Pseudo-exponential coding for light emitting devices and related methods

Abstract

Light emitting devices are disclosed, and more particularly, light emitting devices with pseudo-exponential coding and related methods are disclosed. Pseudo-exponential encoding or pseudo-exponential transformation refers to encoding and decoding techniques that include locating some data bits while introducing at least one data bit offset from exponential transformation. The deviation involves introducing at least one data bit to avoid zero repetition that may exist during decoding. Exemplary light emitting devices include light emitting diode (LED) packages and / or LED displays. Pseudo-exponential coding as described herein provides bit shifting and operations to increase dynamic range while reducing the complexity and size of hardware resources.

Description

Technical Field

[0001] This disclosure relates to light-emitting devices, and more specifically to light-emitting devices with pseudo-exponential coding and related methods. Background Technology

[0002] A light-emitting diode (LED) is a solid-state device that converts electrical energy into light and typically comprises one or more active layers (or active regions) of semiconductor material arranged between n-type and p-type layers with opposite doping. When a bias voltage is applied to the doped layers, holes and electrons are injected into one or more active layers, where they recombine to produce emitted light, such as visible or ultraviolet light.

[0003] LEDs have been widely adopted in various lighting environments, including for backlighting liquid crystal display (LCD) systems (e.g., as an alternative to cold cathode fluorescent lamps) and for direct-view LED displays. Applications utilizing LED arrays include vehicle headlights, road lighting, luminaires, and various indoor, outdoor, and special applications. Desired characteristics of LED devices include high luminous efficiency and long lifespan.

[0004] Large-scale multicolor direct-view LED displays (including full-color LED video screens) typically comprise a large number of individual LED panels, packages, and / or components, with image resolution determined by the distance between adjacent pixels. Direct-view LED displays typically include tri-color displays with arranged red, green, and blue (RGB) LEDs, and dual-color displays with arranged red and green (RG) LEDs. For many LED display systems, it is desirable to form an LED color group for each pixel, such as the primary colors red, green, and blue (RGB), which defines the vertices of a triangle (or polygon) on a chromaticity diagram. This polygon defines the so-called color gamut of the display device, the area of ​​which describes all possible colors that the display device can produce. The driver printed circuit boards used to control LED displays are typically densely packed with electronics, including capacitors for driving the pixels of the display, field-effect transistors (FETs), decoders, microcontrollers, etc. For higher resolution displays, the pixel pitch continues to decrease, and the density of such electronics increases accordingly, resulting in an increase in the number of pixels for a given panel area. This often adds greater complexity and cost to LED panels used in display applications.

[0005] The field is still seeking improved LED array devices with small pixel pitch while overcoming the limitations associated with conventional devices and manufacturing methods. Summary of the Invention

[0006] This disclosure relates to light-emitting devices, and more specifically to light-emitting devices with pseudo-exponential encoding and related methods. Pseudo-exponential encoding, or pseudo-exponential transformation, refers to encoding and decoding techniques that involve positioning certain data bits while introducing at least one data bit in a manner deviating from an exponential transformation. The deviation involves introducing at least one data bit to avoid zero-value repetition that might otherwise be present during decoding. Exemplary light-emitting devices include light-emitting diode (LED) packages and / or LED displays. Pseudo-exponential encoding as described herein provides bit shifting and operation for increased dynamic range while reducing the complexity and size of hardware resources.

[0007] In one aspect, a method for controlling the light output of a light-emitting diode (LED) device includes: receiving compressed data comprising a first set of data bits and a second set of data bits, wherein the first set of data bits is a mantissa and the second set of data bits is an exponent; transferring the mantissa to a first position in an internal register determined by the exponent; introducing a pre-position value into a second position in the internal register determined by the exponent; and driving at least one LED chip. In some embodiments, the first position is associated with a power of 2 when the exponent is not zero, and the first position corresponds to the case where the exponent is equal to 1 when the exponent is zero. In some embodiments, the pre-position value is a single bit with a value of 1. In some embodiments, if the exponent is not zero, the pre-position value is positioned in the next higher significant bit immediately adjacent to the mantissa. In some embodiments, transferring the mantissa to the internal register includes performing a first shift by shifting the mantissa into the internal register; and introducing the pre-position value includes performing a second shift, in which the pre-position value is shifted into the internal register. In some embodiments, performing the second shift includes shifting in data bits with a value of 0 when the exponent is zero, and shifting in data bits with a value of 1 for all other exponent values. The method may further include calculating a correction exponent by taking the inverse of the exponent. In some embodiments, the correction exponent is further corrected by subtracting a value of 1 when the exponent is zero. In some embodiments, an internal register is further shifted according to the correction exponent. The method may further include sending the contents of the internal register to circuitry for driving at least one LED chip. In some embodiments, the circuitry includes at least one pulse width modulation (PWM) processor and driver circuitry coupled to at least one LED chip. In some embodiments, the driver circuitry includes a plurality of current sources configured to provide different current levels based on the output PWM signal of at least one PWM processor. In some embodiments, at least one LED chip, driver circuitry, and at least one PWM processor are integrated together to form an LED package. In some embodiments, the first and second shifts are performed from left to right within the shift register. In some embodiments, the first and second shifts are performed from right to left within the shift register.

[0008] In another embodiment, a light-emitting device includes: a light-emitting diode (LED) chip; a serial interface configured to receive compressed data including a first set of data bits and a second set of data bits, wherein the first set of data bits is the mantissa and the second set of data bits is the exponent; and a decoder including an internal register configured to: transfer the mantissa to a first position in the internal register determined by the exponent; and introduce a pre-bit value to a second position in the internal register determined by the exponent. In some embodiments, the first position is associated with a power of 2 when the exponent is not zero, and the first position corresponds to the case where the exponent is equal to 1 when the exponent is zero. In some embodiments, if the exponent is not zero, the pre-bit value is positioned in the next higher significant bit immediately adjacent to the mantissa. In some embodiments, the decoder is configured to transfer the mantissa by performing a first shift by shifting the mantissa into a shift register, and to introduce the pre-bit value by performing a second shift by shifting the pre-bit value into the internal register. In some implementations, the decoder is configured to perform a second shift in such a way that when the exponent is zero, a bit value of 0 is shifted into the predetermined bit value; or for all other exponent values, a bit value of 1 is shifted into the predetermined bit value. In some implementations, the decoder is configured to calculate a correction exponent shifted into the internal register, wherein if the exponent is zero, the correction exponent is calculated by taking the one's complement of the exponent and subtracting 1, and wherein for all other exponent values, the correction exponent is calculated by taking only the one's complement of the exponent. In some implementations, the internal register is further shifted according to the correction exponent.

