf)
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`2
`
`Heat generated at the junction travels from the die along
`the following simplified thermal path: junction(cid:2)to(cid:2)slug,
`slug(cid:2)to(cid:2)board, and board(cid:2)to(cid:2)ambient air.
`
`For systems involving conduction between multiple surfaces
`and materials, a simplified model of the thermal path is a
`series(cid:2)thermal resistance circuit, as shown in Figure 1A. The
`overall thermal resistance (RΘJ(cid:2)A) of an application can be
`expressed as the sum of the individual resistances of the
`thermal path from junction to ambient (Equation 2). The corre(cid:2)
`sponding components of each resistance in the heat path
`are shown in Figure 1B. The physical components of each
`resistance lie between the respective temperature nodes.
`P = V F * I F
`d
`
`TJunction
`
`TSlug
`
`TBoard
`
`RΘJ-S
`
`RΘS-B
`
`RΘB-A
`
`TAmbient
`Figure 1A. Series Resistance Thermal Count
`
`
`
`MCPCB
`Heat sink
`
`T A
`
`T Junction
`
`T Slug
`
`Die
`Die attach
`Epoxy
`
`T Board
`
`Figure 1B. Emitter Cut(cid:2)Away
`
`Equation 2. Thermal Resistance Model
`RΘJunction(cid:2)Ambient = RΘJunction(cid:2)Slug + RΘSlug(cid:2)Board + RΘBoard(cid:2)Ambient
`
`RΘSlug(cid:2)Board (S(cid:2)B)
`
`Where:
`RΘJunction(cid:2)Slug(J(cid:2)S) = RΘ of the die attach combined with
`die and slug material in contact with
`the die attach.
`= RΘ of the epoxy combined with slug
`and board materials in contact with
`the epoxy.
`RΘBoard(cid:2)Ambient (B(cid:2)A) = the combined RΘ of the surface
`contact or adhesive between the heat
`sink and the board and the heat sink
`into ambient air.
`
`SEC et al. v. MRI
`SEC Exhibit 1024.002
`IPR 2023-00199
`
`

`

`Equation 3, derived from Equation 1 can be used to
`calculate the junction temperature of the LUXEON device.
`
`Equation 3. Junction Temperature Calculation
`TJunction = TA + (Pd)(RΘJ(cid:2)A)
`
`Where:
`= Ambient temperature
`TA
`Pd
`= Power Dissipated (W) = Forward current
`(If ) * Forward voltage (Vf )
`RΘJ(cid:2)A = Thermal resistance junction to ambient
`
`B. Thermal Resistance of LUXEON Light
`Sources
`In LUXEON Power Light Sources, Philips Lumileds has opti(cid:2)
`mized the junction(cid:2)to(cid:2)board thermal path to minimize the
`thermal resistance. The thermal resistance of a LUXEON
`emitter (not mounted on an MCPCB, also called a Level 1) is
`represented by RΘJ(cid:2)S.
`
`The thermal resistance of LUXEON Power Light Sources
`(MCPCB mounted, also called a Level 2) representing by
`RΘJ(cid:2)B, equal to:
`
`RΘJ(cid:2)B = RΘJ(cid:2)S + RΘS(cid:2)B
`
`
`
`LED
`1
`T Junction
`R Θ Junction-Slug
`
`T Slug
`R Θ Slug -Board
`
`LED
`2
`
`LED
`3
`
`LED
`4
`
`LED
`N
`
`…
`
`…
`
`…
`
`T Board
`R Θ Board-Ambient
`T Ambient
`Figure 2. Parallel Thermal Resistance Model
`of Multiple Emitter Products
`
`The RΘJ(cid:2)B of the multiple(cid:2)emitter array is obtained by using
`the parallel resistance equation:
`
`1
`Total_Array_R
`
`Θ
`
`−
`Junction Board
`
`=
`
`LED(1)_R
`
`1
`Θ
`
`−
`Junction Board
`
`+
`
`+
`
`...
`
`LED(N)_R
`
`1
`Θ
`−
`Junction Board
`
`All the parallel resistances can be assumed equivalent, so
`the equation becomes:
`1
`Total _ Array _R
`
`Θ
`
`−
`Junction Board
`
`=
`
`N
`LED _Emitter _R
`
`Θ
`
`−
`Junction Board
`
`Typical values for RΘ are shown in Table 2.
