The present invention relates in general to lamps based on light-emitting diodes (LEDs) and in particular to procedures for tuning the color of light produced by lamps that include multiple LEDs.
With the incandescent light bulb producing more heat than light, the world is eager for more efficient sources of artificial light. LEDs are a promising technology and are already widely deployed for specific purposes, such as traffic signals and flashlights. However, the development of LED-based lamps for general illumination has run into various difficulties. Among these is the difficulty of mass-producing lamps that provide a consistent color temperature.
As is known in the art, not all white light is the same. The quality of white light can be characterized by a color temperature, which ranges from the warm (slightly reddish or yellowish) glow of standard tungsten-filament light bulbs to the cool (bluish) starkness of fluorescent lights. Given existing processes for LED manufacture, mass-producing white LEDs with a consistent color temperature has proven to be a challenge.
Various solutions have been tried. For example, white LEDs can be binned according to color temperature and the LEDs for a particular lamp can be selected from the desired bin. However, the human eye is sensitive enough to color-temperature variation that a large number of bins is required, with the yield in any particular bin being relatively low.
Another solution relies on mixing different colors of light to produce a desired temperature. For example, an LED lamp can include a number of white LEDs plus some red LEDs. The brightness of the red LEDs can be increased to warm the light to the desired color temperature. Such lamps generally require an active feedback mechanism to maintain the color temperature, in part because the LEDs used are not stable in their color characteristics over time. The active feedback mechanism requires a sensor to detect the light being produced, an analyzer to determine whether the light is at the desired color, and an adjustment mechanism to adjust the relative brightness of the white and red LEDs as needed to maintain the desired color. These feedback-loop elements can be a weak point in the system; for example, if the light sensor drifts over time (as most do), so will the color of the light. In addition, incorporating active feedback components into a lamp drives up the cost of manufacturing (and operating) the lamp.
Embodiments of the present invention relate to techniques for tuning the color of an LED-based lamp to a desired color or color temperature. Particular embodiments are adapted for use with lamps that include two or more independently addressable groups of LEDs that each produce light of a different color or color temperature. The lamps can also include a total-internal-reflection (TIR) color-mixing lens to produce light of a uniform color by mixing the light from the different groups of LEDs. The uniform color or color temperature output from the lamp is tuned by controllably dividing an input current among the groups of LEDs. For lamps using LEDs whose color is stable over time, the tuning can be performed once, e.g., during manufacture and/or factory testing of the lamp, and the lamp can thereafter operate at a stable color temperature without requiring active feedback components.
For example, in some embodiments a lamp includes two distinct groups of white LEDs: one group (“warm white”) that produces white light with a warmer color temperature than is desired and another group (“cool white”) that produces white light with a cooler color temperature than is desired. In such lamps, the color temperature can be tuned by controllably dividing an input current between the warm white group and the cool white group. In some embodiments, an optimal division of the input current can be determined based on a linear relationship between a shift in the fraction of current provided to each group and a shift in color-space coordinates (which correspond to color temperature) that obtains over the relevant (small) region in color space; the process is simple, requiring as few as three measurements, and can be highly automated to facilitate mass production of color-tuned lamps.
In other embodiments, a lamp includes three distinct groups of LEDs, for example, warm white, cool white, and red (other non-white colors can also be used). In some embodiments, tuning between the warm white and cool white groups is performed with the red (or other non-white) LED group turned off. Tuning between the “tuned white” light and the red LED group can then be performed, relying on the fact that as long as the current split between warm white and cool white LEDs does not change, the “tuned white” color will not shift with a shift in total current supplied to the white LEDs. Alternatively, triangular interpolation can be used for tuning, relying on the fact that over a small region in color space, the amount of change in the division of current between two groups of LEDs is linearly related to the amount of change in color-space coordinates.
In still other embodiments, a lamp includes four distinct groups of LEDs, for example, warm white, cool white, red, and green (other non-white colors can also be used; for producing white light, the non-white colors are advantageously complementary). Tuning between the warm white and cool white groups is performed with the non-white LED groups turned off. Tuning between the “tuned white” light and the red and/or green LED groups can then be performed, relying on the fact that as long as the current split between warm white and cool white LEDs does not change, the “tuned white” color will not shift with a shift in total current supplied to the white LEDs. Further tuning of the color can be achieved by adding green to the tuned white/red color. Again, triangular interpolation techniques or other linear interpolation can be used over a small region in color space.