[0009] The light-emitting device may further include: a driving circuit coupled to the LED chip, the driving circuit being configured to drive the LED chip based on the contents of an internal register; and a pulse width modulation (PWM) processor coupled to the driving circuit, the PWM processor being configured to receive the contents of the internal register. In some embodiments, the PWM processor includes multiple current sources configured to provide different current levels to the LED chip. In some embodiments, the light-emitting device is an LED package that includes an LED chip and active electronic components integrated within the LED package, the active electronic components including a serial interface, a decoder, a driving circuit, and a PWM processor.

[0010] On the other hand, any of the foregoing aspects, individually or together, and / or the different individual aspects and features as described herein, may be combined for additional advantages. Unless otherwise stated herein, any of the various features and elements disclosed herein may be combined with one or more other disclosed features and elements.

[0011] Those skilled in the art, after reading the following detailed description of the preferred embodiments in conjunction with the accompanying drawings, should understand the scope of this disclosure and implement its additional aspects. Attached Figure Description

[0012] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate several aspects of this disclosure and, together with the specification, serve to explain the principles of this disclosure.

[0013] Figure 1 This is a block diagram illustrating a system-level control scheme for a lighting device employing cascaded communication for series-connected light-emitting diode (LED) packages, based on the principles of this disclosure.

[0014] Figure 2 yes Figure 1 A block diagram of an LED package that can realize pseudo-exponential transformation according to the principles of this disclosure.

[0015] Figure 3 yes Figure 2 A schematic diagram of a portion of an LED package, showing details of the current source for the LED chip used in the LED package.

[0016] Figure 4A This is a schematic diagram illustrating the process for implementing a pseudo-exponential transformation according to the principles of this disclosure.

[0017] Figure 4B This is a schematic diagram illustrating a general process for implementing pseudo-exponential transformations according to the principles of this disclosure.

[0018] Figure 5 It is a diagram illustrating the pseudo-exponential transformation of various combinations of data according to the principles of this disclosure.

[0019] Figure 6 This is a diagram illustrating three different types of exponential decoding and encoding, used to compare the principles of this disclosure. Detailed Implementation

[0020] The embodiments described below provide the necessary information to enable those skilled in the art to practice the embodiments and illustrate the best mode of practice. When reading the following description in conjunction with the accompanying drawings, those skilled in the art will understand the concepts of this disclosure and will recognize that these concepts do not specifically address any applications herein. It should be understood that these concepts and applications fall within the scope of this disclosure and the appended claims.

[0021] It should be understood that although the terms first, second, etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. For example, without departing from the scope of this disclosure, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element. As used herein, the term "and / or" includes any and all combinations of one or more of the related listed items.

[0022] It should be understood that when an element, such as a layer, region, or substrate, is referred to as being "on" or extending "on" another element, the element may be directly on or directly extending onto the other element, or intermediate elements may be present. Conversely, when an element is referred to as being "directly on" or "directly extending onto" another element, no intermediate elements are present. Similarly, it should be understood that when an element, such as a layer, region, or substrate, is referred to as being "on" or extending "on" another element, the element may be directly on or directly extending onto the other element, or intermediate elements may be present. Conversely, when an element is referred to as being "directly on" or "directly extending onto" another element, no intermediate elements are present. It should also be understood that when an element is referred to as being "connected" or "attached" to another element, the element may be directly connected or attached to the other element, or intermediate elements may be present. Conversely, when an element is referred to as being "directly connected" or "directly attached" to another element, no intermediate elements are present.

[0023] Relative terms such as “below” or “above” or “upper” or “lower” or “horizontal” or “vertical” may be used herein to describe the relationship between one element, layer, or region and another element, layer, or region as shown in the figures. It should be understood that these terms, and those mentioned above, are intended to include different orientations of the device other than those depicted in the figures.

[0024] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. As used herein, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well. It should be further understood that, when used herein, the terms “comprises,” “comprising,” “includes,” and / or “including” specify the presence of the stated feature, integral, step, operation, element, and / or component, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or combinations thereof.

[0025] Unless otherwise defined, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. It should be further understood that the terms used herein shall be interpreted as having the same meaning as they have in the context of this specification and the relevant field, and shall not be interpreted in an idealized or overly formal sense unless expressly defined herein.

[0026] Embodiments are described herein with reference to schematic diagrams of embodiments of this disclosure. Therefore, the actual dimensions of layers and elements may vary, and deviations from the illustrated shapes are expected, for example, due to manufacturing techniques and / or tolerances. For instance, areas illustrated or described as squares or rectangles may have circular or curved features, and areas shown as straight lines may have some irregularity. Therefore, the areas shown in the figures are schematic, and their shapes are not intended to show the precise shape of areas of the device, nor are they intended to limit the scope of this disclosure. Furthermore, for illustrative purposes, the dimensions of structures or areas may be enlarged relative to other structures or areas; therefore, these dimensions or areas are provided to illustrate the overall structure of the subject matter and are not necessarily drawn to scale. Common elements between the figures may be indicated herein by common element numbers and may not be described further thereafter.

[0027] This disclosure relates to light-emitting devices, and more specifically to light-emitting devices with pseudo-exponential encoding and related methods. Pseudo-exponential encoding, or pseudo-exponential transformation, refers to encoding and decoding techniques that involve positioning some data bits while introducing at least one data bit in a manner deviating from the exponential transformation. This deviation involves introducing at least one data bit to avoid zero-value repetition that may occur during decoding. Exemplary light-emitting devices include light-emitting diode (LED) packages and / or LED displays. The pseudo-exponential encoding described herein provides bit shifting and manipulation to increase dynamic range while reducing the complexity and size of hardware resources.