`
`or:
`
`Table 2 Typical LUXEON Thermal Resistance
`
`LUXEON Power
`Light Sources
`(RRΘJJ(cid:2)(cid:2)BB) MCPCB
`Mounted
`Level 2
`
`LUXEON Emitter
`(RRΘΘJJ(cid:2)(cid:2)BB) MCPCB
`Mounted
`Level 1
`
`Enter Description
`
`17°C/W
`
`Batwing (all colors)
`Lambertian (Green, Cyan,
`Blue, Royal Blue)
`Lambertain (Red,
`18°C/W
`20°C/W
`Red(cid:2)orange, Amber)
`°C/W = °Celcius (ΔT) / Watts (Pd)
`Note: Consult current data sheet for RΘJ(cid:2)S and RΘJ(cid:2)B
`
`15°C/W
`
`C. Thermal Resistance of Multiple LUXEON
`Products
`The total system thermal resistance of multiple(cid:2)emitter
`LUXEON Products such as the LUXEON Line, Ring or
`multiple Stars can be determined using the parallel
`thermal resistance model as shown in Figure 2. In this
`model, each emitter is represented by individual, parallel
`thermal resistances.
`
`Equation 4. Multiple Emitter to Single Emitter
`Thermal Resistance Relation
`
`Total_Array_R
`
`Θ
`
`−
`Junction Board
`
`=
`
`LED _Emitter _R
`N
`
`Θ
`
`−
`Junction Board
`
`Where:
`LED Emitter RΘJunction(cid:2)Board = RΘJunction(cid:2)Slug + RΘSlug(cid:2)Board
`N = Number of emitters
`
`For example, in a LUXEON Line, there are 12 emitters,
`N=12. The LUXEON Line uses a batwing emitter; therefore,
`the Total Array RΘJ(cid:2)B is: (17°C/W)/12 = 1.42°C/W.
`
`The Total Array RΘJunction(cid:2)Ambient(J(cid:2)A) for the LUXEON Line is:
`
`Total_Array_ RΘJunction(cid:2)Ambient=1.42 + RΘBoard(cid:2)Ambient
`
`The Total Array Dissipated Power must be used in any calcu(cid:2)
`lations when using a Total Array thermal resistance model.
`The Total Array Dissipated Power is the sum of VF * IF for all
`the emitters.
`
`Equation 5. Thermal Resistance of a Multiple Emitter Array
`Δ
`T
`
`Total Array R
`
`Θ =
`
`J(cid:2)A
`
`P
`d _ Total
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`3
`
`Where:
`ΔT
`Pd_Total
`
`= TJunction (cid:2) TAmbient (°C)
`= Total Array Dissipated Power (W)
`
`SEC et al. v. MRI
`SEC Exhibit 1024.003
`IPR 2023-00199
`
`

`

`Inputs/Output of the Thermal
`Model
`You can use a thermal model to predict the junction temper(cid:2)
`ature (TJ) for your application. This section discusses setting
`a goal for a maximum TJ as well as the variables in the right(cid:2)
`hand(cid:2)side of Equation 3 below. You can use variables in the
`thermal model as control factors in your application design.
`
`TJunction = TAmbient + (Pd)(RΘJunction(cid:2)Ambient)
`
`A. Set Limit for Junction Temperature (TJ)
`Good thermal design incorporates TJ limits based on three
`factors:
`Light output with TJ rise
`1.
`Color shift with TJ rise
`2.
`Reliability
`3.
`Consult LUXEON Custom Design Guide for more detailed
`information on light output and color shift with rise in TJ.
`
`11.. LLiigghhtt OOuuttppuutt wwiitthh TTeemmppeerraattuurree RRiissee
`LEDs experience a reversible loss of light output as the TJ
`increases. The lower the TJ is kept, the better the luminous
`efficiency of the product (i.e. the better the light output). Light
`output from red, red(cid:2)orange and amber emitters (based on
`AlInGaP LED technology) are more sensitive to increases in
`junction temperature than other colors.
`
`An example of light output loss associated with temperature
`rise occurs with traffic signals. Signals that are simply retro(cid:2)
`fitted with LED sources may not account adequately for heat
`dissipation. As temperatures rise during the day, the signals
`may dim. Redesigning the signal housing to provide airflow
`to cool the components alleviates this condition.
