GB2467008A

GB2467008A - LED simulation control

Info

Publication number: GB2467008A
Application number: GB0900901A
Authority: GB
Inventors: David Gavin Scott-Maxwell; Richard Peter James Barton
Original assignee: Individual
Current assignee: Individual
Priority date: 2009-01-20
Filing date: 2009-01-20
Publication date: 2010-07-21
Also published as: GB0900901D0

Abstract

An apparatus for simulating a scene uses a number of light engines 601, each engine 601 having three or more LEDs (104, fig 1). Each LED (104, fig 1) in a light engine 601 is individually controlled by a microprocessor 101. The output from the light engines 601 varies dynamically over a range of colours. Information relating to the colours used may be determined from measurements of light output from a real object or scene that the apparatus is simulating. The information may be stored in a look-up table accessed by a pseudo-random number generator.

Description

DESCRIPTION

Technical Field

This invention relates to solid-state lamps, and in particular to the use of LED lamps and light engines to simulate a dynamically changing lighting environment, where the time scales over which the light changes can vary from the response time of the human eye at the shortest to monthly or annual variations.

Background Art

This invention describes an apparatus for simulating by means of light engines a natural scene or object in which variations of both colour and intensity are present.

Examples of the application of the invention include the simulation of a real fire as is used in electric fires, the simulation of light effects caused by the natural movement of waves and sunlight in an aquarium, and the simulation of the effects of time-varying shadows and sunlight on plants.

The nature of the problem and the invention will be illustrated by considering one example of its application by describing a fire simulation apparatus. Flame and fuel simulation may typically be used in an electric fire, or in a flame effect fire that is intended primarily not for heating but simply for providing a focal point to a room.

Simulation gives the aesthetic appeal that a natural fire provides while providing a cheaper, cleaner and more convenient alternative. In addition many modern gas fires are intended more as a focal point or decorative addition to a room than as effective sources of heating, and can be considered primarily as fire simulation equipment.

There is a long history describing the development of electric fires simulating real fires. Patent publications GB104533 and GB108097 from 1917 describe the use of spinning wheels above coloured light bulbs to represent flickering flames and fuel. A more recent patent describing the state of the art is GB1088577, filed in 1966, and describing electric fires incorporating two broad categories of means of simulation of fuel and flames.

The first of the two categories uses a simulated coal or wood bed, typically made of moulded fibreglass or plastic, beneath which a coloured light bulb is placed. A perforated or finned wheel or plate positioned above the bulb and below the fuel bed is free to spin in the convective air currents produced by the hot bulb, thereby giving a flickering effect that simulated real flames and burning fuel.

In the second of the two categories a number of ribbons are suspended inside the appliance, and are moved by air currents or by mechanically means. The ribbons are S illuminated by one or two light sources. The movement of the ribbons can be seen when viewing the fire either directly through a semi-transparent screen or by back projecting an image of the ribbons onto a semi-transparent screen in the fire, hence providing the appearance of flickering natural flames. In many cases a fire will contain simulation methods from both categories.

Although both simulation methods are widely used, both are unsatisfactory in a number of respects. For the method using a spinning wheel the problems are as follows.

* The flickering effect is not random as it is in a natural fire, but repeats on a time scale determined by the rotation of the spinning wheel, even if the holes or blades of the wheel are not placed symmetrically around the wheel.

* The colour is determined by the colour of the light bulb, and does not fluctuate as it does in a natural fire * The energy needed to supply the bulb is wasted. Even though in a heating appliance this is not very significant, in cases where the appliance is intended solely for decorative effect the bulb or bulbs can be a significant source of energy wastage.

* In cases where an attempt is made to reduce this energy wastage by using light-emitting diodes (LEDs) as light sources in place of bulbs as in patent publication W02008062062 the convective air currents may not be sufficiently strong to spin the wheel at the required rate. The simulation apparatus then requires an addition means of producing rotation or other movement.

The use of ribbons illuminated by light sources to simulate flames also has a number of deficiencies.

* Because the simulation involves viewing the reflection of light from moving ribbons through a semi-transparent screen the image of the flames is not very bright, making it difficult to observe the effect in a well-lit room.