Any number of groups of LEDs can be used. LEDs in different groups advantageously occupy non-overlapping regions of color space, and the target color is intermediate between the color-space regions occupied by the different groups.
Applying processes described herein across a number of lamps allows substantial reduction in the color variation from one lamp to the next. In addition, the tuning process can be confined to a relatively small region in color space such that color shift as a function of current shift from one group of LEDs to another can be modeled as a linear relation. Using linear modeling, the appropriate adjustment for a given lamp can be determined from a small number of measurements. Thus, tuning of a lamp can be accomplished quickly, allowing the tuning process to be incorporated into a mass-production environment.
Additional embodiments of the invention relate to tuning apparatus that provide a high degree of automation for the tuning process, suitable for use in mass-production environments.
One aspect of the invention relates to a method for tuning a color produced by a lamp having multiple groups of LEDs, where each group includes at least one LED. Each group of LEDs produces light having a different color, and a current applied to each group of LEDs is independently variable. According to one tuning method, at least two different testing distributions of a total current among the groups of LEDs are established. For each of the different testing distributions of the total current, a color of light produced by the lamp is measured. A target color is defined, and a desired distribution of the total current is determined based at least in part on the measured colors; the desired distribution of the total current produces light having the target color.
In some embodiments, the groups of LEDs can include a group of warm white LEDs and a group of cool white LEDs. Additional groups of LEDs, including groups of non-white LEDs, such as red and/or green LEDs, can also be included. In some embodiments, the groups of LEDs can include at least two groups of cool white LEDs and at least one group of warm white LEDs.
The lamp can include a total internal reflection lens to mix the light produced by the plurality of LEDs, and the measuring of the color of the light can be based on light exiting a front face of the total internal reflection lens. The measuring can be done by a spectrometer (or other color measuring device) external to the lamp, and the lamp itself need not include a spectrometer or other active feedback components for adjusting color.
Another aspect of the invention relates to a method for controlling a color produced by an emitter having independently-addressable warm white LEDs and cool white LEDs. A first value for a color property of the emitter can be measured under a first operating condition in which a maximum current is supplied to the warm white LEDs and a minimum current is supplied to the cool white LEDs. A second value for the color property of the emitter can be measured under a second operating condition in which the maximum current is supplied to the cool white LEDs and the minimum current is supplied to the cool white LEDs. A third value for the color property of the emitter can be measured under a third operating condition in which approximately half of a total current is delivered to the warm white LEDs and the rest of the total current is delivered to the cool white LEDs; the total current is advantageously equal to a sum of the maximum current and the minimum current. Based on the measured first, second, and third values of the color property and a target value of the color property, operating currents, including a first operating current to be supplied to the warm white LEDs and a second operating current to be supplied to the cool white LEDs, can be calculated. A current controller coupled to the emitter can be configured such that when the first operating current is supplied to the warm white LEDs, the second operating current is supplied to the cool white LEDs.
Another aspect of the invention relates to a method for controlling a color produced by a lamp having independently addressable warm white LEDs and cool white LEDs. A first value of a color property of the lamp can be measured while supplying a total current to the warm white LEDs and no current to the cool white LEDs. A second value of the color property of the lamp can be measured while supplying the total current to the cool white LEDs and no current to the warm white LEDs. A third value of the color property of the lamp can be measured while supplying half the total current to the warm white LEDs and half the total current to the cool white LEDs. A first operating current to be supplied to the warm white LEDs and a second operating current to be supplied to the cool white LEDs to achieve a target value of the color property can be determined, with the total current being equal to a sum of the first operating current and the second operating current. The determination of the first and second operating current can be based on the measured first, second and third values of the color property and a proportionality constant that linearly relates a unit of change in a difference between the first and second operating currents to an amount of change in the color property. A control circuit of the lamp can be configured such that when the first operating current is supplied to the warm white LEDs, the second operating current is supplied to the cool white LEDs.
The following detailed description together with the accompanying drawings will provide a better understanding of the nature and advantages of the present invention.
Embodiments of the present invention relate to techniques and apparatus for tuning the color of an LED-based lamp to a desired color temperature. Particular embodiments are adapted for use with lamps that include two or more independently addressable groups of LEDs that each produce light of a different color or color temperature. The lamps can also include a total-internal-reflection (TIR) color-mixing lens to produce light of a uniform color by mixing the light from the different groups of LEDs. The uniform color or color temperature output from the lamp is tuned by controllably dividing an input current among the groups of LEDs. For lamps using LEDs whose color is stable over time, the color tuning can be performed once, e.g., during manufacture and/or factory testing of the lamp, and the lamp can thereafter operate at a stable color temperature without requiring active feedback components.