[0028] In cascaded digital communication, multiple electronic devices are arranged as repeaters to continuously receive serial communications. In the scenario of small-pitch video displays, multiple LED packages are arranged in series as LED pixels to receive cascaded communications. The input signal to each LED pixel is generated by another element (such as a main controller or a previous LED pixel), and the bit stream of the input signal originates from the clock domain of one or more preceding devices. Properly distributing the communication signals to thousands of LED pixels presents challenges. The LED packages need to be small in size to form pixels in high-resolution video displays, and these size constraints introduce further challenges.

[0029] As used herein, the terms “data stream” and “communication channel” are sometimes used interchangeably. However, “data stream” generally refers to a non-physical representation over time of data flowing through at least one set of communication channels and the internal wiring and storage registers within individual components, such as controllers and active electronic components. A data stream can also be referred to as digital communication between two components, such as a controller component transmitting digital communication and a receiver component receiving digital communication. A “communication channel” generally refers to the physical medium through which the data stream is transmitted. For example, a communication channel can include wires with associated electronic components, optical fibers, or even air, as in the case of radio, light, or sound waves. A given physical channel can also be divided in time or frequency to allow multiple “communication channels” to be formed within a medium, such as varying to different frequency bands. In some aspects, a communication channel can be implemented as a serial digital communication channel. Some aspects involve a binary communication channel, which is a single wire referenced to a common conductor (such as ground), which typically holds only one value at a time, either a high voltage or a low voltage (e.g., the digital “0” or “1”) and is controlled by the output registers of the preceding device. This paper also envisions a two-wire differential signal transmission method, but the preferred embodiment shown here uses a single-wire approach, mainly because adding more traces with fine-pitch displays would increase complexity.

[0030] In some aspects, this disclosure relates to light-emitting devices including LEDs, LED packages, and associated LED displays, and more specifically, to active control of LEDs within an LED display. An LED display may include rows and columns of LEDs forming an array of LED pixels. A single LED pixel may include a group of LED chips of the same or multiple colors, wherein exemplary LED pixels include red LED chips, green LED chips, and blue LED chips. In some embodiments, an LED package includes a plurality of LED chips forming at least one LED pixel, and a plurality of such LED packages may be arranged to form an array of LED pixels for an LED display. Each LED package may include its own active electronics configured to receive control signals and actively maintain the operating state of the LED chips for the LED device, such as brightness or grayscale level or color selection signals, while other LED devices are being addressed. In some embodiments, the active electronics may include active circuitry including one or more of driver devices, signal conditioning or conversion devices, memory devices, decoder devices, electrostatic discharge (ESD) protection devices, thermal management devices, detection devices, voltage and / or current sensing devices, command processing devices, and other circuitry. The active electronics also include circuitry facilitating communication with multiple uncorrelated clock domains, including a native clock domain from the controller and a local clock domain originating within the active electronics. In this regard, each LED pixel of the LED display can be configured to operate using active matrix addressing with mixed clock domain communication. The active electronics can be configured to receive one or more of analog control signals, encoded analog control signals, digital control signals, and encoded digital control signals. In this arrangement, strings of LED packages, each containing an active electronics, can be arranged for serial communication, wherein each active electronics receives data from a data stream and transmits data to the next active electronics in the LED package string.

[0031] In active matrix addressing, each LED pixel is configured to actively maintain its operating state or otherwise control its drive state, such as brightness or grayscale level or color selection, while addressing other LED pixels. This allows each LED pixel to maintain or otherwise independently control its drive state, and improves display quality and / or image acquisition using photographic equipment by reducing or eliminating beat frequency effects between the lower-frequency pulses output by the display light and other asynchronous devices (e.g., light sources, other pulse displays, or image capture devices). Therefore, each LED pixel can be configured to maintain its corresponding operating state using continuous drive signals (including pulse width modulation (PWM)) rather than the conventional method of maintaining its corresponding operating state by scanning time-division multiplexed signals between pixel groups, which often results in the addition of low-frequency components to the drive signals associated with passive matrix addressing. For this purpose, each LED pixel may include an active electrical chip or active electrical element, which may include a memory device and have the ability to change the drive conditions of the LED pixel based on the state stored in the memory of the active electrical element. In some implementations, the continuous drive signal is a constant analog drive current, and in other implementations where brightness levels can be controlled via pulse methods such as PWM, the continuous drive signal may refer to a PWM signal that is not interrupted by time-division multiplexing scans of other LED pixels within the array or subarray. In various implementations, the active electronic components include integrated circuit chips, application-specific integrated circuits (ASICs), microcontrollers, or field-programmable gate arrays (FPGAs). In some implementations, the active electronic components may be configured to be programmed or reprogrammed after manufacturing using various memory elements and logic circuits integrated within the active electronic components.

[0032] As used herein, the terms "active electronic chip," "active electronic component," or "active electronic part" include any chip or component capable of changing the driving conditions of an LED based on memory or other information stored within the chip or component. As used herein, the terms "active LED pixel" and "smart LED pixel" are used interchangeably and can both refer to a device comprising one or more LED devices or chips forming a pixel and the active electronic components or chips as described above. In some embodiments, each LED pixel may include a single LED package configured as an active LED package, as described above, comprising multiple LED chips and active electronic components. In this way, the number of individual electronic components required for an LED display can be reduced, such as the separate electronic components located on the back side of the LED panel of the LED display as described above. Furthermore, the overall power consumption required for the LED panel to operate can be reduced.

[0033] The performance of LED displays continues to improve. Previously, LED displays were better suited for static images such as LED signage rather than dynamic applications such as video displays because they lacked many of the performance metrics required for high-quality video image display. The various performance metrics required for LEDs used in high-quality video displays include resolution, contrast ratio, viewing angle, dynamic range, brightness, frame rate, and color gamut. Recent advancements in LED packaging, improvements in LED driver quality, and cost reductions have significantly improved resolution and other requirements. As resolution increases, the packaging density of LED packages on printed circuit boards (PCBs) also increases and becomes more complex. Driving LEDs within LED packages to achieve high dynamic range while maintaining high frame rates and refresh rates remains challenging. This is because current driving technologies require remote drivers located on opposite sides of the PCB containing the LED array. Packaging density limitations necessitate that drivers be shared via time-division multiplexing techniques such as raster scanning. Since the preferred driving technology for LEDs utilizes PWM to set brightness, higher frequencies are required to achieve high dynamic range. Therefore, considering parasitic resistance, capacitance, and inductance, the dynamic range is limited by the highest frequency pulse that can be transmitted. LED packages arranged as pixels for cascaded serial communication can have drivers placed near each LED package, thus eliminating the need for shared drivers, significantly reducing parasitic resistance, capacitance, and inductance, and improving dynamic range.