`
`The chart on the LUXEON product data sheet will help you
`determine a maximum TJ based on the light output require(cid:2)
`ments of your application.
`
`22.. CCoolloorr SShhiifftt wwiitthh TTeemmppeerraattuurree RRiissee
`Emitter color can shift slightly to higher wavelengths with TJ
`rise. Shift values quantifying this effect are included in the
`LUXEON Custom Design Guide. Red, Red(cid:2)Orange and
`Amber color emitters are the most sensitive to this effect,
`although the human eye is more sensitive to color changes
`in the amber region. The importance of this effect depends
`on the color range requirements for the application. If the
`allowed color range is very small, you will need to account
`for color shift when setting your maximum TJ goal.
`
`33.. RReelliiaabbiilliittyy--BBaasseedd TTeemmppeerraattuurree RRaattiinnggss
`To ensure the reliable operation of LUXEON Power Light
`Sources, observe the absolute maximum thermal ratings for
`the LEDs provided in Table 1. The maximum TJ is based on
`the allowable thermal stress of the silicone encapsulate that
`surrounds die.
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`4
`
`Table 1. Maximum Thermal Ratings.
`
`Parameter
`
`LED Junction Temperature
`Aluminum(cid:2)Core PCB Temperature
`Storage/Operating Temperature:
`
`Maximum
`
`120
`105
`
`LUXEON Products without optics
`(Star, Star/C)
`
`(cid:2)40 to 105
`
`LUXEON Products with optics
`(Star/O, Line, Ring)
`
`(cid:2)40 to 75
`
`B. Assess Ambient Temperature Conditions
`The designer must take into account the maximum ambient
`temperature (TA ) the LUXEON Power Light Source will expe(cid:2)
`rience over its lifetime. In most cases, you can use product
`standards to determine the worst case TA. Otherwise, use
`representative maximum TA measurements. Please note that
`the ambient temperatures should include other sources of
`heat such as electronics or heating due to sun exposure.
`
`C. Power Dissipated
`The dissipated power (Pd) can be determined as the forward
`voltage (Vf) of the emitter times the forward current (If). The
`portion of power emitted as visible light (about 10%) is negli(cid:2)
`gible for thermal design.
`
`D. Add Heat Sink to Model
`The RΘB(cid:2)A component of RΘJ(cid:2)A (see Figure 1A) represents
`the heat sink and attachment interface. The responsibility
`for the proper selection of the heat sink thermal resistance,
`RΘB(cid:2)A, lies with the engineer using the product. A process
`for selecting RΘB(cid:2)A is explained in the examples that follow.
`Many resources exist to assist with this selection. Some are
`listed in the resources section at end of this document. The
`following section provides additional guidance to help you
`determine the most suitable heat sink for your application.
`
`Heat Sink Characterization
`
`A. Explanation of Data Charts
`11.. TTeesstt SSeett UUpp
`We tested some typical heat sink configurations on LUXEON
`Stars and Floods including both finned and flat heat sinks.
`We used the following test conditions: free (or natural)
`convection environment with no fan (Figures 3A, 3B, 3C and
`3D) and forced convection in a small wind tunnel (Figure 3E).
`The LUXEON Stars tested did not have optics. The optics do
`not affect the RΘJ(cid:2)B of the LUXEON emitter; however,
`depending on the orientation, they may affect the convection
`flow over the attached heat sink.
`
`SEC et al. v. MRI
`SEC Exhibit 1024.004
`IPR 2023-00199
`
`

`

`22.. HHeeaatt SSiinnkk CChhaarraacctteerriizzaattiioonn CChhaarrtt FFoorrmmaatt
`The following charts (Figures 4 to 9) are intended to guide
`the design engineer in selecting the size and type of heat
`sink required for an application. The charts for 25 mm
`spaced emitters in Figures 4 to 8 show RΘB(cid:2)A on the y(cid:2)axis
`vs. heat sink area required per emitter on the x(cid:2)axis. The
`chart for densely spaced emitters in Figure 9 shows RΘB(cid:2)A
`vs. heat sink area required for the entire array. The heat sink
`type and test set(cid:2)up (Figures 3A to 3E) is referenced in the
`title and discussion of each chart.