Embodiments for tuning lamps with two independently addressable groups of LEDs will be considered first, after which extensions to lamps with larger numbers of groups. As used herein, a “group” of LEDs refers to any set of one or more LEDs that occupies a defined region in color space; the regions are defined such that regions occupied by different groups in the same lamp do not overlap. The lamp is advantageously designed such that the current supplied to each group of LEDs can be controlled independently of the current supplied to other LEDs, and the groups are thus said to be “independently addressable.”
Within housing 102 is an LED package 104. Package 104 includes a substrate 106 on which are mounted individual LEDs 108. Each LED 108 can be a separate semiconductor die structure fabricated to produce light of a particular color in response to electrical current. In some embodiments, each LED 108 is coated with a material containing a color-shifting phosphor so that LED 108 produces light of a desired color. For example, a blue-emitting LED die can be coated with a material containing a yellow phosphor; the emerging mixture of blue and yellow light is perceived as white light having a particular color temperature.
In some embodiments, lamp 100 also includes a control circuit 116 that controls the power provided from an external power source (not shown) to LEDs 108. As described below, control circuit 116 advantageously allows different amounts of power to be supplied to different LEDs 108.
A primary lens 110, which can be made of glass, plastic or other optically transparent material, is positioned to direct light emitted from LEDs 108 into secondary optics 112. Secondary optics 112 advantageously include a total-internal-reflection (TIR) lens that also provides mixing of the colors of light emitted from LEDs 108 such that the light beam exiting through front face 114 has a uniform color. Examples of suitable lenses are described in U.S. Patent Application Pub. No. 2010/0091491; other color-mixing lens designs may also be used. As described below, tuning is advantageously performed based on the color of light exiting through front face 114 of TIR lens 112.
In some embodiments LEDs 108 advantageously include both “warm” and “cool” white LEDs. An example is illustrated in
To facilitate achieving a desired color temperature, the LEDs 108 of lamp 100 are advantageously connected such that cool white LEDs 108a-f and warm white LEDs 108g-l are independently addressable, i.e., different currents can be supplied to different LEDs.
In
Other addressing schemes can also be used; for example, each of the LEDS 108a-l can be independently addressable.
It will be appreciated that lamp 100 described herein is illustrative and that variations and modifications are possible. In one embodiment, lamp 100 can be similar to a LuxSpot™ lamp, manufactured and sold by LedEngin Inc., assignee of the present invention. Those skilled in the art with access to the present teachings will recognize that any lamp that has independently addressable warm white and cool white LEDs can also be used; thus, details of the lamp are not critical to understanding the present invention.
In accordance with some embodiments of the present invention, the currents IC and IW (shown in
Further, it should be noted that in the embodiment shown, lamp 100 does not include any active feedback components. As described below, lamp 100 can be placed into a tuning apparatus and color-tuned during production. Thereafter, lamp 100 can be configured to operate at the desired color temperature simply by maintaining the division (or distribution) of current determined in the tuning process. Provided that the LEDs in lamp 100 can maintain a stable color temperature over time, no further tuning or active feedback is needed during normal lamp operation. Since active feedback is not needed, the cost of manufacture can be reduced as compared to lamps that require active feedback to maintain a stable color temperature.
To understand the tuning process, it is useful to begin by considering the behavior of untuned lamps.
The portion of the CIE color space represented encompasses much of the range associated with white light. The various data points (black diamonds) represent colors measured from a number of LED-based lamps having independently addressable warm white and cool white LED groups, e.g., as described above with reference to lamp 100, under various operating conditions.
More specifically, for purposes of these measurements, a total current ITOT of 1000 mA was supplied to the lamp, and the constraint IC+IW=ITOT was maintained. “Cool white” data, represented by points 302, was measured for each lamp by setting IC=ITOT and IW=0. “Warm white” data, represented by points 304, was measured for each lamp by setting IC=0 and IW=ITOT. “Balanced” data, represented by points 306, was measured by setting IC=IW=0.5* ITOT.