[0034] For LED displays, a dynamic range greater than 1000:1 is considered good, while a dynamic range of approximately 1,000,000:1 is desirable. Some other display technologies, such as LCDs, claim to achieve such high dynamic range through the use of local dimming. PWM is typically applied to LEDs via a constant current driver. With this, the brightness range can range from the minimum pulse width to constant illumination at any given current. Furthermore, the PWM period should be such that the flicker rate of the minimum pulse is imperceptible to the human eye, such as a rate greater than 30 Hz. Therefore, to achieve a dynamic range of 1,000,000:1, a dedicated PWM driver with a 30 MHz clock is required. Some display technologies use 60 MHz, especially if the PWM driver is multiplexed. However, at higher clock frequencies, parasitic power consumption increases significantly, making this approach to achieving high dynamic range less ideal.

[0035] The human eye can perceive a dynamic range of approximately 20 levels (1,000,000:1) under a single lighting condition and adapts to a wider range of different lighting conditions. To accommodate various lighting levels, a higher dynamic range is required. Additional requirements may include providing calibration and thermal compensation. A dynamic range of 16,000,000:1 is likely a reasonable target for next-generation displays. If the upper limit of the PWM clock frequency does not provide the target dynamic range, one approach could involve changing the current. However, the switching of PWM drives from one current to another makes it difficult to match the actual LED brightness as the current changes. Therefore, without extreme measures to calibrate and adjust the PWM signal as compensation, brightness discontinuities associated with current variations may exist. Furthermore, performance differences with temperature make this task even more challenging. Therefore, in most systems using PWM, even with the ability to change the current, a fixed current is chosen for the entire dynamic range of the display.

[0036] Encoding techniques, such as gamma coding or correction, have been implemented to improve bandwidth efficiency while simultaneously enhancing the matching of human brightness perception through data compression. These techniques aim to allocate more bandwidth to brightness levels that the human eye is more sensitive to (such as dark tones), while conserving bandwidth for brightness levels that are difficult for the human eye to distinguish (such as high-bit and multi-color brightness levels). Camera manufacturers often employ different encoding schemes to try and use camera equipment more efficiently, making trade-offs, especially in terms of data bandwidth (file size or the ability to record at a given speed / resolution). While raw (RAW) format is generally preferred for image quality, it is not always practical, and raw format is often unnecessary if good encoding is used. Another consideration for camera manufacturers is the processing power of the processor in any given camera system. Often, faster and more expensive processors may be needed to implement more complex encoding schemes.

[0037] Based on the principles of this disclosure, a pseudo-exponential encoding scheme is implemented that provides efficient bandwidth utilization while reducing the size, cost, and complexity of hardware resources. In this way, this disclosure achieves efficient utilization of high dynamic range while reducing complexity. For example, according to the principles of this disclosure, 8-bit encoded data is subjected to a pseudo-exponential transformation to provide 19-bit or 24-bit data. As used herein, pseudo-exponential encoding / decoding or pseudo-exponential transformation refers to encoding and decoding techniques that involve positioning and / or transforming some data bits while introducing at least one data bit in a manner deviating from the exponential transformation. For example, a first set of bits (i.e., the mantissa) can be shifted into an internal register, and then further shifted into the internal register according to a second set of bits (i.e., the exponent). The second set of bits can be shifted in in a manner that avoids copying "0" values. For example, after initially shifting the mantissa, if the exponent is zero, a single bit "0" is shifted into the mantissa; and for all other exponent values, "1" is shifted into the mantissa. The corrected exponent can then be calculated. Given that further shifting occurs after the aforementioned if / then conditional logic used to avoid duplicate "0" values, the second set of bits can be referred to as a pseudo-exponent. In this context, the term "pseudo-exponent" refers to an exponent that is not used in the conventional manner in traditional coded number systems (such as floating-point number systems). The terms "exponent relation" and "exponent representation" generally have different, well-defined meanings. As used herein, unless otherwise defined, the term "exponent" generally refers to a pseudo-exponent. As used herein, a base-2 exponent or exponent representation is preferred for binary number implementations.

[0038] Figure 1 This is a block diagram 10 illustrating a system-level control scheme for a lighting device employing cascaded communication for series-connected LED packages 12 according to the principles of this disclosure. The lighting device can be implemented as an LED display, and each LED package 12 can form an LED pixel of the display. For this application, the terms LED package and LED pixel are used interchangeably, although it should be understood that an LED package can consist of several LED pixels formed together in one component. Figure 1The exemplary LED string 14 arranged for serial communication is represented by a dashed box. While only a single LED string 14 is shown in detail, one or more other LED strings may also be coupled to the controller 16. The controller 16 may include one or more integrated circuits, such as ASICs, microcontrollers, programmable control elements, and FPGAs. In some embodiments, the controller 16 may be referred to as the master controller of the LED string 14. In other embodiments, the controller 16 may be a sub-controller to which another master controller (not shown) delegates a set of tasks when it belongs to a larger system. The data signal output (Dout) of the controller 16 may be transmitted serially along the LED string 14, and the return data signal input (Din) may be received by the controller 16. The signal may include a raw clock domain signal provided by the controller 16 or another master controller as described above. Figure 1 In this configuration, each LED package 12 or LED pixel is labeled with a number such as “Px 1,1”, where the first number indicates the row and the second number indicates the column. Each LED package 12 includes its own active electronic component 18, which is registered and housed therein, such that each LED package 12 includes logic circuitry that responds to received data signals.

[0039] according to Figure 1 The arrangement of the LED package 12 includes, in important aspects, transmitting high-bit-depth data to the integrated LED driver within each LED package 12 and efficiently converting the data to drive the LED chip within each LED package 12 according to the desired light output level, while improving dynamic range. In some aspects, the LED package 12 and associated active electronics 18 can transform the received data using compression and / or decompression techniques (such as gamma functions or corrections). However, when this transformed data is used as input to a PWM driver, some values ​​may fail to elicit the expected response from the LED chip due to zero-value repetition during exponential decoding. As disclosed herein, an LED package 12 employing pseudo-exponential transformation and related methods are disclosed to avoid zero-value repetition.