`
`33.. DDeeffiinniittiioonn ooff HHeeaatt SSiinnkk SSiizzee
`The following charts quantify heat sink size in two ways. The
`term "exposed surface area" is the sum total of all surfaces
`of the heat sink exposed to convection. The "footprint area"
`quantifies the projected area of the heat sink as shown in
`following diagram.
`
`A finned heat sink can fit more exposed surface area in a
`given foot print than a flat heat sink.
`
`
`
`Foot Print Area
`
`Flat Heat Sink
`
`Finned Heat Sink
`
`B. Heat Sink Characterization Charts (cid:2) 25mm
`Emitter Spacing
`When LUXEON emitters are spaced at least 25 mm apart,
`each acts as a discrete heat source. The charts in figures 4
`to 8 will help you size heat sinks for the LUXEON Star, Line
`and Ring as well as custom boards with individual emitters
`spaced 25 mm or further apart. These charts should also be
`helpful in characterizing heat sinks for custom boards with
`emitter spacing as dense as 20 mm. For boards with more
`densely spaced emitters, use the chart in Section C. The
`following in Figures 4 to 8 show RΘB(cid:2)A vs. heat sink area
`required per emitter in your application.
`
`R2 = 0.9798
`
`35.00
`
`30.00
`
`25.00
`
`20.00
`
`15.00
`
`10.00
`
`5.00
`
`0.00
`
`R THETA b-a (DEG C/W)
`
`0
`
`1
`
`8
`7
`6
`5
`4
`3
`2
`FOOT PRINT AREA (=EXPOSED SURFACE AREA) - in2
`
`9
`
`10
`
`Flat Heat Sink, 0.09" (2.3 mm) Horizontal on insulating foam
`Set(cid:2)up in Figure 3C. Solid Line: Linear Fit of Data
`
`Figure 4. RΘΘBBooaarrdd(cid:2)(cid:2)AAmmbbiieenntt per Emitter vs. Foot Print Area
`
`
`
`Vertical Supports
`
`Fins
`vertical
`
`Figure 3.A. Finned Horz.
`
`Figure 3.B. Finned Vert.
`
`Insulating foam
`
`
`
`
`
`Figure 3.C. Flat Horz.
`
`Figure 3.D. Flat Vert.
`
`Wind tunnel
`
`Fan
`
`HS fins parallel
`to forced air flow
`
`Figure 3.E. Finned Horz. in Wind Tunnel
`
`We tested two types of heat sinks: finned heat sinks and flat
`plates. All heat sinks were aluminum, which is typically the
`best choice because of its excellent thermal conductivity and
`ready, low(cid:2)cost availability. We tested several different sizes
`of flat heat sinks and two sizes of finned heat sinks.
`
`We tested some samples in free convection oriented both
`horizontally and vertically, as illustrated in Figures 3B, 3C
`and 3D.
`
`Finned heat sinks were tested in a small wind tunnel
`enclosed in a control volume. Figure 3E shows the forced
`air set(cid:2)up. We used the same set(cid:2)up to characterize the
`finned heat sinks in free convection by turning the fan off
`(Figure 3A).
`
`We suspended the finned heat sink so that air could circulate
`underneath it.
`
`We used mechanical fasteners to mount the LUXEON Stars.
`The mounting surface of the heat sink was smooth and
`lightly polished. We did not use thermal grease.
`
`We ran all tests in a closed volume test box to control the
`free convection and to improve repeatability. We made all
`measurements at steady state conditions. Initial ambient
`conditions were nominally 25°C, but the ambient tempera(cid:2)
`ture increased as the LEDs reached steady(cid:2)state
`temperatures.
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`5
`
`SEC et al. v. MRI
`SEC Exhibit 1024.005
`IPR 2023-00199
`
`

`

`(horizontal on low(cid:2)conducting insulating foam) configurations
`of a flat heat sink. Most applications probably fall some
`where in between.
`
`When selecting a heat sink for your application, you will need
`to determine the most comparable condition. You will also
`need to assess other factors that might make the RΘB(cid:2)A of
`the larger or smaller than the extremes shown in Figure 5.
`Mounting the heat sink to a conductive surface or at a 45°
`angle, for example, are both factors that would reduce the
`RΘB(cid:2)A compared to the horizontal orientation in Figure 5.