* The light reflected from the ribbons does not change with time in the way in which simulates the appearance of a real flame. Even if a number of light sources such as filament bulbs or LEDs is used, the simulation of the flame does not have the required random nature.

In both categories of simulation method the same key problems are observed. The first is that simulation does not capture the random variation in time of the intensity and pattern of a real fire and the second is that the colours in the simulation do not change with time in a realistic way.

The invention described in this application provides a way of simulating a real fire in which both the random variations of both colour and intensity with time of a real fire can be simulated using light engines preferably incorporating LEDs. In a second application the invention is used to represent the changing patterns of light in an aquarium and in a third implementation the invention is used to represent the changing patterns of light in an internal garden. These applications are not the only ones that can be envisaged and should not be taken to limit the scope of the invention.

Previous patent publications have described the use of LEDs in the simulation of real fires, but in almost all of these publications the LEDs have been used as direct replacements for incandescent and filament lamps. The claimed advantages are primarily the much improved reliability of LEDs, so that lamp replacement is not required as frequently as it is for filament bulbs, and the lower energy dissipation in LED lamps. For example, in W02008062062 the figure illustrating the invention is almost identical with the figure appearing in the earlier patent publication GB 2230335 except that the filament lamps are replaced by LED lamps. In W00235 153 it is expressly stated in the description that the light source may alternatively comprise conventional small size electric bulbs. Neither of these publications discloses prior art relevant to the invention described in the present application.

Other patent applications describe the use of LEDs to simulate a flame effect by means of a display approach and not by means of a lighting approach as described in the present application. Patent publication GB2264555 entitled "Flame effect display" uses typically more than 70 LEDs to define the shape of flames in the fire, the LEDs being of different colours to simulate the colours in a real fire. The spatial variations of colour within the fire are determined directly by the way in which the LEDs are permanently mounted within the simulated fire, and a large number of LEDs is therefore required to give an accurate simulation. Because each LED has to be controlled separately to simulate the variations in intensity at different parts of the flame at different times each LED the cost of manufacturing such displays in very high. Further the fixed colour of each LED has to be carefully selected beforehand, leading to additional cost. Patent publication GB2322188 describes a similar invention in which a large number of individual LEDs are mounted in the fuel bed and switched on and off to create a simulation of the fuel. This invention suffers from similar disadvantages to that described in GB2264555.

The prior art can therefore be divided into two main categories. Within the first category the variations of colour and intensity at different regions of the simulated fire are produced by means of a fixed light source together with mechanical means for producing rotation or other movement that transmits or reflects the light. Within the second much smaller category the simulations use a display approach in which light sources are used to directly define the shape of flames or fuel. These light sources, usually LEDs, are then switched on and off to simulate the colour and intensity variations of a real fire.

In contrast the present invention defines a new category in which the variations in colour and intensity of a scene are simulated by means of a limited number of light engines and not by mechanical means or by a display approach. A small number of light engines, each containing typically three LEDs, is sufficient to produce a realistic simulation of time varying colour changes such as those seen in a fire. The invention produces the simulation with plurality of LEDs, typically ten, that do not have to be carefully selected for colour, and so of low cost.

Disclosure

The invention will be described with reference to the following figures.

Figure la) shows a light engine together with the current supplies and the control microprocessor.

Figure ib) shows an alternative arrangement for driving a light engine.

Figure 2 shows the representation of the light output from the light engine comprising three LEDs in terms of the well-known CIE diagram.

Figure 3 shows the representation of the ensemble of the points corresponding to the light emitted from a real fire by way of example at different times.

Figure 4 shows typical currents through three LEDs in the light engine as a function of time before any smoothing is applied Figure 5 shows typical currents through one of the LEDs as a function of time both unsmoothed and with the inclusion of different forms of smoothing.

Figure 6 shows light engines installed in an electric fire so as to simulate a real fire.

Figure 7 shows light engines installed above an aquarium so as to simulate an underwater environment.