[0040] Figure 2 for Figure 1The block diagram below illustrates an LED package 12 capable of pseudo-exponential transformation based on the principles of this disclosure. According to the embodiments disclosed herein, the active electronic element 18 may include multiple ports represented by a power supply voltage (Vdd), ground (GND), and bidirectional communication ports or digital input / output ports (DIO1 and DIO2). By configuring the DIO1 and DIO2 ports as bidirectional ports, the active electronic element 18 can advantageously detect input signals from a communication channel, then assign one of the DIO1 and DIO2 ports as an input port and the other as an output port. This functionality can be achieved by input / output buffers and / or active switching networks within the active electronic element and electrically connected to the DIO1 and DIO2 ports. This provides flexibility for display layouts where multiple LED packages 12 are connected together for cascaded communication. For example, multiple LED packages 12 can be arranged in multiple rows, where data is cascaded between packages along each row and transmitted in a serpentine manner between rows, such as... Figure 1 As shown. In this arrangement, the bidirectional communication ports allow the LED packages 12 to be mounted in the same orientation and to receive and transmit digital communications from left to right or right to left depending on the row position. In addition to the four ports Vdd, GND, DIO1, and DIO2 on the left side of the block diagram, the active electronics 18 also includes four ports on the right side that connect to LEDs 20-1 to 20-3 of the LED package 12. In this respect, LEDs 20-1 to 20-3 are packaged together with the active electronics 18 in a common LED package 12 to form individual pixels of a large display. As used herein, LEDs 20-1 to 20-3 may also be referred to as LED chips.

[0041] The following describes some components of active electronic component 18; however, it should be understood that active electronic component 18 may include many other components, including memory components, signal conditioning components, thermal management components, electrostatic discharge components, clock components, and oscillators, etc. Figure 2 In this configuration, control logic circuit 22 is configured to receive input data, execute commands according to a command protocol, provide control signals for the operation of LEDs 20-1 to 20-3, report various voltage and / or temperature levels included in the output data, and transmit output data to the next adjacent LED package via DIO1 and DIO2 ports. Control logic circuit 22 can operate in the digital domain and may include input / output buffers electrically connected to DIO1 and DIO2 ports, which assign input and output configurations for bidirectional DIO1 and DIO2 ports.

[0042] In some embodiments, active electronic element 18 may be configured to provide both forward and reverse bias states to LEDs 20-1 to 20-3. In this regard, control logic circuitry 22 may include a reverse bias voltage control output signal configured, via appropriate active elements, to provide LEDs 20-1 to 20-3 with a voltage level close to Vdd or close to GND. Since the term "reverse bias" implies that a high level at the output of control logic circuitry 22 creates a reverse bias condition, the output signal may be directly connected to inverter 24, which is disposed in the driver 26 of active electronic element 18. Therefore, LEDs 20-1 to 20-3 may be forward or reverse biased depending on a specific operating state and / or command received by control logic circuitry 22. Inverter 24 (or inverter logic element) may have sufficient output characteristics to drive LEDs 20-1 to 20-3. Driver 26 may substantially be an analog interface to active electronic components 18 electrically connected to control logic circuitry 22. Driver 26 may include controllable current sources 28-1 to 28-3, which may also be configured as LED sink drivers. Pull-up resistors R1 to R3 may be used to provide a path to Vdd for each of LEDs 20-1 to 20-3, facilitating voltage measurement when configured for reverse bias. Each of current sources 28-1 to 28-3 may be electrically connected to digital output signals LEDs 1 to 3 of control logic circuitry 22. Output signals LEDs 1 to 3 may be provided along multiple wires connected to each of current sources 28-1 to 28-3 for current selection. Output signals LEDs 1 to 3 may be embodied as PWM outputs of control logic circuitry 22 for controlling the operation of LEDs 20-1 to 20-3, as described in more detail later. Figure 3 , Figure 2 Each current source 28-1 to 28-3 includes multiple current sources at different current levels to provide varying current PWM control to each LED chip 20-1 to 20-3. In this regard, each current source 28-1 to 28-3 may be referred to as the total current source for a particular one of the LED chips 20-1 to 20-3, such that each total current source consists of multiple individual current sources at different current levels. Driver 26 may also include a multiplexer 30 electrically connected to the analog-to-digital converter (ADC) and ADC selector of the control logic circuitry 22. Furthermore, driver 26 may include an on-chip temperature sensor configured via multiplexer 30. In some embodiments, the temperature sensor provides thermal compensation for the LEDs 20-1 to 20-3 through thermal compensation curves and / or thermal shutdown.

[0043] Active electronic component 18 also includes a serial interface 32, which is embodied as a module with circuitry configured to decode the input signal of the data stream and convert it into a bit stream in a local clock domain, which can be further processed by control logic circuitry 22. Thus, serial interface 32 can also be referred to as a digital communication receiving device. Digital communication can be received from a controller in the serial string (e.g., Figure 1 16) and / or other LED packages (e.g., Figure 1 12) Receive. Serial interface 32 can receive compressed data and retransmit the data, combined with modified data, in a manner compatible with the entire LED display system to a communication channel connected to other LED packages or other external components.

[0044] In some embodiments, the control logic circuit 22 may include circuitry in the form of one or more PWM processors 34-1 to 34-3 providing output PWM signals to LEDs 1 to 3. The PWM processors 34-1 to 34-3 may be individually configured for each LED chip 20-1 to 20-3, or they may be combined into a single PWM processor for all LED chips 20-1 to 20-3. In some embodiments, the one or more PWM processors 34-1 to 34-3 are configured to transform the input PWM signal and then shift the transformed PWM value to provide the LED chips 20-1 to 20-3 with a pulse width that compensates for the delayed on-time.

[0045] The control logic circuit 22 may also include circuitry in the form of one or more decoders 36-1 to 36-3, each decoder having an output that feeds to the PWM processors 34-1 to 34-3. Decoders 36-1 to 36-3 can be configured individually for each LED chip 20-1 to 20-3, or decoders 36-1 to 36-3 can be combined into a single decoder for all LED chips 20-1 to 20-3. As will be described in more detail later, decoders 36-1 to 36-3 are configured to decode the compressed data received by the serial interface 32 and perform a pseudo-exponential transformation according to the principles of this disclosure. In this way, duplicate "0" values ​​that would otherwise be sent to the PWM processors 34-1 to 34-3 are avoided.