A target color is represented by circle 308, and the goal is to produce colors as close to this target as possible. As can be seen, merely applying equal current to the warm white and cool white LEDs results in balanced data points 306 being scattered about target 308. While the balanced colors are more consistent across different lamps than can readily be obtained by using LEDs of a single white color, further improvement in color consistency can be achieved by tuning the relative currents IC and IW (and consequently the color) on a per-lamp basis. Such tuning in a typical case results in unequal currents being supplied to the warm white and cool white LEDs, with the currents being selected to reduce the lamp-to-lamp variation by bringing the light from each lamp closer to target 308.
Blending light of the colors corresponding to points 402 and 404 results in a color somewhere along line 410. Thus, it may not be possible to produce blended light with a color corresponding exactly to single-color point 408. Accordingly, the aim instead is to reach the closest point to point 408 that is on line 410, i.e., “tuned” point 412 at coordinates (xt, yt). In a typical case (xt, yt) and (xB, yB) are not the same, and (xt, yt) may be different for different lamps; thus, tuning on a per-lamp basis is desired.
In general, the relationship between a change in the relative currents (measured, e.g., as IW/IC) supplied to the warm and cool LEDs and the resulting shift in color temperature is nonlinear. Further, the magnitude of the shift in color temperature resulting from a given change in relative current varies from one lamp to another.
However, as illustrated in
As
is very nearly constant for different lamps. In one embodiment, α is about 0.0008052 mA−1. In other embodiments, the applicable ratio α can be determined by measuring a sampling of lamps.
Accordingly, referring to
where α is the constant ratio defined in Eq. 1. Setting
I
C0=0.5*(ITOT+Iδ) (Eq. 3)
and
I
W0=0.5*(ITOT−Iδ) (Eq. 4)
can be expected to produce light of color (xt, yt).
Based on the foregoing, a rapid tuning procedure can be applied to tune an LED lamp.
At block 602, the input current to the LED lamp (or settings on potentiometers within the lamp) is adjusted such that IC=ITOT and IW=0. At block 604, the color of the resulting light is measured, e.g., as (xC, yC). Conventional spectrometers or other known instruments can be used for this measurement and all color measurements described herein.
At block 606, the input current to the LED lamp (or settings on potentiometers within the lamp) is adjusted such that IW=ITOT and IC=0. At block 608, the color of the resulting light is measured, e.g., as (xW, yW).
At block 610, the input current to the LED lamp (or settings on potentiometers within the lamp) is adjusted such that IC=IW=0.5*ITOT. At block 612, the color of the resulting light can be measured, e.g., as (xB, yB).
At block 614, a current shift Iδ that will produce a tuned color (xt, yt) is computed using the linear relation observed above. More specifically, (xt, yt) can be computed as the nearest point to (xs, ys) that is on the line between measured (xC, yC) and (xW, yW) (see
x
t
=x
C
+u(xW−xC)
y
t
=y
C
+u(yW−yC) (Eq. 5)
where
Then, Iδ can be computed using Eq. 2.
At block 616, the operating currents IC0 and IW0 can be determined using Eqs. 3 and 4.
At block 618, to confirm the computation, operating currents IC0 and IW0 can be applied to the lamp. The resulting color can be measured and compared to the predicted (xt, yt).
It will be appreciated that process 600 is illustrative and that variations and modifications are possible. Steps described as sequential may be executed in parallel, order of steps may be varied, and steps may be modified, combined, added or omitted. In addition, while the embodiment described takes the measurements used to calculate Iδ at the “extreme” points and the “mid” point of possible current splits, those skilled in the art will appreciate that other points could also be used. For example, if desired, measurements could be taken at 10/90 and 90/10 current splits, and at the midpoint some other intermediate point. As long as three distinct measurements at three distinct current splits are made, the process above can be used to determine a current split to achieve a desired tuned color temperature (or color). In some embodiments, the target value is advantageously close to the midpoint between the warm and cool color temperatures, as this allows the lamp to operate at highest efficiency (i.e., maximum lumens per LED die). This can be reliably achieved by selecting the warm white and cool white LEDs such that the target value is near the midpoint; in one embodiment, the warm white and cool white LEDs are selected such that the tuned color will always be reached with a warm/cool current split somewhere in the range between 30/70 and 70/30. However, no particular target value is required; tuning can be achieved at any point that lies between the two groups in color temperature space.
In some embodiments, process 600 can also include further fine-tuning of the color. For example, a least-squares fit can be used to determine the distance between the target point on the blackbody curve and the line between measured xC and xW, and this can be used to modify the current split to fine-tune the color.