In all implementations of the present invention illumination is by means of one or a plurality of LED light engines. For the purposes of this invention a light engine is taken to be a number of LEDs of different colours mounted together on a metal-cored printed circuit board or other means of removing heat and allowing for electrical contact, as is well known in the art. Each of the LEDs is controlled separately by a current supply such as is available commercially and each current supply is controlled in turn by a microprocessor as shown in figures la) and ib). Different implementations of the control arrangement are possible as will be obvious to those skilled in the art.

In the implementation shown in figure la) the microprocessor or other microcontroller 101 has three separate outputs marked R, G and B in figure la) corresponding in this example to red, green and blue LEDs, although alternative colours can also be used.

Each output provides a control signal through a data link 103 which is connected to a constant current source 102. The magnitude of the current provided by each constant current source is determined by the control signal from the microcontroller, and is supplied to an LED 104.

An alternative implementation of the LED controller is shown in figure ib) where the microcontroller has a serial data link 105 connected between the output labelled RGB X in figure ib) and a second microcontroller 106 that provides three separate constant current outputs. Each constant current output is connected to one of the three LEDs 104.

These two implementations are described by way of example only, and it will be obvious to those skilled in the art that a variety of other implementations of the LED control circuits is possible. The appropriate choice will be determined by considerations of performance and cost.

Light emitted from the light engine can conveniently be represented on the well-known CIE diagram shown in figure 2, where the light from each LED in the light engine is represented by two co-ordinates known as the X and Y coordinates and which can be plotted as a point on the CIE diagram. The X and Y coordinates of light corresponding to single wavelengths defines an area 201 on the CIE diagram. Each LED in the light engine corresponds to a different point as illustrated in figure 2 for the case, taken by way of example, of three LEDs. A triangular area 202 on the CIE diagram can be defined by joining the points corresponding to the chromaticity coordinates of the three LEDs. By varying the current through each LED in the light engine light corresponding to any X and Y coordinate within this area 202 can be produced.

In many applications it is important to ensure by means of lenses or diffusers in front of the light engine that light from all the LEDs in the light engine is mixed so to give a uniform colour. One of the advantages of this invention is that for the simulation of a fire this is not necessary, thereby reducing cost. For most applications using three LEDs the colours would be chosen to correspond to red, blue and green, although other choices are possible and have advantages in other implementations of this invention.

Light emission from a real object such as a fire, taken by way of specific example, varies not only from point to point in the object but also with time. Both the intensity and the colour of the light emitted from a small region of the fire change in a pseudo-random way with time. The appearance of the fire changes on time scales that vary from the response time of the human eye to very long times of several hours, although for the purposes of simulation of a fire prior art has typically considered changes on time scales of between 0.1 second and 1 second fixed by the motion of the mechanical simulation mechanisms. In the present invention the time scales of the fluctuations can be varied so that the invention can simulate both the rapid flickering seen in the early stages of combustion and also the slower variations seen in a mature fire.

The colour of the light emitted from a fire can be conveniently represented by means of points on the CIE diagram. Light from a small region of the fire at any one time will have specific X and Y coordinates that correspond to a point on the CIE diagram shown in figure 3, but at a different time the light will in general correspond to a different point. Measurements made of the light emitted at a large number of different times will when plotted on the CIE diagram be restricted to an area of this diagram 301 corresponding to the range of colours emitted by the fire.

The representation on the CIE diagram of a series of measurements of a typical real fire is shown in figure 3 where they are restricted to an area or region of the diagram 301 localised in the red and orange parts of the spectrum. This area 301 will be referred to subsequently as the simulation area. For other applications the simulation area will in general fall in another region of the DE diagram. For light observed underwater, for example, the simulation area could be limited to the blue/green region of the CIE diagram.

For each point on the CIE diagram representing the colour of the fire the intensity of the light output must also be specified. This intensity will change from a measurement made at one time to a measurement made at a later time so that each point within the simulation area also has a specified intensity. In some cases the intensity may be correlated with the colour.

Measurements of both the colour and the intensity can be made by measuring the spectrum of the light emitted from a region of the fire. The intensity of the light as a function of wavelength, usually known as the emission spectrum, can be used to calculate the CIE coordinates following procedures well known to those skilled in the art. The total intensity can be obtained by integrating the spectrum over wavelength.