[0046] Figure 3 yes Figure 2 A schematic diagram of a portion of the LED package 12 shows details of the total current source 28-1 for the LED chip 20-1. Although Figure 3 Discussed within the context of the total current source 28-1 used to provide a varying current PWM to the LED chip 20-1, it should be understood that the principles described also apply to... Figure 2 Each of the total current sources 28-1 to 28-3. In this way, total current sources 28-2 and 28-3 can also be configured to supply current to each of them in the same manner. Figure 2 LED chips 20-2 and 20-3 provide varying current PWM. This is illustrated through an example. Figure 3 The total current source 28-1 is shown as having six current sources labeled 1X, 2X, 4X, 8X, 16X, and 32X. However, the principle described can be applied to any number of multiple current sources for the LED chip 20-1.

[0047] As in Figure 3 As shown, each current source 1X, 2X, 4X, 8X, 16X, and 32X is connected to the same LED chip 20-1 to provide different PWM current levels. For example, current source 2X provides twice the current of current source 1X, current source 4X provides four times the current of current source 1X, and so on. In this way, each subsequent current source 1X, 2X, 4X, 8X, 16X, and 32X provides twice the current level of the previous one. As shown, each current source 1X, 2X, 4X, 8X, 16X, and 32X is connected to a different output terminal of the output PWM signal LED 1 (e.g., different wires or conductive paths). In this way, the output PWM signal LED 1 is assigned to individually control individual switches 38-1 to 38-6, such as metal-oxide-semiconductor field-effect transistors (MOSFETs), associated with each current source 1X, 2X, 4X, 8X, 16X, and 32X. Therefore, based on the expected brightness level of LED chip 20-1, the output PWM signal LED 1 can selectively turn on one or more of current sources 1X, 2X, 4X, 8X, 16X, and 32X during the PWM cycle, thereby providing an increased non-zero brightness level. In some embodiments, the current levels are set in a power of 2 step sequence to conveniently match the current with the binary number encoded in the parallel output of the PWM signal LED 1, thereby achieving a linear relationship between the current and the binary output value. However, other multi-current schemes are envisioned, such as setting at least two or even all current sources to the same current. For example, a particular configuration may have at least two identical current sources (e.g., 1X, 1X) while having other different current sources (e.g., 2X, 4X, 8X, ...). In other embodiments, to compensate for other nonlinearities or deviations within the system, the current levels may be intentionally adjusted to values ​​different from those corresponding to strict powers of 2.

[0048] like Figure 2 and Figure 3As shown, LED driver 26, current sources 28-1 to 28-3, PWM processors 34-1 to 34-3, and decoders 36-1 to 36-3 are advantageously integrated within the active electronics 18 of LED package 12. In this respect, LED package 12 can form LED pixels configured for active matrix addressing in an LED display. However, the principles disclosed herein are applicable to other implementations, such as light-emitting devices (e.g., systems or displays) where the circuitry is separate from each LED package or LED pixel. Regarding Figure 3 The principle described above for changing the current level within the PWM cycle can be combined with that in Figure 4 and... Figure 5 This is achieved through a pseudo-exponential transformation as described.

[0049] Figure 4A This is a schematic diagram illustrating a process 40 for implementing a pseudo-exponential transformation according to the principles of this disclosure. In the first step 42, the LED device receives compressed data from a data stream. Figure 2 In the application scenario of the LED package 12, compressed data can be received by the serial interface 32. While the described principle applies to various data sizes, as an exemplary implementation, the compressed data will be described in 8-bit data format. The 8-bit data may include a first set of data bits (i.e., the mantissa) and a second set of data bits (i.e., the exponent). For reference, the 8-bit data can be represented as "eeee_mmmm", where "m" is the mantissa and "e" is the exponent. As described herein, the compressed data is decompressed or decoded into an internal register. Figure 2 In the application scenario of LED package 12, decoders 36-1 to 36-3 include internal registers, which may also be referred to as shift registers. As used herein, internal registers or shift registers can refer to any memory element capable of holding data. By way of example, the 8-bit data received in the first step 42 can be decoded into a 24-bit register, which can typically be represented as "0000_0000_0000_0000_0000_0000".

[0050] In the second step 44, the mantissa is transferred to the internal register. By way of example, for a serial bit-shifting implementation, the mantissa can be shifted from left to right for the first time, in which case the 24-bit format of the 24-bit register can be represented as "mmmm_0000_0000_0000_0000_0000". Although this implementation discusses the application scenario of shifting from left to right, the principle described also applies to shifting from right to left.

[0051] In step 46 of the third step, a pre-position value is introduced into an internal register. In the serial bit shift example, a second shift is performed, shifting in at least one data bit (i.e., the pre-position value) based on the exponent. The 24-bit register can then be represented as "bmmm_m000_0000_0000_0000_0000", where "b" is the data bit shifted in during step 46. The shift setting of the "b" value is to prevent the exponent value from being repeatedly 0 in all cases where the mantissa is 0. In this way, if the exponent is zero, the value of "b" is "0". For all other non-zero exponent values, the value of "b" is "1".

[0052] In step 48, the mantissa and prepositioned values ​​are repositioned based on the exponent. In this way, the corrected exponent (hereinafter referred to as eeee') can be calculated. When the exponent value is zero (i.e., "eeee=0000"), the corrected exponent (eeee') can be calculated by taking the inverse of the exponent and subtracting "1". For all other non-zero exponent values, eeee' is calculated simply by taking the inverse of the exponent, without subtracting "1". This calculation can be expressed by the following formula, where the inverse operator is "~", which means inverting all bits.

[0053]

[0054] As indicated, if the exponent value is non-zero, the process can continue without performing a decrement operation. The eeee' value is not shifted into the 24-bit register in the same way as the mantissa. Instead, it is shifted from the left to zero according to the value of eeee (i.e., the 24-bit register is shifted right eeee' times).

[0055] In step 50, the contents or data bit value of the 24-bit register are sent to other circuits (e.g., Figure 2 The PWM processors 34-1 to 34-3 and current sources 28-1 to 28-3 drive one or more LED chips.

