LED driver

Abstract

An LED light emitting diode driver (100), the LED driver comprising: an electrical driver output (182') to which the LED driver (100) is arranged to provide a current in dependence on a binary control signal; and a timer function (132') configured to update a timer counter variable at a timer frequency, the timer function (132') being configured to compare, for each update, the timer counter variable to a set comparison value, and the timer function (132') being configured to cause the binary control signal to switch between a first binary state and a second binary state in the event that the timer counter variable has reached the set comparison value, wherein a duration of a repeating control signal pattern of the binary control signal is longer than a time taken for the timer counter variable to complete a full cycle, and wherein the LED driver (100) is arranged to update the set comparison value at least twice during each individual control signal pattern.

Description

Technical Field

[0001] This invention relates to an LED (light-emitting diode) driver and a method for operating the LED driver. Background Technology

[0002] LED light sources are more energy-efficient than many other types of light sources, such as halogen or incandescent light sources. Therefore, sales of LED light sources have increased in recent years. At the same time, considerable resources have been invested in the further development of LED light sources themselves and in developing various methods to power them.

[0003] Typically, low DC voltages (such as 12V or 24V DC voltage) are used to drive LED light sources, which are provided by what is known as LED drivers. To control the luminous intensity of an LED light source, an LED driver typically controls the power delivered to the LED light source by either controlling the voltage while keeping the current constant or by controlling the current while keeping the voltage constant.

[0004] LED light sources can generally be dimmed across a relatively wide luminous intensity spectrum. However, it has proven difficult to provide an LED driver capable of dimming LED light sources across an extremely wide dimming range, such as up to complete darkness.

[0005] Specifically, dimming an LED light source across a wide dimming range is difficult without considering the individual dimming behavior of the LED light source to be dimmed. It is desirable to provide an LED driver that offers sufficient dimming across a wide dimming range for many different types of LED light sources.

[0006] It has been found that dimming conventional LED drivers across such wide dimming ranges can cause one or more of the following problems: flickering light; unpleasant noise; increased wear on electrical components; noticeable dimming steps and / or color shift at low light power; cable crosstalk; uneven or unpredictable illumination levels at different dimming settings and along the LED strip (“LED strip”); and high peak current requirements.

[0007] Some of these problems can be partially or completely overcome by selecting expensive or short-life components.

[0008] Therefore, it is desirable to provide a low-cost LED driver that has a long service life and avoids one or more of the problems mentioned above. Summary of the Invention

[0009] This invention relates to an LED (light-emitting diode) driver.

[0010] LED drivers may include a binary control signal source arranged to provide a binary control signal as a square wave having a control signal pattern that repeats over time to form the binary control signal, the control signal pattern having an instantaneous time average at each time point, the instantaneous time average being the time-averaged control signal pattern used at the time in question.

[0011] An LED driver may include a first channel electric driver output, which is arranged to provide drive current to drive the LED based on a binary control signal.

[0012] An LED driver may include a power transistor having a power transistor control terminal and a pair of different terminals, one of which is connected to the output of a first channel driver.

[0013] An LED driver may include a driving circuit that is connected to the control terminals of a power transistor.

[0014] The drive circuit may include a switch arranged to provide electrical coupling between a power transistor control terminal and a low or high voltage based on an instantaneous low or high value of a binary control signal.

[0015] The drive circuit is configured to receive or generate a level signal that is provided as an increasing or decreasing function of the instantaneous time average, or as an increasing or decreasing function of a control signal that is in turn related to the instantaneous time average.

[0016] The drive circuit may also include a variable resistor section arranged to provide a variable resistance along the electrical coupling, the variable resistance varying according to the level signal, thereby causing an increase in the instantaneous time average value, resulting in an increase in the instantaneous current flowing to or out of the power transistor control terminal.

[0017] In some implementations, the drive circuit includes a signal averaging section that is arranged to provide a level signal as or as a value corresponding to an instantaneous time average or the reciprocal of the instantaneous time average.

[0018] In some implementations, the signal averaging section includes a low-pass filter or digital-to-analog converter circuit arranged to provide the level signal as a smoothed and / or time-averaged version of the binary control signal, or to provide the level signal as a smoothed and / or time-averaged signal calculated based on the binary control signal, wherein the value of the level signal at a specific time is determined to correspond to the overall level of the binary control signal at the same specific time.

[0019] In some implementations, the level signal is a current or voltage corresponding to an instantaneous time average.

[0020] In some implementations, the power transistor is a FET (field-effect transistor).

[0021] In some implementations, the switch includes a switching transistor having a gate / base terminal, a source / emitter terminal, and a drain / collector terminal, the gate / base terminal of the switching transistor being connected to a binary control signal in the form of a voltage or current that varies according to the value of a binary control signal.

[0022] In some implementations, the switch includes two switching transistors, each with its gate / base terminal connected to a binary control signal in the form of a voltage or current that varies according to the value of a binary control signal. Together, the two switching transistors provide a push-pull circuit.

[0023] In some implementations, two switching transistors are arranged between two different voltage potentials.

[0024] In some implementations, the power transistor control terminal is coupled to a point between two switching transistors.

[0025] In some implementations, the switching transistor is a FET or a BJT (bipolar junction transistor).

[0026] In some implementations, the variable resistor section includes a current throttling section that is arranged to throttle the current that can be electrically coupled, and thus limit the rise and / or fall rate of the electrical signal supplied from the power transistor to the drive output when the switch electrically couples between low and high voltages.

[0027] In some implementations, the variable resistor portion includes a current-controlled transistor, which in turn includes a current-controlled gate / base terminal, a current-controlled source / emitter terminal, and a current-controlled drain / collector terminal.

[0028] In some implementations, the current flowing through the electrically coupled transistor flows across the current-controlled transistor via the current-controlled source / emitter terminals and the current-controlled drain / collector terminals.

[0029] In some implementations, the level signal is coupled to the current-controlled gate / base terminal, such that an increase in the instantaneous time average results in a higher instantaneous current across the current-controlled source / emitter terminal and the current-controlled drain / collector terminal.

[0030] In some implementations, the current-controlled transistor is a BJT.

[0031] In some implementations, the current-controlled transistor is connected in series with the first current-controlled resistor along an electrical coupling.

[0032] In some implementations, the variable resistor portion includes a first current control transistor that is configured such that current flowing through the electrical coupling between the first switching transistor and the control terminal of the power transistor passes through the first current control transistor.

[0033] In some implementations, the variable resistor portion includes a second current control transistor that is configured such that current flowing through the electrical coupling between the second switching transistor and the control terminal of the power transistor passes through the second current control transistor.

[0034] In some implementations, the binary control signal source is arranged to generate a binary control signal in response to the LED dimming value by varying both the pulse width and the number of pulses in the control signal mode.

[0035] In some implementations, the binary control signal source is arranged to generate a binary control signal in response to an LED dimming value by widening the pulses of the control signal pattern as the LED dimming value increases, and then increasing the number of pulses in the control signal pattern.

[0036] The present invention also relates to a method for driving an LED, the method comprising: providing power to a driver output of an LED driver connected to the LED, thereby causing a binary control source to provide a binary control signal and causing a power transistor to provide a drive current to the driver output.

[0037] In some implementations, the method further includes adjusting a binary control signal, thereby modifying the instantaneous time average.

[0038] In some implementations, the method further includes: adjusting the dimming state of the LED driver; and mapping the adjustment of the dimming state to a corresponding adjustment of the binary control signal.

[0039] The present invention also relates to a method for configuring an LED (light-emitting diode) driver for dimming an LED, the LED driver being arranged to provide a variable current to the LED, the time average of the variable current corresponding to a dimming value, the variable current being modulated using a transistor of the LED driver based on a binary control signal having a control signal pattern that repeats over time to form the binary control signal, the control signal pattern having an instantaneous time average at each time point, the instantaneous time average being averaged over time over the control signal pattern used at the time in question, the method comprising the steps of: providing a variable current to the LED or different loads subsequently connected to the LED driver. The variable current has several different instantaneous time averages of a binary control signal; for each of the several different instantaneous time averages, a corresponding measurement of a quantity is obtained, which is then related to at least one of or at least one of the following: first, the average current flowing through the LED or load, and second, the average light intensity emitted by the LED, thereby obtaining a first relationship of measurement of the absolute or relative current / light intensity change according to the instantaneous time averages; configuration information is determined based on the first relationship to map each of the several different absolute or relative current / light intensity values ​​to the corresponding value of the instantaneous time average; and the configuration information is stored.

[0040] In some implementations, the relative current and / or light intensity are relative to the maximum current or light intensity, respectively.

[0041] In some embodiments, the method further includes: identifying a second relationship between an instantaneous time average and an absolute or relative current / light intensity; identifying a parametric model of the second relationship, the parametric model comprising one or more at least piecewise continuous functions determined by a set of parameters, the parametric model being used to determine a mapping of each of several different absolute or relative current / light intensity values ​​to a corresponding instantaneous time average; and determining values ​​of parameters that provide a fit of at least one function to the second relationship, the second relationship being evaluated for each of the measured values ​​and corresponding instantaneous time averages.