`
`
`
`
`
`
`
`35.00
`
`30.00
`
`25.00
`
`20.00
`
`15.00
`
`10.00
`
`5.00
`
`0.00
`
`R THETA - DEG C/W
`
`0
`
`10 11 12 13
`9
`8
`7
`6
`5
`4
`3
`2
`1
`SURFACE AREA EXPOSED TO FREE CONVECTION - in 2
`
`Flat Heat Sink
`
`Finned Heat Sink
`
`Figure 6.
`
`RΘBoard(cid:2)Ambient per Emitter in Free Conv.
`Horizontal Flat Heat Sink (cid:2) Set(cid:2)up Fig. 3A vs. Horizontal Finned Heat Sink (cid:2)
`Set(cid:2)up Fig. 3C
`
`Aavid Heat Sink #65245
`Total surface area = 25 in2
`
`
`
`
`
`
`
`35.00
`
`30.00
`
`25.00
`
`20.00
`
`15.00
`
`10.00
`
`5.00
`
`0.00
`
`R THETA - DEG C/W
`
`0
`
`1
`
`2
`
`7
`6
`5
`4
`3
`FOOT PRINT AREA - in 2
`
`8
`
`9
`
`10
`
`Flat Heat Sink
`
`Finned Heat Sink
`
`Figure 7.
`
`55.. FFiinnnneedd ((FFiigg.. 33AA)) vvss.. FFllaatt HHeeaatt SSiinnkkss ((FFiigg.. 33CC)) iinn FFrreeee
`((NNaattuurraall)) CCoonnvveeccttiioonn
`We tested two finned heat sinks with identical 2 in2 foot print
`areas, but different exposed surface areas. Increasing the
`number and length of fins on the heat sink increases the
`surface area. The fin spacing was about 0.25 in. Figure 6
`shows RΘB(cid:2)A per exposed surface area for finned heat sinks
`and flat heat sinks. The heat sinks plotted in Figure 6 are
`horizontal (Set(cid:2)up Figure 3A for finned, Figure 3C for flat).
`
`The finned heat sinks required more exposed surface area
`for a given RΘB(cid:2)A compared to the flat heat sinks. This shows
`that a flat heat sink can be effective in thermally managing
`LUXEON Power Light Sources with 25 mm emitter spacing.
`
`In order to fully utilize the surface area on the finned heat
`sinks, the fins must lie in parallel with the convection airflow.
`The finned heat sinks would probably have a slightly lower
`RΘB(cid:2)A if oriented vertically (Set(cid:2)up Figure 3B).
`
`22.. HHoorriizzoonnttaall,, FFllaatt HHeeaatt SSiinnkk ((FFiigg.. 33CC)) iinn FFrreeee ((NNaattuurraall))
`CCoonnvveeccttiioonn
`As exposed surface area increases, thermal resistance
`decreases. Figure 4 illustrates this relationship with a flat,
`horizontal heat sink, which is close to linear.
`
`In the horizontal orientation, only a single, upward(cid:2)facing
`surface of the flat heat sink is exposed to convection. The
`bottom surface contacts the insulating foam. This is the least
`efficient orientation for convection, resulting in the highest
`expected thermal resistance.
`
`33.. HHoorriizzoonnttaall ((FFiigg.. 33CC)) vvss.. VVeerrttiiccaall OOrriieennttaattiioonn ((FFiigg.. 33DD)) iinn
`FFrreeee CCoonnvveeccttiioonn
`When the flat heat sink is oriented vertically, the surface area
`doubles, as both sides are exposed to free convection. This
`results in a more efficient heat sink within the same foot print
`area. This effect is illustrated with respect to the foot print
`area in Figure 5.
`
`Horiz. Orientation -- Exposed Surf.
`Area=1 x Foot Print Area
`Vert. Orientation -- Exposed Surf.
`Area=2 x Foot Print Area
`
`35.00
`
`30.00
`
`25.00
`
`20.00
`
`15.00
`
`10.00
`
`5.00
`
`0.00
`
`R THETA b-a (DEG C/W)
`
`0
`
`1
`
`2
`
`3
`
`6
`5
`4
`FOOT PRINT AREA - in2
`
`7
`
`8
`
`9
`
`10
`
`Flat Heat Sink 0.09" (2.3 mm) Thick (cid:2) Horz. Set(cid:2)up Fig. 3C (cid:2) Vert. Set(cid:2)up Fig 3D
`
`Figure 5. RΘΘBBooaarrdd(cid:2)(cid:2)AAmmbbiieenntt per Emitter in Free Convection
`Vs. Foot Print Area.