[0056] As mentioned above, Figure 4AProcess 40 avoids the problem of repeated zero values ​​in all cases where the mantissa is zero, thus achieving a higher dynamic range. Since "1" is shifted into all non-zero exponent values ​​in the third step 46 above, the decompression does not strictly follow standard exponential decompression. Therefore, process 40 can be called pseudo-exponential encoding / decoding and / or pseudo-exponential transformation. Table 1 below provides examples of pseudo-exponential transformations for 8-bit compressed data (eeee_mmmm) with different exponent values ​​after the completion of the fourth step 48. For example, the first row gives an example with an exponent value of zero, such that 0000_mmmm, after the fourth step 48, results in 24-bit register data of 0000_0000_0000_000m_mmm0_0000.

[0057] Table 1

[0058] As can be seen from Table 1, in the case where the compressed data contains a 4-bit exponent and a 4-bit mantissa, the rightmost 5 bits of the 24-bit shift register are never used. However, other implementations with more mantissa bits or more exponent bits can utilize the entire 24-bit register.

[0059] Although Figure 4A The description is based on a serial bit shift implementation, but the principle is generally applicable to implementations that perform bit shifting in any order (including parallel, serial, or other operations). In general scenarios, the above refers to... Figure 4A The aspects described can be based on the following Figure 4B This is achieved through the principle of [the principle]. In this way, Figure 4B This is a schematic diagram illustrating a general process 52 for implementing a pseudo-exponential transformation according to the principles of this disclosure. Figure 4B In the middle, the first step 54 is related to the previous text. Figure 4A The same as described in the first step 42. Figure 4B In this process, steps 56 and 58 can be performed in any order. Step 56 involves transferring the mantissa to an internal register, and the mantissa will eventually reside in a first position determined by the exponent. Step 58 involves introducing a pre-position value into an internal register, and the pre-position value may eventually reside in a second position, also determined by the exponent. In some embodiments, a first and a second position may be set simultaneously with the transfer of the mantissa and the pre-position value to the internal register. In other embodiments, the mantissa and the pre-position value may reach the first and second positions, respectively, after the initial transfer to the internal register. The following paragraphs describe how the first and second positions can be calculated.

[0060] exist Figure 4B In the process, the compressed data received at step 42 in the first step can be referred to as the compressed data received by... + The encoded value consists of bits, where It is the number of bits in the exponent. It refers to the number of bits in the mantissa. The decoded value is ultimately stored in a storage system that can hold... The bits are in the internal register (steps 56 in the second step and 58 in the third step). Depending on the number of bits, there may be unused bits in the internal register, or some least significant bits may be missing. Extra bits It is expressed by the following formula.

[0061]

[0062] When extra bits When the value is positive, this extra bit will be padded with zeros at the least significant bit of the internal register. When the value is negative, the indicated number of bits may be lost. Using an 8-bit example with a 24-bit register, the extra bits... The following calculations can be followed.

[0063]

[0064] In this general example, the data can be located according to the following formula, where the result is the decoded value finally stored in the result register, m is the mantissa value, and e is the exponent value.

[0065]

[0066] In the above formula, if the exponent e is zero, the position of the mantissa m depends on the number of extra bits. Otherwise, as shown in the first term, the mantissa m will be shifted again according to the exponent e, with the starting position of the shift satisfying the condition that the position when e=1 is the same as the position when e=0. As shown in the second term, using... The "1" is added to the next higher-order bit position, becoming the most significant bit of the mantissa m.

[0067] For example, in the absence of extra bits ( When =0), the above formula can be expressed as follows.

[0068]

[0069] Therefore, the top condition ( The expression represents the true mantissa-exponent relationship, and if e=0, the pseudo-exponent principle described in this paper is equivalent to it. The top condition will equal the bottom condition. The equation is almost satisfied when m reaches its maximum value (for all exponents e). For the example with 4 mantissas, The maximum value is ,as well as 16 makes the bottom condition almost identical to the top condition. It deviates from the true mantissa-exponent relationship when the mantissa is less than its maximum value, but yields better results. The pseudo-exponent principle described in this paper is monotonic, and numerical repetition only occurs when the result register's bit width is insufficient to hold the full range of values.

[0070] Figure 5 This is a diagram illustrating the pseudo-exponential transformation of various combinations of data according to the principles of this disclosure. The x-axis represents the input (i.e., compressed) data value, and the y-axis represents the internal decompressed value. The plot lines are a series of three-digit markers separated by colons, in the format (exponent bits + mantissa bits):exponent bits:d, where "exponent bits + mantissa bits" represents the total number of bits in the compressed format and can be referred to as the input bit width, "exponent bits" represents the number of bits of the exponent and can be referred to as the exponent width, and d represents the number of bits after decoding and decompression and can be referred to as the internal bit width. In this regard, the plot line 8:4:24 represents 8-bit compressed data with a 4-bit exponent, which is decoded into 24 bits in the internal register. In some embodiments, the number of bits after decoding and decompression can be greater than or less than the internal bit width supported by the shift register. When the bit width of the internal register is less than the bit width of the input data, the least significant bit is shifted out and lost, resulting in a lower dynamic range because it is limited by the bit width of the shift register. For comparison, a 24:0:24 line with a zero exponent width (i.e., no exponent) is provided to indicate that the encoding is off. As shown, while plot lines 8:4:24 and 12:4:24 may not cover the full range at lower levels, they achieve excellent pseudo-exponential transformations, improve dynamic range, and provide a visually pleasing display response for human observers. For the 8:4:24 case in Table 1, it can be seen that the full range of the internal shift register is not fully utilized (in the case of eeee=0000, the mantissa is not shifted to the rightmost end of the shift register, missing 5 bits). Plot line 16:4:24 covers the full range, although with several initial repetition values. Plot line 24:4:24, with the highest input width, shows a reduction in resolution in the first few thousand values. Therefore, there is a trade-off between resolution and range, with some preferred implementations using a 4-bit exponent width and a variable number of bits for the mantissa width. For some applications, a lower exponent width cannot adequately extend the dynamic range.