[0042] In some implementations, the configuration information includes parameter values.

[0043] In some implementations, the second relationship includes The quotient or reciprocal of the quotient, where g is a non-constant function with respect to x; x is the instantaneous time average; and y is the absolute or relative current / light intensity.

[0044] In some implementations, one or more functions include at least one first-order or higher-order polynomials.

[0045] In some implementations, one or more functions comprise a polynomial of degree 2 or higher.

[0046] In some implementations, one or more functions comprise two first-order or higher-order polynomials interconnected by multiplication with an sigmoid function.

[0047] In some implementations, one or more functions are based on functions Defined by, among which about The absolute value of the first derivative, in other words It is decreasing, among which It is an absolute or relative current / light intensity.

[0048] In some implementations, one or more functions include functions defined based on two different polynomials and an sigmoid function that multiplies the two polynomials together.

[0049] In some implementations, the two polynomials are chosen to approximate the overall shape of the second relation on either side of a local minimum or maximum value of the second relation by selecting parameter values.

[0050] In some implementations, the sigmoid function is defined based on two functions s' and s'', which are chosen to produce a continuous function at the expected local maximum of the absolute or relative current / light intensity in the second relationship, such that the derivative of the continuous function does not change sign.

[0051] In some implementations, the configuration information includes configuration information for determining and storing at least one of the following: first, at least two different pulse numbers for the control signal mode; second, at least two different temperatures of the LED driver when the measurement values ​​are obtained; and third, at least two different maximum allowable currents to be supplied to the LED driver.

[0052] The present invention also relates to a method for operating an LED driver to drive an LED, the method comprising: configuring the LED driver using a method of the type described above; and operating the LED driver by determining a binary control signal corresponding to a desired current / light intensity, the determination of the binary control signal being performed using configuration information, and providing a variable current based on the determined binary control signal to the driver output of the LED driver.

[0053] In some implementations, determining the binary control signal includes adjusting the width of one or more pulses within the binary signal pattern.

[0054] In some implementations, determining the binary control signal includes adjusting the number of pulses within the binary signal pattern.

[0055] In some implementations, the pulses within a binary signal pattern are adjusted to be distributed across a predetermined set of available pulse positions across the binary signal pattern.

[0056] In some implementations, determining the binary control signal includes: determining the number of pulses within the control signal pattern; and selecting the binary control signal based on the interpolation between the corresponding first and second expressions of the binary control signal.

[0057] In some implementations, the first expression of the binary control signal is determined based on a configuration parameter of the number of first pulses within the control signal pattern.

[0058] In some implementations, the second expression of the binary control signal is determined based on a configuration parameter for the number of second pulses within the control signal pattern.

[0059] In some implementations, interpolation is performed on the current number of pulses within the control signal pattern, which is between the first number of pulses and the second number of pulses.

[0060] In some implementations, the first and second expressions of the binary control signal are determined based on the corresponding pulse width within the control signal pattern.

[0061] In some implementations, interpolation is performed on the first and second expressions of the binary control signal using the square root of the normalized value of the current pulse count as a weighting factor.

[0062] The method also relates to an LED driver, the LED driver comprising: a first channel electric driver output, the LED driver being arranged to provide current to the first channel electric driver output according to a first channel binary control signal; and a first channel timer function configured to update a first channel timer counter variable at a first channel timer frequency, the first channel timer function being configured to compare the first channel timer counter variable with a set first channel comparison value for each update, and the first channel timer function being configured to switch the first channel binary control signal between a first binary state and different second binary states if the first channel timer counter variable has reached or exceeded the set first channel comparison value.

[0063] In some implementations, the duration of the repeated first-channel control signal pattern of the first-channel binary control signal is longer than the time taken for the first-channel timer counter variable to complete a full cycle, and the first-channel control signal pattern includes one or more pulses.

[0064] In some implementations, the LED driver is arranged to update the set first channel comparison value at least twice during each individual first channel control signal mode.

[0065] In some implementations, the frequency of the first channel timer is at least as high as or higher than the predetermined clock frequency of the CPU included in the LED driver.

[0066] In some implementations, the first channel timer function uses a DLL (delay-locked loop) function, which is configured to divide the timer input clock period into several fractional steps.

[0067] In some implementations, the first channel timer function is cyclic.

[0068] In some implementations, the LED driver further includes: a RAM (random access memory) region; and a comparison value copying function operable to update a first channel timer function using an updated set first channel comparison value from the RAM region.

[0069] In some implementations, the comparison value copying function is a DMA (Direct Memory Access) channel or an ISR (Interrupt Service Routine) executed in response to an IRQ (Interrupt Request).

[0070] In some implementations, the comparison value copy function is configured to respond to an interrupt that occurs when the first channel timer counter value reaches its full cycle value, and this response includes copying a new first channel comparison value from the next or subsequent memory location in the RAM region.

[0071] In some implementations, the DMA channel is configured as a ring, such that after it has reached the end of the RAM region, it resumes copying from the beginning of the RAM region again.

[0072] In some implementations, the size of the RAM region is at least 4 comparison values, such as at least 8 comparison values, such as at least 16 comparison values.

[0073] In some implementations, the CPU is configured to update the RAM region in response to an updated dimmer setting received from the LED driver by using an updated set of first channel comparison values ​​specifically selected to reflect the updated dimmer setting.

[0074] In some implementations, the individual pulse of the binary control signal is longer than the time it takes for the first channel timer counter variable to complete a full cycle.

[0075] In some implementations, the LED driver is arranged to vary the number of pulses provided across the first channel control signal mode in order to achieve different desired instantaneous time averages of the first channel control signal mode.

[0076] In some implementations, the LED driver is arranged to vary the pulse width of one or more pulses in a set of one or more pulses provided across the first channel control signal mode in order to achieve different desired time-averaged currents supplied to the output of the first channel electric driver.

[0077] In some implementations, the LED driver is arranged to gradually increase the time-averaged current applied to the output of the first channel driver by first increasing the pulse width of one or more pulses provided across the first channel control signal mode and then increasing the number of pulses provided across the first channel control signal mode.

[0078] In some implementations, the LED driver is arranged to distribute at least one pulse across the first channel control signal pattern onto a set of predetermined and distributed time pulse positions across the first channel control signal pattern.

[0079] In some implementations, the distribution is based on the bit reversal of the binary representation of a number, which in turn describes or corresponds to the index of a pulse within the first channel control signal pattern.

[0080] In some implementations, the duration of the first channel control signal mode is constant across different desired time average currents supplied to the first channel electric driver output.

[0081] In some embodiments, in addition to the first channel driver output, the LED driver further includes: a second channel driver output, the LED driver being arranged to provide current to the second channel driver output according to a second channel binary control signal; and a second channel timer function, the second channel timer function being configured to update a second channel timer counter variable at a second channel timer frequency, the second channel timer function being configured to compare the second channel timer counter variable with a set second channel comparison value for each update, and the second channel timer function being configured to switch the second channel binary control signal between a third binary state and different fourth binary states if the second channel timer counter variable has reached or exceeded the set second channel comparison value.

[0082] In some implementations, the duration of the repeating second-channel control signal pattern of the second-channel binary control signal is longer than the time taken for the second-channel timer counter variable to complete a full cycle, and the second-channel control signal pattern includes one or more pulses.

[0083] In some implementations, the LED driver is arranged to update the set second channel comparison value at least twice during each individual second channel control signal mode.

[0084] In some implementations, the duration of the first channel control signal mode is an integer multiple of the duration of the second channel control signal mode, such as being the same as the duration of the second channel control signal mode.

[0085] In some implementations, the time offset of the first channel binary control signal relative to the second channel binary control signal is less than the offset value of the first channel control signal mode.

[0086] In some implementations, the LED driver includes three or more electrical driver outputs, and the LED driver is arranged to apply a corresponding current to each such driver output according to a corresponding binary control signal.

[0087] In some implementations, each of the binary control signals has a corresponding repeating control signal pattern whose duration is an integer multiple or fraction of the duration of each of the repeating control signal patterns of the other binary control signals, such as being the same as the duration of each of the repeating control signal patterns of the other binary control signals.

[0088] In some implementations, the LED driver is arranged to time-offset each of the binary control signals in the binary control signals by a value less than the duration of the first channel control signal mode.

[0089] In some implementations, the corresponding time offsets of at least three control signal modes are distributed across a set of predetermined and distributed time offsets of the control signal modes across the first channel.

[0090] In some implementations, the corresponding time offsets of at least three control signals are different.

[0091] In some implementations, the time offset is calculated based on the bit reversal of the binary representation of a number, which in turn describes or corresponds to the index of the corresponding electrical driver output.