`
`In the vertical orientation, the thermal resistance decreases
`noticeably as the exposed surface area doubles. The total
`surface area of the horizontal heat sink equals the foot print
`area. For the vertical heat sink, the total surface area is
`double the foot print area.
`
`The vertical heat sink is also more efficient due to the nature
`of free convection. Bouyant, warm air moving over a vertical
`surface is more efficient than air that moves vertically away
`from a horizontal surface.
`
`As the foot print areas approach 9in2, the RΘB(cid:2)A of the two
`orientations begin to converge. This indicates that as foot
`print areas approach 9in2 per emitter, heat sink orientation is
`not influencial. Also, with areas greater than 9in2 per emitter,
`there are diminishing reductions in the RΘB(cid:2)A. The lower limit
`for RΘB(cid:2)A with increasing area will approach about 10 to 11
`°C/W.
`
`44.. RRaannggee ooff EEffffiicciieennccyy wwiitthh FFllaatt HHeeaatt SSiinnkkss
`The two conditions shown in Figure 5 represent the most
`efficient (vertical, 2 convective surfaces) and least efficient
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`6
`
`SEC et al. v. MRI
`SEC Exhibit 1024.006
`IPR 2023-00199
`
`

`

`We characterized three types of heat sinks using 12 and 18
`emitter LUXEON Floods. The results are shown in Figure 9.
`All heat sinks were vertically orientated with free convection
`on all sides. We tested both flat plate (see Figure 3D test set(cid:2)
`up) and finned heat sinks (see Figure 3B.)
`
`Figure 9 should be most useful in sizing heat sinks for custom
`applications that use ten to twenty emitters. However, it can
`also be used as a rough guide for sizing heat sinks for appli(cid:2)
`cations with about 3 to 20 densely spaced emitters.
`
`Attachment to Heat Sinks
`
`A. Mechanical Attachment
`We recommend mounting LUXEON Power Light Sources
`(Level 2 products) directly to a heat sink with mechanical
`fasteners for best performance. You can use fasteners when
`mounting to a smooth machined or extruded metal surface.
`The addition of thermal grease (e.g. Wakefield Eng. Thermal
`Compound) can minimize air gaps and improve thermal
`contact to castings and uneven surfaces.
`
`B. Adhesive Attachment
`Tapes and adhesives can aid in thermal contact with most
`surfaces. Philips Lumileds utilizes Amicon E 3503(cid:2)1 as the
`epoxy for attaching LEDs onto boards. The thermal proper(cid:2)
`ties of Amicon and a double sided Bergquist tape are shown
`in Table 3.
`
`Adhesives are available from many sources, such as, Epo(cid:2)
`Tek, Dow Corning, 3M, and others, however, the customer
`must perform a thorough evaluation of the adhesive in terms
`of thermal performance, manufacturability, lumen mainte(cid:2)
`nance, and mechanical durability.
`
`Furthermore, Philips Lumileds does not recommend adhe(cid:2)
`sives containing hydrocarbons such as amine, heptane,
`hexane, and other volatile organic compounds.
`
`66.. FFiinnnneedd HHeeaatt SSiinnkkss RReedduuccee FFoooott PPrriinntt SSiizzee
`The Figure 7 shows RΘB(cid:2)A per foot print area for finned heat
`sinks and flat heat sinks. Each of the finned heat sinks had 2
`in2 footprints. The flat heat sinks have footprints equal to the
`exposed area. A flat heat sink needs about 6 in2 footprint to
`match the RΘB(cid:2)A of a 2 in2 foot print finned heat sink. If foot
`print size is a major design constraint, a finned heat sink
`offers an efficient solution.
`
`The lower limit for RΘB(cid:2)A using a 2 in2 footprint finned heat
`sink is about 10 to 11°C/W. A heat sink typical of this
`performance is an AAVID heat sink extrusion part # 65245.
`A 1.6 in length of this heat sink extrusion has 25 in2 total
`surface area with a 2 in2 footprint. RΘB(cid:2)A for this heat sink is
`plotted in Figure 7. Looking at Figure 5, a 9 in2 vertical flat
`heat heat sink (18 in2 total surface area) would have about
`the same RΘB(cid:2)A.