[0071] Figure 6This is a diagram illustrating three different types of exponential decoding and encoding to compare the principles of this disclosure. The x-axis represents the input data of the 8-bit example, and the y-axis represents the output. The plot line labeled "Exponential Relationship" refers to the standard exponential relationship, which would be a good alternative to gamma correction, and a similar implementation could exist in camera applications. However, this example requires a large amount of mathematical computation and therefore complex hardware requirements, which may not be suitable for all applications, such as light-emitting devices and / or LED packages as pixels in displays. The plot line labeled "Standard Exponent" refers to a true mantissa-exponent implementation that does not follow the principle of avoiding zero-value repetition in this disclosure. In this way, undesirable decompression is produced, such that when the mantissa is 0, each exponent value is 0, and the plot lines overlap. The resulting plot line corresponds to a poor dynamic range visual effect. The plot line labeled "Pseudo-Exponent" represents the above for... Figure 4B The principles of this disclosure are as described above. Therefore, the pseudo-exponential encoding of this disclosure provides superior performance while avoiding the use of complex hardware that requires extensive mathematical calculations.

[0072] Therefore, the above is aimed at Figures 1 to 6 The described pseudo-exponential transformation technique offers significant advantages over standard exponential decompression. Existing methods use gamma functions or power-law expressions, which do not suffer from the problem of repeated values, but require complex hardware to perform the calculations. The pseudo-exponential encoding and / or transformation described in this paper provides a simpler method and apparatus, employing simple bit operations such as shifting, inversion, and single decrementing. Therefore, this reduction in complexity allows for a smaller die size of the device, thereby lowering associated costs.

[0073] It is conceivable that any of the foregoing aspects and / or the various aspects and features described herein may be combined to obtain additional advantages. Unless otherwise stated herein, any of the various embodiments disclosed herein may be combined with one or more other disclosed embodiments.

[0074] Those skilled in the art will recognize improvements and modifications to the preferred embodiments of this disclosure. All such improvements and modifications are considered to be within the scope of the concepts disclosed herein and the appended claims.

Claims

1. A method for controlling the light output of a light-emitting diode (LED) device, the method comprising: Receive compressed data including a first set of data bits and a second set of data bits, wherein the first set of data bits is the mantissa and the second set of data bits is the exponent; The mantissa is transferred to the first position in the internal register determined by the exponent; The preposition value is introduced into the second position in the internal register determined by the index; and Drive at least one LED chip.

2. The method according to claim 1, wherein, When the exponent is not zero, the first position is related to the power of 2; and when the exponent is zero, the first position is equivalent to the case where the exponent is equal to 1.

3. The method according to claim 1, wherein, The predefined bit value is a single bit with a value of 1.

4. The method according to claim 1, wherein, If the exponent is not zero, the preposition value is positioned in the next higher significant bit immediately following the mantissa.

5. The method according to claim 1, wherein: Transferring the mantissa to the internal register includes performing a first shift by shifting the mantissa into the internal register; and Introducing the preposition value includes performing a second shift, in which the preposition value is shifted into the internal register.

6. The method according to claim 5, wherein, Performing the second shift includes shifting in data bits with a value of 0 when the exponent is zero, and shifting in data bits with a value of 1 for all other exponent values.

7. The method according to claim 6 further includes calculating a modified exponent by taking the inverse of the exponent.

8. The method according to claim 7, wherein, When the index is zero, the correction index is further corrected by subtracting the value 1.

9. The method according to claim 8, wherein, The internal register is further shifted according to the correction exponent.

10. The method according to claim 1, further comprising: The contents of the internal register are sent to the circuitry used to drive the at least one LED chip.

11. The method according to claim 10, wherein, The circuit includes at least one pulse width modulation (PWM) processor and a driving circuit connected to the at least one LED chip.

12. The method according to claim 11, wherein, The drive circuit includes multiple current sources configured to provide different current levels based on the output PWM signal of the at least one PWM processor.

13. The method according to claim 12, wherein, The at least one LED chip, the driving circuit, and the at least one PWM processor are integrated together to form an LED package.

14. The method according to claim 5, wherein, The first shift and the second shift are performed from left to right within the shift register.

15. The method according to claim 5, wherein, The first shift and the second shift are performed from right to left within the shift register.

16. A light-emitting device, comprising: Light-emitting diode (LED) chip; A serial interface is configured to receive compressed data comprising a first set of data bits and a second set of data bits, wherein the first set of data bits is the mantissa and the second set of data bits is the exponent; as well as The decoder, including internal registers, is configured to: The mantissa is transferred to the first position in the internal register determined by the exponent; and The preposition value is introduced into the second position in the internal register determined by the index.

17. The light-emitting device according to claim 16, wherein, When the exponent is not zero, the first position is related to the power of 2; and when the exponent is zero, the first position is equivalent to the case where the exponent is equal to 1.

18. The light-emitting device according to claim 16, wherein, If the exponent is not zero, the preposition value is positioned in the next higher significant bit immediately following the mantissa.

19. The light-emitting device according to claim 16, wherein, The decoder is configured to transmit the mantissa by performing a first shift by shifting the mantissa into a shift register, and to introduce the pre-position value by performing a second shift to shift the pre-position value into the internal register.

20. The light-emitting device according to claim 19, wherein, The decoder is configured to perform the second shift in such a way that when the exponent is zero, the pre-position value is shifted into bit value 0; or for all other exponent values, the pre-position value is shifted into bit value 1.

21. The light-emitting device according to claim 20, wherein, The decoder is configured to calculate a correction exponent shifted into the internal register, wherein if the exponent is zero, the correction exponent is calculated by taking the inverse of the exponent and subtracting the value 1, and wherein for all other exponent values, the correction exponent is calculated by taking only the inverse of the exponent.

22. The light-emitting device according to claim 21, wherein, The internal register is further shifted according to the correction exponent.

23. The light-emitting device according to claim 21, further comprising: A driving circuit is connected to the LED chip, and the driving circuit is configured to drive the LED chip based on the contents of the internal register. as well as A pulse width modulation (PWM) processor is connected to the drive circuit, and the PWM processor is configured to receive the contents of the internal register.

24. The light-emitting device according to claim 23, wherein, The PWM processor includes multiple current sources configured to provide different current levels to the LED chip.

25. The light-emitting device according to claim 23, wherein, The light-emitting device is an LED package, which includes the LED chip and active electronic components integrated within the LED package. The active electronic components include a serial interface, a decoder, the driving circuit, and the PWM processor.

Images