[0092] In some implementations, the time offset is scaled over the duration of the first channel control signal mode.

[0093] In some implementations, the clock frequency of the first channel timer function is at least 50 MHz, such as at least 100 MHz, such as at least 150 MHz.

[0094] In some implementations, the LED driver is a constant voltage LED driver.

[0095] In some implementations, the first channel binary control signal is arranged to switch between a first binary state and a second binary state at a frequency of at least 20 kHz, preferably at least 25 kHz, and then switch back to the first binary state.

[0096] In some implementations, the LED driver also includes a DC bus, which comprises ceramic capacitors. Attached Figure Description

[0097] Figure 1a It is a graph showing how the current changes with the voltage; Figure 1b This is a graph showing the logarithmic change of current with respect to voltage, where current and voltage correspond to... Figure 1a The current and voltage shown; Figure 1c It is a graph showing the changes in current with respect to voltage at several different logarithmic values; Figure 2 This is an overview of LED drivers; Figure 3 This is a flowchart illustrating the interrelationships between the first, second, and third aspects; Figure 4a It is a diagram showing a voltage signal pulse train with straight edges; Figure 4b It is a graph showing a voltage signal pulse train with sloping edges; Figure 4c It is a series of charts showing the progress of pulse train design for increased PWM-like duty cycles, from the top chart to the bottom chart; Figure 4d It is a schematic diagram of a defined pulse; Figure 5 This is an overview of LED drivers; Figure 6 This is a chart illustrating the order in which different pulses are introduced in a pulse mode; Figure 7 This is a flowchart illustrating how to introduce pulses in pulse mode; Figure 8 It is a graph showing how light intensity changes according to the duty cycle of a PWM-like circuit; Figure 9 Similar to Figure 8 However, it shows information from a log-log graph; Figure 10 These are graphs showing curve fitting and graphs showing the sigmoid function; Figure 11 This is a flowchart for operating an LED driver; Figure 12 This is an overview diagram illustrating LED driver configuration settings; Figure 13 This is a graph showing an alternative S-shaped function; Figure 14 This is a flowchart for operating an LED driver; Figure 15 This is an overview of the LED drivers connected to the LED load; Figure 16 This is an overview of LED drivers; Figure 17 This is a circuit diagram showing a push-pull circuit; Figure 18 This is the circuit diagram for an LED driver; Figure 19 and Figure 20 This is an example Figure 18 A series of circuit diagrams for analysis of a portion of the circuit diagrams; Figure 21 It is a graph showing a set of curves illustrating how voltage changes with output voltage and temperature; Figure 22 It is a graph showing a set of curves illustrating how voltage changes with output voltage and resistance; Figure 23 This is an overview of LED drivers; Figure 24 This is the circuit diagram for an LED driver; Figure 25 This is the circuit diagram for an LED driver; Figure 26a and Figure 26b The shape of the corresponding voltage pulse, which varies with time, is shown; and Figure 27 This is a flowchart for operating an LED driver.

[0098] The accompanying figures share the same reference numerals for the same or corresponding parts. Some reference numerals are used with or without one or more apostrophes, where a reference numeral without any apostrophes usually refers to that part, while multiple apostrophes following the reference numeral indicate a specific channel. For example, an LED driver switch is generally referred to as 180, while 180'' refers to an LED driver switch for the second channel. Detailed Implementation

[0099] It may be difficult to predict the behavior of any particular LED light source across a wide range of drive power. Specifically, dimming behavior may not be linear, or it may even be undefined.

[0100] To understand this, we can study the behavior of an LED connected in series with a current-limiting resistor (a common setup) in terms of light power as a function of voltage. The current through the LED is described by the Shockley diode equation: , in It is the current through the diode. It is the current scaling factor. It's voltage. It is thermal voltage, and It is the ideal factor. For LEDs, Typically on the order of nA, thus making the exponent term dependent on normally occurring currents. Dominant. Therefore, the expression can be simplified to , Furthermore, solve : .

[0101] As an example of how current changes when an LED is connected in series with a resistor, a typical LED can be observed to have a maximum forward current of 0.03A. The LED used in this example is the Nichia NSPW500BS. The parameter values ​​can be easily read in the software program LTSpice. =0.27nA and =6.79. Thermal voltage Through expressions Among the obtained, It is Boltzmann's constant. It is absolute temperature, and It is the elementary charge. At room temperature (300K), this equation produces... =25.852mV.

[0102] This then received .

[0103] At 0.03A, =3.25V. If an LED and a resistor connected in series are driven at 5V, then at the nominal current, there will be 1.75V across the resistor, which produces resistance. Therefore, the equation for voltage under a given diode current becomes: .

[0104] The graph of this equation can be seen in... Figure 1a And the version with logarithmic scale current is in Figure 1bAs shown in the figure, the graph has an approximately exponential region due to the LED and an approximately linear region due to the resistor. The light power from the LED is proportional to the current, and the human eye has high sensitivity at lower light levels.

[0105] In the above calculations, all physical constants are approximate values. Fluctuations during manufacturing will cause parameter values ​​to shift between diodes, resulting in LEDs that do not have exactly the same properties and characteristics. Figure 1c The result is shown after slightly shifting the input parameter value. From Figure 1c It is clear that at lower voltage levels, the fluctuations become more pronounced. For LED strips that consist of many individual LEDs, this means that the LEDs will start emitting light in different ways as the voltage drops, which is undesirable.

[0106] To address this issue, the properties of each LED can be precisely measured during or after manufacturing, and only LEDs with similar properties can be used together. This is called "binning" and is quite expensive. Alternatively, LEDs can be avoided at lower voltage levels and instead always used at higher voltage levels, such as full power.

[0107] In the latter case, conventional LED drivers use a switching mode to dim the LED light source, rapidly switching the LED on and off at a sufficiently high frequency so that the resulting flicker is imperceptible to the human eye. This type of driver, using pulsed power, provides a more perceptibly uniform dimming behavior for a wide range of LED light sources with varying properties. Such LED drivers are called "constant voltage" drivers.

[0108] One way to provide this pulsed power to an LED light source is by using an LED driver with pulse width modulation (PWM). A PWM signal can be described in terms of pulse width and period time (both measured in seconds) or alternatively in terms of its frequency (measured in Hz) and duty cycle (pulse width as a percentage of period time).

[0109] In a microcontroller-based PWM driver, a timer can be used to count the number of clock cycles and compare that count to a specific comparison value. As long as the counter is below the comparison value, the output signal is high (e.g., at a predetermined full LED power); and when the timer exceeds the comparison value, the output signal goes low (e.g., at zero LED power). The timer then resets upon reaching the cycle time. Therefore, the time resolution of the PWM driver is determined by the clock frequency of the timer, and the PWM frequency is the clock frequency divided by the cycle length (measured in clock cycles). For example, if a timer with a clock frequency of 25MHz uses 100,000 steps (clock cycles), a PWM signal with... A PWM signal with a periodic frequency (PWM frequency). The time resolution is... In this case, the minimum achievable duty cycle is for a comparison value of 1, and the duty cycle is 1 / 1e5 = 0.001%.

[0110] The human eye is highly sensitive to low lighting levels, meaning that the perception of light is primarily based on relative changes in light intensity. A change from 100% light intensity to 10% light intensity is perceived as roughly the same as a change from 10% to 1%. In other words, dimming an LED driver from 100% to 1% of the time-averaged light intensity is perceived as dimming the LED light source to about twice the degree of dimming compared to dimming from 100% to 10%.

[0111] At the same time, the human eye can perceive extremely small amounts of light. For example, it has been found that when an LED light source with a nominal current of 200mA is driven with a total current as small as 5μA (which is 1 / 40,000th of the nominal current), light emitted from the LED light source can be perceived.

[0112] To dim this LED light source to such a low emitt power using PWM, a duty cycle of 0.0025% would be required. If this number is rounded down to 0.001%, it means that a microcontroller with a clock frequency of 25MHz using a PWM frequency of 250Hz would be almost unable to achieve such a low light level.

[0113] However, even if such low-level lighting were possible, the increment from the lowest dimming level to the second lowest dimming level would be 100%, which would be a very noticeable light intensity increment for the human eye. Specifically, this would be visible when using color-adjustable lighting (such as RGB-controlled lighting), as the colors would change significantly due to brightening or dimming one color channel to near the lowest possible dimmer setting.

[0114] It has been shown that the human eye cannot perceive changes in light intensity smaller than approximately 1.5%. Therefore, if an LED driver is to dim continuously to darkness and be perceived as continuously dimming across the entire dimming intensity range, all dimming steps across the entire intensity range must be less than 1.5%. If the maximum step size is set to 1% for margin, such a 1% step size needs to be taken at approximately 0.001% of the light intensity, close to the lowest dimming level. This means that at these levels, the light intensity resolution needs to be at most 0.00001%.