`
`FAN OFF (FREE)
`FAN ON (FORCED)
`
`25.00
`
`20.00
`
`15.00
`
`10.00
`
`5.00
`
`0.00
`
`R THETA - DEG C/W
`
`4 5 6 7 8 9 10 11 12 13 14
`0 1 2 3
`SURFACE AREA EXPOSED TO CONVECTION -
`in2
`
`Figure 8. RΘΘBBooaarrdd(cid:2)(cid:2)AAmmbbiieenntt per Emitter (cid:2) Free Conv. (Test Set(cid:2)up Fig. 3A)
`vs. Forced Conv. (Test Set(cid:2)up Fig. 3E) (cid:2) 42f/min (12.8m/min) Air
`Flow with Fan On
`
`BB.. HHeeaatt SSiinnkkss iinn FFrreeee CCoonnvveeccttiioonn -- DDeennssee EEmmiitttteerr SSppaacciinngg
`When LUXEON emitters are densely packed, they function
`as a single heat source. This chart will help you characterize
`the LUXEON Flood as well as custom Level 2 Boards with
`emitter spacing between 9 and 13 mm. This chart can also
`be used to characterize heat sinks for clustered emitters,
`with spacing up to about 19 mm. For wider spacing, use the
`charts in Section B. The following chart in Figure 9 shows
`the Total Array RΘB(cid:2)A vs. heat sink area required for the total
`array. It is the total array RΘB(cid:2)A shown in Figure 2, which is
`the thermal resistance model for multiple emitter products.
`
`1. Flat HS (Fig 3C)
`
`2. Flat HS (Fig. 3C)
`
`3.Finned HS (Fig. 3B)
`
`140
`120
`100
`80
`Surface Area of Heat Sink -- in2
`
`160
`
`4.00
`
`3.50
`
`3.00
`
`2.50
`
`2.00
`
`1.50
`
`1.00
`60
`
`R Theta Board-Amb - Deg C/W
`
`Figure 9. High Density Emitter Heat Sink
`Total Array Thermal Resistance (Board to Ambient)
`vs. Surface Area Exposed
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`7
`
`SEC et al. v. MRI
`SEC Exhibit 1024.007
`IPR 2023-00199
`
`

`

`TTaabbllee 33 TTyyppiiccaall TThheerrmmaall RReessiissttaanncceess ooff GGlluueess aanndd TTaappeess..
`LLeevveell 11 MMoouunnttiinngg (cid:2)(cid:2) EEmmiitttteerr
`SSlluugg ttoo BBooaarrdd
`AAddddeedd RRΘΘsslluugg(cid:2)(cid:2)bbooaarrdd ((°°CC//WW))
`ppeerr EEmmiitttteerr
`00..004444 iinn22 ((2288 mmmm22))
`SSlluugg AArreeaa
`4.5
`
`GGlluueess
`approx.
`0.05”
`thick
`TTaappeess
`
`AAddhheessiivveess
`Amicon E3503(cid:2)1
`
`LLeevveell 22 MMoouunnttiinngg (cid:2)(cid:2) BBooaarrdd
`ttoo HHeeaatt SSiinnkk
`AAddddeedd RRΘΘBBooaarrdd(cid:2)(cid:2)HHeeaatt__SSiinnkk__TToopp ((°°CC//WW))
`ppeerr EEmmiitttteerr
`11 iinn22 ((662255 mmmm22))
`BBooaarrdd AArreeaa
`*
`
`MMaannuuffaaccttuurreerr
`IInnffoorrmmaattiioonn
`Emerson & Cuming(cid:2)Belgium
`Ph: 0032/ 14 57 56 11
`
`Bond Ply 105
`(0.005” thick)
`
`14
`
`3°C/W
`
`The Bergquist Company
`www.bergquistcompany.com
`
`Before selecting an adhesive or interface material be sure to determine its suitability and compatibility with LUXEON, your manu(cid:2)
`facturing processes, and your application. Philips Lumileds uses Amicon 3503(cid:2)1 from Emerson and Cuming. This epoxy may be
`purchased from multiple distributors. Some examples of these distributors may be found in the Philips Lumileds Resource Guide
`at www.philipslumileds.com.