Information on the colour and the intensity is most conveniently represented by means of a look-up table, although other means are also possible. Each point within the simulation area is described by three numbers, two corresponding to the X and Y colour co-ordinates and the third corresponding to the intensity. In another equivalent implementation of the look-up table the three numbers correspond to the drive currents needed for three LEDs in the light engine. This second implementation is often to be preferred because non-linearities such as the variation of LED output with current can be taken into account.

The colour co-ordinates and the intensity can also be obtained very conveniently by using a video camera to capture an image of the fire at a given time. An analysis of representative pixels within the video image can give the colour coordinates and also provides a direct measure of the intensity variations of the fire. A preferred method of analysing the data is to use the data from the video image to give the look-up table directly. The video image gives the intensity of the red, green and blue components of the image and can be used to give the colour co-ordinates and the intensity of the light output captured by the pixel. If the three intensities in the video image are used directly a correction factor should in principle be applied to relate the colour co-ordinates of the red, green and blue components of the image to those of the LEDs in the light engine.

The time variation of the video image also provides a measure of the required sampling frequency needed to capture the random fluctuations although other ways of obtaining this information are obvious to those skilled in the art.

The use of the phrase "sampling frequency" in the previous paragraph does not imply that the changes in the colour and intensity occur at a well defined frequency. In practice the use of a fixed frequency is more straightforward in any implementation, but random sampling with time can also be used.

The analysis of the colour of the real fire in terms of the colour coordinates can be repeated at other parts of the fire to give a representation of the way in which the colour of the fire varies as a function of position. It has been found that three positions are sufficient to give a realistic simulation in the case of a real fire, as will be described. More positions could be used but with the penalty of extra cost as more light engines would be needed for the simulation. In other implementations and applications of this invention a better representation can be achieved with additional light engines.

Simulation of the fire is achieved by driving the LED light engines so that the colour coordinates and intensity corresponding to the light output from each light engine vary pseudo-randomly with time. This can be achieved by using a pseudo-random number generator within the microprocessor. Preferably at any instant the generated random number is used to select one entry from the look-up table to specify the colour and intensity of each light engine and so to define the currents that must be supplied to each of the LEDs in the light engine at that instant. This is repeated after a short time interval shown as T in figure 4 to give in general a different current. Figure 4 shows the variation of the currents through the red R, blue B and green G LEDs as a function of time. The currents change randomly with time in the case of a fire to give randomly fluctuating colour coordinates for the light output with colour coordinates limited to the region of the simulation area. The sampling rate of the random number generator (lIT) is chosen to correspond to the observed time variation of the colour of the real fire. In figure 4 sampling is carried out for convenience at fixed time intervals T, but random sampling is also possible.

In order to avoid abrupt changes of colour with time a smoothing function is incorporated in the microprocessor control to ensure that the colour of the simulation does not change abruptly with time. This can be done either by means of a simple interpolation within the microprocessor, digital filtering or by introducing a low-pass filter into the output of the microprocessor as is familiar to one skilled in the art.

Figure 5b) shows the result of smoothing the currents through each LED by means of a linear interpolation of the original currents shown in figure 5a), and figure 5c) shows the currents after smoothing with a digital filter.

Other methods of generating the output from the light engines can be envisaged and the implementation described above should not be taken to limit the scope of the invention. Any method that gives a light output that corresponds to a limited region of the CIE diagram, the simulation area, can be used.

An array of light engines operating as described above can be used in a number of applications all relying on the basic principle of using a random variation of the currents through the LEDs to represent the simulation area on the CIE diagram defined by means of observations on real scenes. Specific implementations are defined below.

In the first implementation the light engines are used to simulate a real fire to overcome the problems described in the section on the background to the invention.