In
It is noted that, based on the degree of scatter, the improvement is greater in the CIE-x coordinate than in CIE-y. Since the human eye is less sensitive to change in CIE-y, tuning based on CIE-x (e.g., using process 600) is found to yield satisfactory results.
Tuning as described herein can be practiced with any lamp with an emitter having independently addressable groups of warm white and cool white LEDs. In some embodiments, selection of the LEDs for the warm white and cool white groups can optimize tunability. For example,
Using the processes described above, a lamp whose emitter contains warm white LEDs from bin 802 and cool white LEDs from bin 804 can be tuned, e.g., to a point along line 806. The exact point will in general depend on the variations in particular LEDs in a given lamp; dotted lines 808 indicate some of the possibilities. As indicated, even with a relatively large manufacturing tolerance for the LEDs, a small tuned projection (line 806) can be achieved.
In other embodiments, rather than selectively choosing LEDs to produce a given color temperature, the manufacturer can produce an emitter with one group of LEDs above the blackbody curve and another group of LEDs below the blackbody curve without targeting a particular color temperature. The lamp can be tuned to a point on the blackbody curve using techniques described above, and thereafter the lamps can be binned according to their tuned color temperature.
For purposes of providing lamps with a desired color, blackbody curve 902 can be segmented into a number of bins as indicated by boxes 910. The size of the bins can be chosen such that variations in color are imperceptible or nearly so. Each lamp can be assigned to a bin based on the point on blackbody curve 902 to which it tunes.
In some embodiments, further improvements in tuning can be provided by using lamps that include more than two independently addressable groups of LEDs of different colors. For example, in addition to cool white and warm white, it is possible to include red and/or green LEDs in an emitter.
By way of illustration of a three-group embodiment,
Point 1110, at coordinates (xt1, yt1), represents a tuned color for the warm white and cool white LED groups. By performing process 600 described above (or a similar process), with no current supplied to the red LED group, a suitable division of current between the warm white and cool white groups (operating currents IW0 and IC0) can be determined, such that light of color (xt1, yt1) is produced. Thereafter, current distribution between the white LEDs and the red LED can be tuned to bring the color closer to (xs, ys), while maintaining the relative currents between the warm white and cool white LEDs. Specifically, a constant current ITOT can be divided as follows:
I
TOT
=I
R+β(IW0+IC0), (Eq. 7)
for 0≦β≦1. That is, during this phase of tuning, the currents supplied to the warm white and cool white LED groups are held in a fixed relation to each other (i.e., IW0/IC0 is constant) so that the effective color temperature (“net white”) of the warm white and cool white groups is constant, and the total current to the white LED groups (i.e., β(IW0+IC0)) is adjusted relative to the current IR to the red LED group, keeping ITOT constant. A process similar to process 600 can be used to determine values for IR and β such that the resulting color is at the closest point along line 1112 to point (xs, ys), i.e., point 1114, which has coordinates (xt2, yt2). For tuning between the net white color and the red color, a different constant α′ would be used.
Next, tuning can be performed between the net white color and the red LED group. More specifically, at block 1204, IR in Eq. 7 is set to zero, β is set to 1, and a color (xβ, yβ) is measured. (This may be the same color as (xt1, yt1) in
As with process 600, it is not necessary to use the “endpoint” cases at blocks 1204 and 1206. In a typical embodiment, the target color (xs, ys) lies on the well-known blackbody curve in color space, line 1116 between points (xC, yC), (xW, yW) is close to the blackbody curve, and red color point (xR, yR) is far from the blackbody curve. In such cases, (xt1, yt1) is already quite close to (xs, ys), and a small contribution from the red LED is used to fine-tune the color. Thus, a better linear interpolation may be obtained by using an intermediate value in place of the IR=1 endpoint at block 1206. For example, it may be sufficient to use (IR=0.3*ITOT, β=0.7).
Process 1200 is particularly effective in embodiments where the red LED color is situated in color space such that moving the color along line 1112 in
As
In this embodiment, the three groups of LEDs include a first cool white group 1302 with a color temperature “above” the blackbody curve (dashed line 1308), a second cool white group 1304 with a color temperature “below” blackbody curve 1308, and a warm white group 1306. By adjusting the relative current distributed to LED groups 1302, 1304, and 1306, the color can be tuned to any point within triangle 1310. In some embodiments, tuning to a range of points on blackbody curve 1308 (e.g., color temperatures of about 4500 K to about 2800 K) with high precision can be achieved. Thus, for example, a desired color temperature (xs, ys) (point 1312) on blackbody curve 1308 can be produced by tuning.