[0115] Furthermore, generating short PWM pulses risks having wavelengths comparable to those of LED strips using controlled LEDs. LED strips are typically 5 to 10 meters long to avoid unacceptable voltage drops across them. The speed of light in copper is about half that in a vacuum, meaning a 40-ns pulse entering an LED strip would have a pulse length of approximately 6 meters. ( ).

[0116] A 6m pulse is on the same order of magnitude as the length of a typical LED strip. Since the pulse length is comparable to the length of the LED strip, this means that microwave theory, rather than traditional voltage equations, must be used to model the load. In other words, the generated pulse will be reflected to some extent at the ends of the LED strip, and the pulse, along with its reflection, risks creating a standing wave pattern along the LED strip. This, in turn, means that the applied voltage will not be uniform along the LED strip, and individual LEDs will emit light at different powers depending on the characteristics of the applied voltage pulse.

[0117] At the same time, it is also necessary to avoid visible flicker and noise originating from the driver or any power source used. Since PWM involves flicker by definition, the flicker needs to be kept at a frequency that is not perceptible as uncomfortable. This is achieved by using a sufficiently high PWM frequency.

[0118] The standard IEEE PAR1789 (“IEEE Recommended Practices for Modulating Current in High-Brightness LEDs for Mitigating Health Risks to Viewers”) defines “flicker-safe” driving with loads alternating between 0% and 100% as using a driving frequency of at least 3 kHz. The uncomfortable effects of flicker perceived by humans decrease as the average light intensity decreases. The inventors have noted that maintaining a frequency above 3 kHz is less critical when operating near the lowest light intensity of a driver capable of dimming to complete darkness.

[0119] At a 3kHz PWM frequency and a 0.001% duty cycle, the pulse width becomes extremely short; 0.001% of the 333μs period time is 3.33ns. This is equivalent to a 300MHz PWM frequency, and for a square wave retaining its first overtone, this implies a frequency in the GHz range. At such high frequencies, the wavelength becomes so short that the voltage supply cable begins to act as an antenna, potentially causing crosstalk with nearby cables. Standing waves in the voltage supply cable can also occur, resulting in, for example, LED strips supplying different light intensities along different sections of their length.

[0120] To mitigate these issues, the sharp current derivative of the PWM signal can be removed, resulting in a reduction or elimination of the highest frequency. However, this means the LED light source will spend more time in total across the voltage range between 0V and its nominal voltage (e.g., 12V or 24V). Since the behavior of the LED light source may be undefined in this range, non-uniform light may be produced during dimming. Therefore, for an LED light source to operate acceptablely with a 3kHz PWM signal and a 0.001% duty cycle, the light intensity emitted by different LEDs in the same LED light source (or “LED load”) should preferably exhibit the same behavior in terms of the time-averaged current variation in light output across the range from 0V up to its nominal voltage. This is possible, but then it is conventionally necessary to carefully sort the LED light sources to be used with the same LED driver to minimize differences in their forward voltages.

[0121] However, at higher light intensities, the current derivative will contribute only a very small share of the total light, so far from the minimum dimming intensity, any difference between individual LED light sources becomes negligible.

[0122] For many LED drivers, unwanted noise becomes a problem as the delivered power varies. For example, a constant-voltage LED driver will draw varying current as the delivered power changes, which can cause any capacitors and inductors in the power supply to start making noise. If the PWM frequency is within the human audible sound spectrum, such noise can become noticeable when the current demand is high enough.

[0123] By using a PWM frequency exceeding approximately 25kHz, any generated noise will be imperceptible to the human ear. This high PWM frequency also allows for the use of surface-mount capacitors while still providing sufficient capacitance to draw a relatively smooth current, resulting in smaller current and voltage variations in the power supply. Smaller voltage variations mean that any electrolyte capacitors in the power supply are not under as heavy a load during operation, effectively extending their lifespan.

[0124] For these reasons, a PWM frequency of at least 25kHz is advantageous, provided the current drawn by the LED light source is not negligible. Conversely, if a higher PWM frequency is chosen, ceramic capacitors, which are typically associated with long lifespans, can be used.

[0125] The LED driver implementations described herein can be constant voltage LED drivers. They can be used to drive individual LEDs and / or LED strips, for example, at a constant voltage of 12V or 24V. LED strips are typically 5m or 10m in length and can include multiple individual LEDs, which can then be connected in series with one or more resistors. Groups of one or more individual LEDs in the LED strip can be connected in parallel, allowing the user to adapt the total length of the LED strip by cutting it at predetermined points along its length.

[0126] Figure 2 An exemplary LED driver 100 with N channels is illustrated. The presence of N channels in the LED driver means that it can simultaneously drive N LED loads, such as N different LEDs and / or LED strips. The channels can be driven independently, allowing each channel to be dimmed independently of the others. For multi-color LEDs, different colors can be controlled on different channels, using RGB (red, green, blue) LEDs as an example. As an example, the LED driver 100 may have twelve channels, but it should be understood that in various embodiments, the number of channels may be at least one, at least two, at least four, or at least ten; and / or the number of channels may be at most one hundred, at most fifty, or at most twenty.

[0127] LED driver 100 includes microcontroller 130, which is arranged to accept a corresponding control input 110', 110'', 110''' for each of N channels. Each such control input 110', 110'', 110''' may be in the form of a digital decoded number or an analog signal representing the desired dimming value of the channel in question.

[0128] The microcontroller 130 includes a CPU 131 and / or other hardware computing logic circuitry, such as a GPU. The microcontroller 130 may also include corresponding timer functions 132', 132'', 132''' for each of the channels. Each such timer function 132', 132'', 132''' may further include corresponding DLL (Delay-Locked Loop) functions 133', 133''', 133'''. The microcontroller 130 may also include a RAM (Random Access Memory) region 134 and / or a comparison value copying function 135. For each channel, the microcontroller 130 may include corresponding binary control signal generators 136', 136'', 136''' and / or corresponding pulse compensators 137', 137'', 137'''. These components will be described in more detail below.

[0129] The LED driver 100 (and specifically, the microcontroller 130) may include control software configured to execute on the CPU 131 and / or any other computing hardware component of the microcontroller 130, and, when executed, to perform the steps described herein performed by the microcontroller 130. Depending on the circumstances, one or more of the steps described herein may also be performed using purely hardware-based logic components. Each of components 132, 135, 136, and 137 may be implemented as: a discrete hardware component; a software module; or a combination thereof. The LED driver 100 may also include a communication bus for communication between the various hardware and software components of the LED driver 100 described herein.

[0130] In addition, the LED driver 100 includes or is connected to an internal or external power supply 120. The LED driver 100 may also include a DC bus 121, which can then provide a constant voltage of at least eight volts and / or up to 48 volts. The power supply 120 may be arranged to power a microcontroller (e.g., using 3.2V) and / or the LED driver switches 180', 180'', 180''' (e.g., at 5.0V) of the LED driver 100 via the LED bus 121. One such LED driver switch 180', 180'', 180''' for each channel is arranged to provide a variable current to the corresponding driver output of the channel in question. The time average of the variable current corresponds to the desired dimming value for that channel. The DC bus may include ceramic capacitors.

[0131] The LED driver 100 may also include corresponding drive circuits 140', 140'', 140''' for each channel. The drive circuit 140 is arranged to provide an analog switch control signal to the LED driver switch 180 to control the LED driver switch 180 to provide drive current to the LED driver output. The drive circuit 140 may include software-implemented functionality, but in this example, the functionality of the drive circuit 140 is fully or at least partially implemented in hardware.

[0132] The LED driver 100 also includes features for overcurrent and overtemperature protection. These features are standard in themselves and will not be described in detail herein.

[0133] The inventors have developed the LED driver described herein based on three aspects that can be used independently or freely combined.

[0134] In the first aspect, the pulses generated by the microcontroller 130 have dynamic frequencies, such that low frequencies are used for low intensities and high frequencies are used for higher intensities.

[0135] In the second aspect, a model is developed that allows for the reduction of the mismatch between expected light intensity and actual light intensity.

[0136] In a third aspect, the drive circuit 140 is designed to have dynamic rise and fall times, such that smooth pulses with long rise and fall times are used for low intensity, while faster pulses with shorter rise and fall times are used for higher intensity.

[0137] The interrelationships between these aspects are... Figure 3 The simplified overview shown is for illustrative purposes.

[0138] from Figure 3 Starting at the top and moving downwards, control inputs are generated, identified, or received by the microcontroller 130. Control inputs can be any suitable format of analog or digital signal that can be interpreted by the microcontroller 130. Generally, the input signal can represent the desired dimming level, such as on a scale from 0% (no light emitted from the LED) to 100% (full light). In the example shown, the control value received by the microcontroller 130 represents a "10% dimming level".