Figure 6 shows an electric fire containing three light engines 601 under a diffuse screen 602. Two of the light engines are shown in the cut-away drawing, one mounted in the centre and one on the right of the fire. Other arrangements and number of light engines are possible. Imitation fuel 603 is placed on the diffuse screen. This imitation fuel can be artificial coal or artificial logs, decorative ceramic pebbles or similar material that only partially obscures the light from the light engines. Light from the light engines is either reflected from the imitation fuel or passes through the fuel to be reflected from the back of the fire which is ideally made of diffusely reflective material 604 such as brushed stainless steel or similar material.

The electric fire is generally also provided with a heating apparatus although such an apparatus does not form part of this invention.

In operation each of the light engines is driven by way of example using the drive circuits shown in figure la) or figure ib) and illuminates the imitation fuel in such a way that in general each element of the fuel is illuminated on different sides from different light engines. The colours will change as the pseudo random light emission from the light engines spans the colour space defined by the simulation area on the CIE diagram in a way that simulates the colour variations of a real fire. Each of the light engines can be controlled by a separate microprocessor and current source, although it is also possible to use sequential addressing from a single microprocessor.

In this implementation the sampling rate with which the light output is changed is typically 5Hz, although lower or higher rotes can be used. In addition the simulation area and hence the look-up table can change with time in order to represent the changing stages of a real fire, as can the sampling rate for the look-up table. In the early stages of the life of a fire the rate of change will be higher and the colour tend to be yellow, but in the later stages the rate of change will be lower and the colour tend more towards the orange and red.

In a second application the invention is applied to the illumination of an aquarium as illustrated by way of example in figure 7. A plurality of light engines 701 is used to illuminate fish 702 (in case the drawing is not self-explanatory), underwater corals 703 and other structures. In the simplified figure 7 only two light engines are shown.

A similar approach to that used in the simulation of a fire can be used to produce a look-up table for the light engines that will reproduce the desired underwater environment in terms of colour and intensity. In this application the simulation area would be expected to lie predominantly within the blue and green regions of the CIE diagram.

In an extension of this application additional LEDs operating in the visible and near ultraviolet (UV) in the range of wavelengths 380nm to 450nm can be added to the light engines to generate actinic effects in corals.

In a further extension of this application the invention can be used to simulate the diurnal variations of light in the underwater environment for sea-life centres and for fish breeding. In this application the output from the light engines would simulate not only the short term variations in light colour and intensity but also the slower variations of daylight and moonlight, and the even slower monthly or seasonal variations.

In a third application the invention is applied to the illumination of plants and other vegetation in an indoor or garden environment. Using measurements on real outdoor scenes to define the simulation area, the invention would simulate the dynamic effects produced by variations in sunlight resulting from the motion of the sun or clouds, and simulate the fluctuations produced by wind.

It can be seen that the applications and implementations described here are by way of example only, and that the invention is equally applicable to a range of other dynamic lighting environments.

Claims

CLAIMS1. A simulation lighting apparatus in which light engines are used for time-dependent illumination and where a. the light engines are driven by a microcontroller capable of providing independent currents to each of the lamps in the light engine and b. the light output from the light engines is controlled so as to represent or simulate the appearance of a real scene and c. the colour of the light emission from the light engines is constrained to vary over a restricted area of the CIE chromaticity colour space known as the simulation area.
2. An apparatus as described in claim 1 in which the lamps are LEDs of different colours.
3. An apparatus as described in claim 1 in which the simulation area is defined by means of measurements on a real object.
4. An apparatus as described in claim 1 in which the light output is varied pseudo-randomly with time in both colour and intensity over the simulation area.
5. An apparatus as described in claim 1 in which the currents through the lamps in the light engine are determined by means of a look-up table based on measurements on a real object.
6. An apparatus as described in claims 1 to 5 in which the look-up table is defined by means of a video image of the real object.
7. An apparatus as described in claims 1 to 6 in which a plurality of light engines is used to simulate a real fire.
8. An apparatus as described in claims 1 to 6 in which a plurality of light engines is used to illuminate an aquarium, reptilium or similar environment.
9. An apparatus as described in claims 1 to 6 in which a plurality of light engines is used to illuminate or simulate an underwater environment.
10. An apparatus as described in claims 1 to 6 in which a plurality of light engines is used to simulate or illuminate dynamic outdoor light effects in a group of plants.