Next, at block 1404, a division of the cool-LED current ICO between groups 1302 and 1304 is optimized. Holding ICO and IWO constant, IC1 and IC2 can be varied to shift the color toward the desired point (xs, ys).
The embodiments of
In some embodiments, more than three groups of LEDs can be used. For example, some embodiments may have two warm white groups (bracketing the blackbody curve) and two cool white groups (also bracketing the blackbody curve), for a total of four groups of LEDs. In still other embodiments, both red and green LED groups can be provided in addition to the warm white and cool white groups, thus providing four groups of LEDs.
The process for tuning with four groups can be similar to process 1200 (
It will be appreciated that the tuning processes for multiple groups of LEDs described herein are illustrative and that variations and modifications are possible. Any number of groups of LEDs can be provided, and tuning can be done by successively adding the next group to an optimal blend of previous groups, or by interpolating between multiple vertex locations associated with different mixtures of light from the different groups.
In some embodiments described above, an assumption is made that the change in color is linearly related to the change in relative currents between groups of LEDs when total current to all groups is held constant. This assumption works well for small regions in color space, particularly if the LEDs are chosen to have equal flux densities. In this case, an approach to tuning with two groups can include defining at least two reference points in color space, corresponding to at least two different distributions of a fixed total current between the groups of LEDs in a lamp, where the reference points are chosen such that the target color is intermediate between them, then applying linear interpolation to tune the current distribution such that the resulting light closely approximates the target color. Where more than two groups of LEDs are provided, at least three reference points in color space can be chosen such that the target color lies within a polygon (e.g., a triangle) defined by the reference points, and triangular interpolation and/or other interpolation techniques can be used to tune the current distribution such that the resulting light closely approximates the target color.
More generally, the change in color need not be linearly related to change in relative currents between the LED groups. Blending of light from independently-addressable LED groups having different colors or color temperatures can be used to tune a lamp regardless of whether the assumption of a linear relationship holds. In some cases where the assumption of linearity does not hold, the actual nonlinear response can be modeled for a family of lamps. Alternatively, a tuning algorithm can proceed by a “search” strategy that tests different divisions (or distributions) of currents among the LED groups and adjusts the current division iteratively based on color measurements. One search strategy can include shifting the current division by a fixed step size (e.g., 50 mA) between color measurements. Another search strategy can be based on a half-interval search technique, similar to a binary search. Starting from an assumption that the extremes of the current distribution bracket the target color temperature, the color temperature with an equal distribution of current can be measured. The next measurement can be taken with a current distribution halfway between equal and the extreme that should pull the result closer to the desired temperature, and this can be repeated until the desired color temperature is reached. A particular search strategy is not critical to the present invention.
In order to facilitate tuning, the total current applied to all groups is advantageously held constant during tuning; tuning is achieved by varying the distribution of the fixed total current to different groups (or, equivalently, the fraction of total current applied to each group).
The tuning processes described herein are straightforward and predictable, allowing for automated implementation, e.g., in a manufacturing environment. Examples of apparatus capable of implementing the tuning processes described herein will now be described.
Adjustment fixture 1802 can incorporate mounting features for holding a lamp 1812 in place during tuning Adjustment fixture 1802 also provides for delivery of light from lamp 1812 into optical fiber 1804 (e.g., a conventional optical fiber with a diameter of 100 microns). For example, adjustment fixture 1802 can include retention elements that hold optical fiber 1804 in position relative to lamp 1812 so that light from lamp 1812 falls onto the end of optical fiber 1804. In some embodiments, adjustment fixture 1802 can provide lenses or other optical elements, e.g., to focus the light from lamp 1812, thereby increasing the light incident on the end of optical fiber 1804.
Spectrometer 1806 can be of conventional design, such as the commercially available Ocean Optic USB4000 spectrometer. Any device capable of measuring light color and communicating its measurements to a computer can be used.