[0139] Then, a pulse signal corresponding to the control input is generated according to the functionality of the first aspect. More specifically, the pulse signal is a periodic binary signal generated by one or more timer functions 132 of the microcontroller 130. The periodicity of the periodic binary signal may coincide with the lowest supported frequency of the microcontroller 130, and the highest supported frequency of the microcontroller 130 may be a multiple of the lowest supported frequency. In the example, the lowest supported frequency is 250 Hz, and the highest supported frequency is 32 kHz.

[0140] In this document, the resulting periodic binary signal is generally referred to as a "binary control signal". The binary control signal may be in the form of a square wave, consisting of alternating low and high values ​​(such as 0 and 1). The binary control signal may have a control signal pattern; in other words, it has a binary pattern that repeats over time to form the binary control signal. At each given point in time, the control signal pattern has an instantaneous time average. This instantaneous time average is a value determined as the time average over the control signal pattern used to form the binary control signal at that time. The control signal pattern may change over time, for example, in response to a change in control input. In some implementations, the (potentially time-varying) repetitive control signal pattern completely defines the binary control signal over time, such that the binary control signal does not include any additional waveforms other than the repetitive control signal pattern. In any case, the control signal pattern can be arranged to always completely define the instantaneous time average of the binary control signal.

[0141] like Figure 3 As shown, the generated binary control signal may include pulses of different pulse widths. Generally, the control signal pattern may include one or more pulses, each pulse having the same or different pulse widths, and / or the time between individual pulses may be the same or different. Thus, the resulting binary control signal is similar to a PWM signal, but may be more complex than a conventional PWM signal. In some embodiments, the binary control signal has a duration of at least 0.1 ms, such as at least 0.5 ms, such as at least 1 ms, or even at least 2 ms. In some embodiments, the binary control signal has a duration of up to 1000 ms, such as up to 500 ms, such as up to 100 ms, such as up to 50 ms, such as up to 20 ms, or even up to 10 ms.

[0142] Then, based on the functionality of the second aspect (and specifically, the pulse compensator 137 of the microcontroller 130), a compensated binary control signal can be generated based on the binary control signal. This compensation can then take into account a determined model with respect to the properties of at least one of the following: the drive circuit 140, the LED drive switch 180, the LED load 10 actually powered by the LED driver 100, and a predetermined typical LED load to be powered by the LED driver 100.

[0143] More specifically, since any pulse rise time or fall time has a non-zero width for any drive circuit, the actual relative current drawn from the LED load will not perfectly match the average duty cycle of the binary control signal. If the instantaneous pulse frequency of the binary control signal is increased while keeping the average PWM duty cycle constant, the pulse width becomes narrower, and the rise and fall times will have a greater impact on the overall match between the average duty cycle and the average relative current drawn from the LED load 10. Therefore, the model can be designed to approximate the inverse model of the drive circuit 140, such that the compensated signal is tailored to the specific behavior of the drive circuit 140 (and correspondingly to the LED drive switch and LED load 10 or a typical LED load).

[0144] like Figure 3 As shown, the compensated binary control signal may have a format corresponding to the uncompensated binary control signal, such as having a straight edge (which may be a pure binary signal), but the pulse configuration may be changed, for example, in terms of the individual pulse width in the control signal mode.

[0145] Maintaining uniform light intensity across the entire length of an LED strip at low light intensities requires pulses that are physically much longer than the LED strip itself. Therefore, pulse widths that are too short cannot be used, and thus the pulses must be longer without causing excessive light. If the rise and fall times are too long, the pulse becomes too wide, which also fails to promote maximum power output. However, for higher currents and higher frequencies, this results in more power dissipation than required. According to the functionality of the third aspect (and more specifically, the driver circuit 140), the analog control signal is provided as a (non-pure binary) signal with dynamic rise and fall times, which depend on the instantaneous time average of the binary control signal.

[0146] Then, the analog control signal is provided to the LED driver switch 180, which is located at the LED driver output 182 (see...). Figure 5 and Figure 15 The current to be supplied to the LED load 10 is generated at point 120. The current is supplied using power supply 120. The average current supplied to the LED load 10 relative to the maximum current can be the input to the control input 110 of the LED driver 100.

[0147] It should be recognized that each of the first, second, and third aspects offers different advantages in terms of the final pulse current supplied to the LED load 10, but not all three aspects are necessary in all embodiments. Specifically, the first, second, and third aspects can be used individually, in pairs, or in combination of all three to provide the various advantages described herein and desired or required in the particular use case considered. For example, the control input can be in the form of a PWM-like signal directly provided to the functionality according to the second aspect. The binary control signal can be used as is and directly provided to the functionality according to the second aspect. The compensated control signal can be directly provided to the LED driver switch 180. And so on.

[0148] Figure 4a An exemplary binary control signal is shown in the form of a square voltage (Y-axis) waveform that varies according to time (X-axis), alternating between low values ​​(0V in this case) and high values ​​(24V). The binary control signal is arranged to control the on / off or high / low state of an LED load (such as a single LED lamp or LED strip as described above). It should be appreciated that the binary control signal may alternatively be a current signal. Furthermore, Figure 4a The signals shown are analog signals, but it should be recognized that in some implementations, the binary control signal can be a decoded signal, such as a digital decoded signal conveying information about the corresponding waveform. The binary control signal can also be digitally generated by the timer 132 of the microcontroller 130 and used as... Figure 4a The analog signal output is shown in the figure.

[0149] For a binary control signal in the form of a repetitive square wave with only one pulse, the pulse width divided by the period width represents the percentage of time the binary control signal is active. As discussed above, this is called the duty cycle, for example, in PWM. Figure 4a In the provided example, the pulse width is 60 clock cycles, and the period time is 200 clock cycles, resulting in a duty cycle of 60 / 200 = 30%. The frequency is provided as the reciprocal of the period time. Timers integrated into conventional microcontrollers typically use an internal clock to count the number of clock cycles until a desired maximum counter value is reached, after which the counter restarts. The counter value is compared to a compare value, and the microcontroller generates a signal that is high if the counter is below the compare value and low if the counter is above the value, thus generating a PWM signal.

[0150] The time resolution of this signal is determined by the speed of the internal clock: the faster the clock counts, the more frequently the signal can be changed from low to high or vice versa. Most conventional microcontrollers use a master clock that can be scaled down to generate the timer clock. However, some newer microcontrollers have timers that use delay phase-locked loops (DLLs), effectively providing the same effect as scaling up the master clock.

[0151] In an embodiment of the invention, the LED driver 100 may be a constant voltage LED driver arranged to drive the LED load 10 at a constant voltage. Instead of smoothing current changes to achieve various dimming levels, the LED driver 100 may be arranged to provide a voltage signal that switches from zero to a nominal constant voltage (such as 12V or 24V) or from some other low value above zero to said nominal constant voltage. Then, according to, for example... Figure 4a The square wave signal shown may be switched according to a square wave signal that has been compensated and / or modified as described in conjunction with the various embodiments described herein. It should be understood that this also means that the drive current supplied to the LED load 10 will be a varying (pulse) current as the voltage switches between its low and high values.

[0152] As mentioned above, there are problems with using the PWM method to generate extremely low light levels in LED load 10, both because the dimming step size at extremely low light levels becomes relatively large even with a fast microcontroller, and because of the problem of extremely short pulses generating standing waves along the driven LED strip.

[0153] Another thing to consider is how the voltage supplied to the LED load 10 changes over time; in other words, the voltage derivative as it switches from a low value to a high value and vice versa. This is in Figure 4b As shown in the example, this figure corresponds to Figure 4a The waveform shown is a variation of the voltage applied by the LED driver 100 to the LED load 10.

[0154] As from Figure 4b It is clear that the voltage takes a certain amount of time to change from a low value to a high value, and vice versa, thus causing the LED load 10 to emit light before the rising voltage has reached its maximum level and correspondingly before the falling voltage has reached its minimum level. The nature of the voltage derivative at the edge of the voltage supplied to the LED load 10 affects the behavior of the light emitted by the LED load 10 at a low light value. The challenge is to enable the LED driver to achieve uniform and smooth illumination at a low light value in terms of both the dimming behavior of each individual LED and the light emitted along the LED strip.

[0155] This invention addresses the problem by defining a minimum acceptable PWM frequency for use at low light levels, thereby allowing a sufficiently wide pulse width to achieve a defined low average optical power over time without significant pulse edge effects due to voltage derivatives. The optical power is adjusted from a low or minimum dimming value, and the pulse width can initially be increased until a specific pulse width is reached where edge effects are negligible relative to the duty cycle, and at this point, another pulse can be introduced into the repetitive pulse pattern. The newly added pulse can have a narrower pulse width than the existing pulses. It should be understood that this signal is then no longer a PWM signal, but a more complex signal.

[0156] Then, the pulse width of the newly added pulse can be increased as the dimming value increases, up to a specific pulse width, and then additional pulses can be added, and so on.