`
`Best Practices for Thermal Design
`• A flat, aluminum heat sink can be as effective as a finned
`heat sink when emitters are spaced at least 25 mm apart.
`• A finned heat sink is an effective solution to minimize foot(cid:2)
`print area.
`• For maximum thermal performance using a flat heat sink,
`allow an exposed surface area of about 9in2 per emitter
`(with 25 mm emitter spacing).
`• A LUXEON Flood requires a flat heat sink with an exposed
`surface area of 36in2 to operate at room temperature
`(25°C).
`• Where practical, use mechanical fasteners to mount heat
`sinks to smooth and flat surfaces.
`
`Evaluating Your Design
`Use the charts in Figures 4 to 9 to approximate the heat sink
`size, as well as its orientation and shape.
`
`To do so, you must first determine the required RΘB(cid:2)A, per
`emitter, given both the thermal and optical requirements of
`your application. Then based on the required RΘB(cid:2)A, you can
`use the data in the charts to define your heat sink require(cid:2)
`ments. General steps for doing this follow.
`
`For single or multi(cid:2)emitter applications with 25mm spacing,
`you can approximate heat sink requirements using Figures 4
`to 8. For applications with dense emitter spacing such as the
`LUXEON Flood, use Figure 9.
`
`A. Steps to Select Minimum Size Heat Sink
`SStteepp 11)) Determine allowable RQJ(cid:2)A
`With TJ as the constraining variable, you can use the
`following equation:
`TJ = TA+(P)(RΘJ(cid:2)A)
`
`Enter the absolute maximum TJ and the worst case oper(cid:2)
`ating conditions TA into the equation. You may need to
`specify a maximum TJ lower than 120°C in order to achieve
`the optical performance required for your application. See
`the LUXEON Custom Design Guide for more information.
`
`The dissipated power per string, P, can be determined by:
`P = (VF)(IF)
`Solve for RΘJ(cid:2)A using:
`
`Θ
`
`Junction
`
`R
`
`−
`
`Ambient
`
`(T
`
`Junction
`
`=
`
`T
`
`Ambient
`
`)
`
`−
`P
`
`Thermal Design Using LUXEON Power Light Sources App Brief AB05 (6/06)
`
`8
`
`SEC et al. v. MRI
`SEC Exhibit 1024.008
`IPR 2023-00199
`
`

`

`SStteepp 22)) Subtract the RΘJ(cid:2)B (found in Table 1, also check
`current product data sheet) of LUXEON emitter from RΘJ(cid:2)A
`to obtain the target RΘB(cid:2)A.
`
`possible to an emitter base (Figure 10). Evaluate the design
`at the expected ambient temperature range, ambient air flow
`and with any additional heat loads.
`
`You can monitor temperatures using a surface probe
`temperature meter, though this is not practical for applica(cid:2)
`tions in enclosures. In general, thermocouples offer the most
`practical temperature monitoring solution.
`
`Recommended thermocouple (TC) attachment:
`1.Locate TCs on the hottest areas of the board. Examples
`are: near the center of a cluster array of emitters or near
`any heat producing electronics.
`
`2.Locate the TCs as close as possible near the base of an
`emitter. Do not mount TC tip on lead traces. Do solder or
`mount TCs to the emitter solder pads.
`
`3.If using small diameter TCs (J(cid:2)type) or adhesive mounted
`TCs, they can be taped flat to the top of the board, with
`the TC tip at the base of the emitter.
`
`4.If using a larger T or K(cid:2)type TC, it may not be possible to
`tape the TC tip flat on the board, which would lead to
`inaccuracies. In this case, drill a hole, just larger than the
`TC dia. in the top of the board, 0.03" deep. (Figure 11)
`Bend the TC tip at right angle. For better contact, dip the
`TC tip in a conductive paste (e.g. Wakefield Eng, Thermal
`Compound). Insert the TC tip and secure the TC wire with
`tape or glue to keep the TC tip fully inserted.
`
`Thermocouple
`
`Figure 10. Location of thermocouple to Monitor TBBooaarrdd.
`
`Drilled Hole
`
`Figure 11. Thermocouple tip inserted in board.
`
`SStteepp 33)) Using the calculated RΘB(cid:2)A as a target, review the
`