Programmable potentiometer 1810, which can also be of conventional design, can be connected to current input points of lamp 1812. Potentiometer 1810 can include variable resistors and the value of each resistor can be programmed, e.g., in response to a control signal. Potentiometer 1810 is advantageously arranged to apply resistances to divide an input current ITOT provided by current source 1818 into a current distribution for each group of LEDs in lamp 1812. For example, in the case where lamp 1812 includes cool white and warm white LEDs, IC can be delivered to the cool white LEDs while IW is delivered to the warm white LEDs in lamp 1812. For example, as shown in
Control system 1808 can be implemented using, e.g., using a computer system of conventional design, including a central processor (CPU), memory (e.g., RAM), display device, user input devices (keyboard, mouse, etc.), magnetic storage media (e.g., a hard or fixed disk drive), removable storage media (e.g., optical disc, flash-based memory cards), and the like. (In the interest of simplicity, these conventional components are not illustrated.) In one embodiment, control system 1808 is based on a Linux platform; however, a particular platform is not required. Control system 1808 can implement a single-color adjustment algorithm 1822, e.g., using program code that can be stored in memory and executed by the CPU. As described below, algorithm 1822 can implement aspects of process 600.
Control system 1808 can also implement a spectrometer driver 1824 that can receive color data from spectrometer 1806. In various embodiments, spectrometer driver 1824 can include a physical interface (e.g., Universal Serial Bus (USB) or the like) compatible with spectrometer 1806 and associated control software (executable by, e.g., a CPU or other processor of control system 1808) that can be used to direct the spectrometer to take readings and to provide data. In some embodiments, spectrometer driver 1824 in some embodiments can also provide code related to interpreting the data, e.g., converting measurements received from spectrometer 1806 into CIE color-space coordinates or other desired format.
Control system 1808 can also implement a potentiometer driver 1826 that can control operation of programmable potentiometer 1810. In various embodiments, potentiometer driver 1826 can include a physical interface (e.g., Universal Serial Bus (USB), I2C or the like) compatible with potentiometer 1810 and associated control software (executable, e.g., by a CPU or other processor of control system 1808) that can be used to instruct the potentiometer to set its variable resistances to specified values. The values can be specified by single-color adjustment algorithm 1822.
User interface 1828 can include standard interface components, such as a keyboard, mouse, track ball, track pad, touch pad, display screen, printer, etc., along with associated software executed by the CPU of control system 1808 to control and communicate with the interface components. Via user interface 1828, a user can communicate with single-color adjustment algorithm 1822 to control operation thereof. For example, the user can control starting and stopping of a tuning process and view data associated with tuning processes (e.g., plots similar to those of
Operation of apparatus 1800 can proceed as follows. First, an LED-based lamp 1812 (e.g., corresponding to lamp 100 of
Alternatively, in some embodiments, the lamp itself may include programmable potentiometers. For example,
Apparatus 1900 also includes a robotic arm 1930 that is operable by robotic driver 1932 to pick up a lamp (e.g., lamp 1912) from a location holding lamps to be tuned and place lamp 1912 into adjustment fixture 1802. Robotic arm 1930 is further operable by robotic driver 1932 to remove lamp 1912 from adjustment fixture 1802 after tuning and place lamp 1912 into a location designated for holding tuned lamps. Robotic driver 1932 can be controlled by a suitable robotic-control subsystem 1934, which can be implemented using hardware and/or software incorporated into control system 1908. Conventional techniques for robotic control systems can be used to implement robotic arm 1930, driver 1932 and control subsystem 1934. In some embodiments, adjustment fixture 1802 may include movable members that extend to hold lamp 1912 in place and retract to release lamp 1912. Such members can also be operated under control of robotic driver 1932, allowing full automation of the process of inserting lamps into the adjustment fixture for tuning and removing them when tuning is complete.
Apparatus 1900 allows for a fully automated tuning procedure, in which a lamp 1912 is inserted into adjustment fixture 1802 and connected to adjustment interface 1910. Robotic arm 1930 can be used to remove human intervention from the process of inserting and removing lamps from the adjustment fixture. Control system 1908, which can include components similar to those of control system 1808 of
At block 2006, control system 1808 operates apparatus 1900 to determine a current distribution that produces the desired color. For example, single-color adjustment algorithm 1822, which can implement any of the tuning processes described above, can be executed to determine a distribution of a total current among the groups of LEDs in lamp 1912 that produces the desired color. At block 2008, operating resistances for potentiometer 1914 that produce the desired current distribution are determined. For example, in one embodiment with two groups of LEDs, the principle that IW/IC=RC/RW can be used together with the operating currents IW0 and IC0 (determined at block 2006) to select appropriate resistances. This computation can be incorporated into single-color adjustment algorithm 1822. At block 2010, potentiometer 1914 is programmed with the operating resistances determined at block 2008; for instance, single-color adjustment algorithm 1822 can communicate the operating resistances to potentiometer driver 1826, which communicates the resistances to potentiometer 1914 via adjustment interface 1910.