[0157] Once the desired maximum number of pulses has been reached in the repetitive control signal mode, the pulse width of all pulses can be increased simultaneously to further increase the dimming value until the full light spot, at which the pulses completely fill the signal and the signal is actually at its high value for 100% or close to 100% of the time.

[0158] Figure 4c An example of how this can be achieved is shown. Figure 4c Each line in the diagram illustrates two complete cycles (each cycle 4,000 μs long) of the repetitive waveform signal supplied to the LED load 10, where the dimming value increases from top to bottom. For clarity, the voltage derivative is slightly exaggerated. Figure 4c Each row in the table is represented by the instantaneous time-averaged control signal mode value of the row in question, as marked on the left side of the table. Figure 4c The X-axis in the diagram represents the clock cycle.

[0159] In the first waveform, a 250Hz signal is introduced. Due to the finite derivative and the short pulse, the pulse does not have time to reach the high value of 24V before it ends. However, as the dimming value increases, the voltage at the pulse peak increases. Once the pulse has reached a sufficiently wide pulse width to make the edge behavior negligible, an additional pulse is introduced (see line 3). The pulse width of this added pulse is then increased until the specified pulse width is reached. Thereafter, new pulses are introduced between existing pulses until the maximum number of pulses has been reached. Finally (from line 7 onwards), the pulse widths of all pulses are increased simultaneously until the maximum light is reached.

[0160] An alternative way to achieve this would be to increase the pulse width up to a specific pulse width and then reduce the period time of the repetitive signal. However, as will be seen below, this is sometimes not a feasible solution in implementations where a fixed total time for the repetitive signal pattern is desired (such as for multi-channel LED drivers).

[0161] Figure 5 This is another view showing an exemplary LED driver 100 with three channels, for example, for driving an RGB LED or LED strip 10. Three different dimming values ​​(such as red, green, and blue) are received by the LED driver 100, and specifically by a microcontroller 130. The microcontroller 130 outputs a corresponding binary control signal to each of the three analog drive circuits 140', 140'', 140'''. Each of the drive circuits 140', 140'', 140''' generates a corresponding analog control signal, which is fed to corresponding switches 180', 180'', 180''', each of which in turn has driver output terminals 182', 182'', 182''', which provide a corresponding changing (pulse) voltage signal to each of the three corresponding inputs on the LED load 10, providing the desired dimming and color settings.

[0162] First aspect of the invention Now let's turn to the first aspect, see reference... Figure 2 and Figure 5 The LED driver 100 includes a first channel electric driver output 182', which is arranged to supply current to the first channel electric driver output according to a first channel binary control signal provided by a binary control signal generator 136.

[0163] The first channel timer function 132' can be configured to update the first channel timer counter variable at the first channel timer frequency. The first channel timer function 132' is configured to compare the first channel timer counter variable with a set first channel comparison value for each such update. Furthermore, the first channel timer function 132' is configured to switch the first channel binary control signal between a first binary state and different second binary states if the first channel timer counter variable has reached or exceeded the set first channel comparison value. The corresponding configuration can also be applied to any second channel or additional channel timer function 132.

[0164] In some implementations, the clock frequency of timer function 132 is at least 50 MHz, such as at least 100 MHz, such as at least 150 MHz. In some implementations, the binary control signal is arranged to switch between a first binary state and a second binary state at a frequency of at least 20 kHz, such as at least 25 kHz, and then switch back to the first binary state. For at least some of the incoming dimming values, this switching can be an average switching frequency over time.

[0165] As described above, smoothly dimming to complete darkness without using a high-resolution duty cycle of conventional PWM is difficult. Worse still, calibration will be rather coarse, resulting in poor in-channel matching.

[0166] In practice, the inventors used an STM32G484 microcontroller purchased from ST Microelectronics as the microcontroller 130. This microcontroller includes a CPU 131 capable of performing floating-point calculations at a clock frequency of 170 MHz, and it also features direct memory access (DMA) functionality, allowing data to be copied from one memory region to another without directly involving the CPU 131. The STM32G484 also features six "high-resolution timers" (also called "HRTIMers") serving as corresponding timer functions 132 for each of the supporting channels. The HRTIMers are based on DLL technology and are capable of dividing each clock pulse into 32 parts. This provides the microcontroller with… The lowest time resolution. Furthermore, the STM32G484 has two output channels per timer, allowing the construction of an LED driver 10 with a total of twelve channels using the principles described herein, based on a 16-bit counter. If the PWM cycle time is set to... If the longest possible cycle time of using this counter is , then the implementation is... The PWM frequency. Therefore, using the same comparison value is not a possible way to generate high-resolution but low-frequency binary control signals.

[0167] However, the comparison value can be updated before each HRTIM cycle, as explained below.

[0168] If, for example, a 32kHz PWM signal is desired, the required number of HR (high-resolution) clock cycles can be calculated first: Since this value is higher than 65503, it is divided into three parts: The cycle time is set to a lower nearest integer, in this case 56666, resulting in a total repetitive control signal mode cycle time of three 56666 cycles, or a total of 169998 HR cycles and a PWM frequency of approximately 32000.38Hz.

[0169] As an example, assume the desired duty cycle of the corresponding PWM signal is 84.31393%. This then produces... The total comparison value of the HR loop. It should be noted that this is greater than both 56666 and greater than... Therefore, the two first comparison values ​​are each set to 56666, resulting in an output control signal that remains high during the first two runs of the HRTIM cycle. The third comparison value is set to the remaining value. .

[0170] In other words, at HR clock cycle 0, the comparison value is set to 56666, and the counter variable starts from 0. The resulting control signal is at a logic high value (e.g., 0 or corresponding to 3.3V or "all light"). At HR clock cycle 56666, the counter variable reaches the end of the HRTIM cycle and is reset to 0. The comparison value is set to 56666 again. The control signal remains high. At HR clock cycle 113332, the counter variable is reset to 0 again, and the comparison value is set to 3000. The control signal remains high. At HR clock cycle 143332, the counter variable reaches the comparison value of 3000, triggering a switch in the resulting control from high to logic low (e.g., 0 or corresponding to 0V or "no light"). The counter variable continues to increment until it reaches the end of the HRTIM cycle. Then, the output control signal is triggered again, now back to high, and the process restarts from the beginning with the counter variable reset to 0.

[0171] By updating the comparison value after each HRTIM cycle, arbitrary low frequencies can be generated while maintaining the desired high time resolution. Using the cycle time from the example above again, a 250Hz PWM frequency can be generated by combining the number of cycles 128 times: This means The defined comparison values ​​can be used to generate a repeating binary control signal pattern that varies between 250 Hz and 32 kHz.

[0172] Generally, the duration of a repetitive control signal pattern in binary control signals can be longer than the time it takes for the timer / counter variable to complete a full cycle. This is illustrated in the example above, where the counter variable reaches the end of the HRTIM cycle several times before reaching the full length of the control signal pattern. More generally, the repetitive control signal pattern may comprise one or more pulses.

[0173] The LED driver 100 may be arranged to update the set comparison value at least twice during each individual control signal mode, or even at least three times or more, as seen in the example above.

[0174] In principle, timer function 132 may have a clock frequency independent of the clock frequency used by CPU 131 of microcontroller 130. However, in some embodiments, CPU 131 may be arranged to operate at a predetermined clock frequency, and the clock frequency of timer function 132' (timer frequency) may be at least as high as or higher than the predetermined clock frequency of CPU 131.

[0175] As mentioned in the example above, timer function 132 can use DLL 133, which can then be configured to divide the timer function 132 clock (the timer input clock period, where "input" refers to the fact that the clock enters the timer peripheral) into several fractional steps. This is equivalent to increasing the timer resolution of timer function 132 without increasing its clock frequency.

[0176] As also mentioned in the example above, timer function 132 can be cyclic, meaning that in the example above, after reaching the end of each HRTIM cycle, the counter variable repeatedly restarts from its initial value, and as the step passes through the 132-cycle timer function, the counter variable is compared with the current comparison value. In this document, the terms "HRTIM cycle" and "timer cycle" are used interchangeably, but it should be understood that the more general term "timer cycle" also applies to "HRTIM period." In the specific example discussed herein, the HRTIM timer uses a DLL to divide the clock cycle, but in a more general case, other specific mechanisms can be used to define the appropriate timer cycle.

[0177] As described, an exemplary HRTIM cycle of 56,666 clock cycles means that the comparison value needs to be changed after that time has elapsed, which in this example means more than 96,000 times per second. If the task were executed by CPU 131, it would need to dedicate a large portion of its capacity to the task. Instead, the inventors propose performing the comparison value update in the background, such as using a DMA channel.

[0178] That is, in the example above, the HRTIM unit is controlled by a master clock that is reset after 56,666 HR clock cycles. Each reset consumes all currently used comparison values ​​(for each channel) and requires new comparison values. The DMA channel is configured such that each time new values ​​are needed, they are read directly from a circular RAM buffer, provided as RAM region 134 or in RAM region 134, holding 384 values ​​for each channel. Because the buffer is circular, value retrieval restarts from scratch after all 384 values ​​have been copied, and the contents of the RAM buffer do not need to be changed if the incoming dimming value has not changed. This offloads the update work from CPU 131. Once the dimming value is updated, different algorithms are configured to update the data in the RAM buffer as needed. In this example, this then needs to be done at most once the RAM buffer has been reached, i.e., 250 times per second. This update can then be performed by CPU 131 with significantly lower computational cost. Then, the longest possible repeating control signal pattern without changing the values ​​in the RAM buffer is defined by multiplying the number of clock cycles in each HRTIM cycle by the number of comparison values ​​held in the RAM buffer. In various implementations, the RAM buffer may be sized to hold at least 2, such as at least 4, such as at least 8, at least 16, or even at least 32 comparison values ​​at any given time. It should be appreciated that the comparison values ​​can generally be integers.

[0179] Therefore, the comparison values ​​in the RAM buffer can be configured to collectively define a repeating control signal pattern, which in turn forms a binary control signal when the control signal pattern is repeated. When the dimmer setting changes, the RAM buffer needs to be updated using the corresponding updated set of comparison values, which together define a new repeating control signal pattern reflecting the updated dimmer setting. To this end, the CPU 131 can be configured to update the RAM region 134 in response to the updated dimmer setting received by the LED driver 100 using an updated set of comparison values ​​specifically selected to reflect the updated dimmer setting.

[0180] Generally, the comparison value copy function 135 can be configured to update the timer function 132 using an updated set comparison value (such as one such comparison value for each channel supported by the LED driver 100) from the RAM area 134.

[0181] Generally, for a microcontroller unit, there are several different ways for non-CPU components or peripheral components (such as timer function 132) of the microcontroller in question to receive and process updated information. As a first example, a polling interface is used, where the CPU feeds an updated value to the component, and the CPU waits for the value to be consumed before feeding the next value. This is not a highly efficient way to use CPU resources, as the CPU will be locked during the time the component spends consuming the value. As a second example, an interrupt-controlled interface can be used. The value or buffer is then fed to the component, which consumes these values ​​in the background. Once the value is fully consumed, an interrupt is generated, which can be caught by the CPU and used to trigger an update of the buffer data. As a third example, a DMA-controlled interface can be used. The component is then connected to RAM region 134, and the DMA channel is configured to perform application-specific data usage. Once the component needs the next value, it signals this to the DMA channel, which then takes over providing the next value to the component. The value is copied from the RAM buffer by the DMA channel, and an interrupt can be configured to be generated automatically at specific points, such as when the RAM buffer has been fully read or half-read. For example, a first interrupt can be triggered after a first share (such as half) of the RAM buffer comparison value has been consumed, indicating that the first share can be updated, and a second interrupt can be triggered after a second share (such as the other half) of the RAM buffer comparison value has been consumed, indicating that the second share can be updated, and so on. The CPU 131 can then catch the interrupt and thus take action on it, such as by updating the value in the RAM buffer. Since the CPU 131 knows the current position of the process in the read loop, it can update the RAM buffer value that is not currently being read.

[0182] In an exemplary embodiment of the invention, a second or third example may be used to consume and / or update the RAM buffer in RAM region 134. Therefore, the comparison value copying function 135 may be or include a DMA channel, which may be a circular or cyclic DMA channel, such that after it has reached the end of RAM region 134, it resumes copying from the beginning of RAM region 134 again. The DMA channel may be arranged to notify CPU 131, for example, by using an interrupt, when a specific point has been reached (such as half or all of the value has been consumed), thereby providing CPU 131 with information about the current read state of RAM region 134, so that CPU 131 can therefore take action and thus update different portions of RAM region 134.

[0183] Alternatively, the comparison value copying function 135 may be or include an ISR (Interrupt Service Routine) executed in response to an IRQ (Interrupt Request), the ISR being configured to signal to the CPU 131 via an interrupt that a new comparison value from the RAM buffer is needed. When the IRQ signal is emitted, the CPU 131 executes the ISR, which is configured to update the timer function 132 with the new comparison value from the RAM buffer. Thus, the comparison value copying function 135 may be configured to respond to an interrupt generated by an IRQ with or in response to a timer counter value reaching its full cycle value, the response then including copying the new comparison value from the next or subsequent memory location in the RAM region 134. Further alternatively, the IRQ may be used to alert the CPU 131 via an interrupt that the RAM buffer has been fully consumed and that it is time for the CPU to update the comparison value in the RAM buffer.

[0184] As illustrated in the example above, a repetitive control signal pattern can be defined as longer than the HRTIM period; in other words, longer than the time it takes for the timer / counter variable to complete a full cycle. Furthermore, one or more individual pulses forming part of the repetitive control signal pattern can be longer than the HRTIM period, meaning that at least one comparison value will be equal to or greater than the number of HR clock cycles in one HRTIM period. Thus, the repetitive control signal pattern is, in a sense, independent of the HRTIM period, because both individual pulses and the entire control signal pattern can extend beyond the boundaries of a single HRTIM period in time.

[0185] The mechanism used to define the repetitive control signal will be explained and illustrated in more detail below. Generally, the LED driver 100 can be arranged to vary at least one of the following (such as both) in a way that different desired time-averaged currents (or voltages) can be supplied to the driver output 182: firstly, the number of pulses provided across the repetitive control signal pattern, and secondly, the corresponding pulse width of one or more pulses in the set of one or more pulses provided across the repetitive control signal pattern.

[0186] As will be seen below, the LED driver 100 can generally be arranged to gradually increase the time-averaged voltage or current applied to the driver output 182 as the dimming value gradually increases from a low level to a high level by first increasing the pulse width of one or more pulses provided across the repetitive control signal pattern, and then increasing the number of pulses provided across the repetitive control signal pattern to further increase the light intensity. As will be seen, this can then be repeated by increasing the corresponding pulse width of one or more of these added pulses, followed by adding one or more additional pulses, and so on.

[0187] Continuing with the example discussed above, in the sense of PWM, a "pulse period" can be defined as the concatenation of three comparison values, such that 128 individual PWM pulse periods together are used to define the repetitive control signal pattern, and thus the instantaneous shape of the binary control signal. It should be recognized that the repetitive control signal pattern can have any arbitrary shape, for example, such that the individual pulses are distributed in a more irregular manner compared to the string forming such "PWM pulses," but this is the case for this example. It should be noted in particular that one or more portions of the general method described regarding how to introduce and grow individual pulses can be applied to other pulse patterns across the repetitive control signal pattern. When a particular portion of a pulse pattern can be analyzed according to PWM pulses, it is referred to herein as "PWM-like." It should be understood that this terminology is used to facilitate understanding of the invention and is not intended to limit the described embodiments to PWM in its conventional sense (having a single repetitive pulse whose duty cycle varies over time).

[0188] Using this type of three-compare PWM pulse cycle, the threshold pulse width (defined as the width reached before the introduction of a new pulse) is set to a value below 169998, such as 10000. For the lowest optical power, across the repetitive control signal mode, only one PWM pulse will be non-zero, but as the dimming value increases, the pulse width of this single PWM pulse will increase until it reaches the threshold pulse width in question. At this point, a second PWM pulse is introduced, and this second PWM pulse begins to grow with the dimming value increasing. Once the second PWM pulse has reached the threshold pulse width, a third pulse is introduced and begins to grow, and so on. When all 128 pulses have reached the threshold pulse width, the comparison value is adjusted so that the corresponding pulse width of all PWM pulses widens simultaneously. It should be noted that the comparison value can be set such that all PWM pulses are processed in the same way as described, but alternatively, such that different PWM pulses have slightly different properties, such as being associated with different threshold pulse widths before the introduction of the next pulse or at the position where the pulse in question is placed in the repetitive control signal mode. This provides a way to customize the pulse definition algorithm for highly precise dimming, even on hardware platforms with various hardware constraints, such as the minimum or maximum number of clock cycles for comparison values ​​and the like. By controlling the pulse width separately, degradation of bit resolution can be avoided, and therefore the number of effective bits in the control signal can be kept constant. This is given by an expression for the total number of clock cycles in a repetitive control signal pattern, which in this example results in: Ones.

[0189] The time spent at a high level (rather than a low level) relative to the total time of the repetitive control signal pattern defines a PWM-like "duty cycle".

[0190] Generally, the LED driver 100 may be arranged to distribute at least one pulse of a repetitive control signal pattern across a set of time pulse positions that are predetermined and / or uniformly or substantially uniformly distributed (e.g., achieving the precision allowed by the distribution mechanism) across the repetitive control signal pattern. In some embodiments, this distribution may be based on a bit-inverted representation of a number that in turn describes or corresponds to an index of a pulse within a first channel control signal pattern. This will now be illustrated by example.