At block 2012, the operating currents can be tested by measuring the operating color (x0, y0) while lamp 1912 remains in adjustment fixture 1802. In some embodiments, at block 2014, the color can be fine-tuned with a further adjustment, e.g., in response to the measurement at block 2012 and a least-squares fit to a blackbody curve.
At block 2016, after the final tuning is completed, lamp 1912 can be removed from adjustment fixture 1802. Potentiometer 1914 advantageously remains programmed with the operating resistances determined in process 2000 so that lamp 1912 will produce light of the tuned color whenever operating power is supplied.
After block 2016, process 2000 can end. In some embodiments, additional lamps can be tuned to the same color temperature by repeating process 2000 (starting from block 2004) for each lamp.
It will be appreciated that the process 2000 described herein is illustrative and that variations and modifications are possible. Steps described as sequential may be executed in parallel, order of steps may be varied, and steps may be modified, combined, added or omitted.
A similar process can be used with apparatus 1800 of
It should be noted that in ordinary use (after process 2000), lamp 1912 does not require any feedback mechanism to preserve the color tuning Potentiometer 1914 can remain in its programmed state for the life of the lamp, delivering the desired currents to keep the color tuned. The color will not shift as long as the LEDs within lamp 1912 remain color-stable throughout their lifetime. White LEDs capable of lifetime color stability to within acceptable tolerances are known and can be used in lamp 1912 or other lamps described here. Thus, there is no need for an active feedback process during ordinary use of the lamp and no need for a color sensor that is stable over the lifetime of the lamp. Accordingly, an external active feedback loop, e.g., as shown in
In some embodiments, lamp 1912 can include control circuitry to maintain a desired distribution of an input current to the different groups of LEDs. For example, programmable potentiometers can be used as described above. Once the current is tuned, the programmable potentiometers can store the resistance values corresponding to the desired color. In other embodiments, the lamp can include memory circuits (e.g., programmable read-only memory, flash memory or the like) that can store information indicating the desired distribution of current. Thus, for example, a fixture in which the lamp is installed can include a current controller capable of reading the stored information and providing input currents to each group of LEDs based on the desired distribution. Other techniques can also be used to store or retain the tuning information (e.g., the desired current distribution) within a lamp. In some embodiments, the lamp may be capable of operating at a user-selectable one of a number of different target colors (or color temperatures), e.g., by use of an external control switch to select a color or the like. The tuning process can be modified to determine a distribution of input current to produce each target color, and the lamp can store information indicating the distribution associated with each color; in operation, the lamp can retrieve the desired distribution based on the setting of the control switch.
Further, since ordinary use of lamp 1912 does not require a feedback loop, the various components of the feedback loop used for tuning can be external to lamp 1912 and removed after tuning, as is the case for apparatus 1900 of
While the invention has been described with respect to specific embodiments, one skilled in the art will recognize that numerous modifications are possible. For example, the invention is not limited to a particular lamp geometry or form factor or as to the number and type of LEDs. The particular current values and tuning constant values mentioned herein are also illustrative, and other values may be substituted. The number of groups of LEDs, number of LEDs in any group, and/or the color of a group can be varied. In general, a tunable lamp will include at least two groups of LEDs, with each group occupying a non-overlapping region in color space. The size of the region will depend in part on the manufacturing processes and tolerances used to produce the different groups of LEDs; where a group includes multiple LEDs, those LEDs can be randomly scattered within the associated color-space region. The regions allowed for different groups are advantageously chosen such that the desired (tuned) color is intermediate between the regions occupied by the different LED groups.
Thus, although the invention has been described with respect to specific embodiments, it will be appreciated that the invention is intended to cover all modifications and equivalents within the scope of the following claims.
This application is a continuation of U.S. application Ser. No. 13/106,808, filed May 12, 2011, entitled “Tuning of Emitter with Multiple LEDs to a Single Color Bin.” The disclosure is also related to commonly-assigned U.S. application Ser. No. 13/106,810, filed on May 12, 2011 (now U.S. Pat. No. 8,513,900, issued on Aug. 20, 2013). The disclosures of both applications are incorporated by reference herein.
Number | Date | Country | |
---|---|---|---|
Parent | 13106808 | May 2011 | US |
Child | 14091914 | US |