[0191] Therefore, let The determined (desired) pulse width for the 128 individual PWM pulses included in the repetitive control signal pattern, where Furthermore, let The threshold pulse width before introducing additional PWM pulses. This generates a value based on the frequency before reaching the highest frequency. The sequence, where ,and For having pulse width The number of PWM pulses. If The three comparison values ​​represent the PWM pulse period and the desired average PWM duty cycle (instantaneous intensity). Total HR clock cycle length Then the sequence will satisfy the equation .

[0192] The mapping can now be determined. .if Then all pulses fall after each other, followed by a long pause, resulting in a large 250Hz component, which in turn risks producing flickering and unpleasant sounds. To address this, mapping... It can be designed to maximize the resulting PWM frequency. One way to achieve this is to... Represent it as a binary number, such as a 7-bit number, and then reverse or swap the bits and use a mapping. For the first value, implement the following:

[0193] This is Figure 6 The diagram visually illustrates this, where the height of each bar symbolizes the order in which the corresponding PWM pulses are introduced, with higher bars being introduced earlier (the first bar is for k=0). This method of introducing pulses across repetitive control signal patterns ensures that the pulses are distributed as widely as possible without consuming excessive computational resources. It should be recognized that this distribution mechanism can also be used for less regular pulse distributions, allowing the introduction of pulses to be distributed using bit swapping or similar mapping mechanisms even if the pulses themselves are not perfectly uniformly distributed across repetitive control signal patterns.

[0194] Therefore, for a binary system where numbers are represented by a series of 0s and 1s, the bit reversal process is as described above. However, this reversal concept is not limited to the binary domain. It extends to more complex systems. In a more general case, for example, using a mixed radix of prime numbers. Unlike a binary system where each number has a base of 2, a mixed radix system involves numbers having different bases at different positions. In a mixed radix of prime numbers, prime numbers are ordered from highest to lowest (the lowest prime is the least significant position). The order is then reversed to lowest to highest (the highest prime is the least significant position). Then, given a starting frequency, such as 250Hz, the higher end of the interval (which is 32kHz in the example above) can be arbitrarily chosen, for example, 24kHz. Furthermore, if the ratio between the highest and lowest frequencies is not a power of 2, a mixed radix can be used instead. This can potentially lead to even better dimming performance. As a concrete example, suppose the ratio is... Since 96 can be decomposed into Therefore, a mixed radix with the base {3,2,2,2,2} can be used to represent the number. The bases are sorted from highest to lowest such that the lowest base has the lowest significant value. The representation becomes {s5,s4,s3,s2,s1,s0}. Now, if a number is expressed in this base, the possible values ​​range from 0 to 95 and are described by the sequence (s5,s4,s3,s2,s1,s0). The representation is then reversed to {s0,s1,s2,s3,s4,s5} and observed using the mixed radix {2,2,2,2,3}. Thus, the “bit reversal” process described above is generalized here as “number reversal” and then consists of: reversing the sequence to (s0,s1,s2,s3,s4,s5) and using the base (2,2,2,2,2,3) to compute a new index, where the bases are now sorted from lowest to highest. Based on this, an index is calculated, where the sequence (0,1,2,3,...) will be mapped to (0,48,24,72,12,60,36,84,6,...).

[0195] In this and other embodiments, the duration of the repetitive control signal pattern across different desired times supplied to the driver output 182 can be constant. In other words, in some embodiments, the duration does not change with the dimming value, but rather the concentration of the high-value clock cycle across the control signal pattern is changed by altering the pulse pattern.

[0196] As an alternative, a pulse signal, for example, at a frequency of 250 Hz, can be allowed to increase in pulse width as the dimming value increases until a specific threshold pulse width is reached. Thereafter, as the dimming value increases further, the cycle time can be continuously reduced to introduce pulses more frequently over time. Once the frequency reaches a specific maximum value, such as 32 kHz, the pulse width can be increased again until a 100% duty cycle is achieved. This type of modulation is particularly effective for single-channel LED drivers 100 and can also be implemented using the multi-HTRIM periodic pulse structure generally described above.

[0197] However, for a multi-channel LED driver 100, the fact that the duration of the corresponding repetitive control signal mode is not the same for different channels means that there will be drift in terms of instantaneous current demand across different channels. To illustrate the consequences, it can be assumed that there is a PWM duty cycle. and Two channels, In the latter example, this means the corresponding PWM frequency. and Not equal, that is When the corresponding channel is ahead (high value of the binary control signal), an increased current demand occurs, which in turn causes a temporary drop in voltage on DC bus 121. The magnitude of this drop depends particularly on the size of the LED load 10 and the capacitance on DC bus 121. This means that if the channels are active simultaneously, a specific current demand will occur, but if they are active at different times, different current demands will occur. Interference will then appear at the sum and difference of the corresponding frequencies, meaning that the frequency... It will have a non-zero amplitude. For example, This could lead to low-frequency flickering, which is undesirable. Therefore, using, for example, according to the example above... Configuration compared to The configuration is generally more suitable for multi-channel LED driver 100.

[0198] Generally, the relative current demand across different channels can be arranged to be stable rather than drifting. To this end, a corresponding periodic signal (such as a repetitive control channel mode) can be generated on each channel, where the periodicity is the same for all channels, and where the periodicity can be as long as the longest PWM period of the PWM pulses that together form the periodic signal. In other words, the signal can consist of a repetitive sequence, and this sequence is generated as a series of individual class PWM pulses. Because the sequence is periodic, the voltage on the DC bus 121 will also be periodic with the same period, resulting in a constant light contribution from the pulse as long as the pulse does not move within the sequence. If the pulse moves to a different position, there is a risk of notches appearing in the dimming curve. Once a pulse has been introduced at a specific point in the sequence as the dimming value increases, the LED driver 100 can be configured to hold the introduced pulse at that position in the sequence as long as the dimming value increases.

[0199] Generally, in addition to the first channel driver output 182', the LED driver 100 may also include a second channel driver output 182'' and any third channel driver output 182''' or other driver outputs. Similar to the first channel driver output 182', the LED driver 100 may be arranged to supply current to the second channel driver output 182'' and any other driver output 182''' according to a corresponding binary control signal (such as a second channel binary control signal). The LED driver 100 may also correspondingly include a second channel timer function 132'' and any third channel timer function 132''' or other timer functions, configured to update the corresponding second channel timer counter variable, third channel timer counter variable, or other channel timer counter variable at the second channel timer frequency, third channel timer frequency, or other channel timer frequency. The second-channel timer function 132'', the third-channel timer function 132'', or an additional channel timer function can be configured to, for each update, compare the second-channel timer counter variable, the third-channel timer counter variable, or the additional channel timer counter variable with a set second-channel comparison value, the third-channel comparison value, or the additional channel comparison value. The second-channel timer function 132'', the third-channel timer function 132'', or the additional channel timer function can be configured to, when the timer counter variable in question has reached or exceeded the set comparison value in question, switch the binary control signal in question between corresponding sets of binary states, such as switching between a third binary state and different fourth binary states for the second-channel binary control signal. Therefore, this functionality corresponds to the general description above regarding binary control signals, and specifically regarding the first-channel binary control signal.

[0200] Then, the duration of the repeated second-channel control signal pattern, third-channel control signal pattern, and / or other-channel control signal pattern can be longer than the time taken for the second-channel timer / counter variable, third-channel timer / counter variable, and / or other-channel timer / counter variable to complete a full cycle. As with the first-channel control signal pattern, the second-channel control signal pattern, third-channel control signal pattern, or other-channel control signal pattern can each consist of one or more pulses.

[0201] Furthermore, the LED driver 100 can then be arranged to update the set second-channel comparison value, third-channel comparison value, and / or additional-channel comparison value at least twice during each individual second-channel control signal mode, third-channel control signal mode, and / or additional-channel control signal mode.

[0202] In such cases, the duration of the first channel control signal mode can be an integer multiple of the duration of the second, third, and / or other channel control signal modes, such as being the same as the duration of the second, third, and / or other channel control signal modes. This will result in no intra-channel drift. It should be noted that if the duration of the first channel control signal mode is a shorter duration, the same effect is achieved if the first channel control signal mode is an even fraction of the duration of the second, third, and / or other channel control signal modes.

[0203] To achieve the most stable DC voltage possible, it is desirable for the LED driver 100 to draw the smoothest possible current. If we assume, for example, a 10% corresponding PWM duty cycle across all channels, and that the activated binary control signal sequence activates all channels simultaneously in the sequence, then the current demand will be unnecessarily high, followed by an unnecessary long pause before the next current demand occurs.

[0204] Conversely, if an offset is introduced at the channel level, and this offset is different for each channel, then current demands will not occur simultaneously on each channel, resulting in a smoother overall current demand. As the frequency content of the current demand increases, the capacitor on the LED driver 100 does not need to supply current to the circuit for a long period of time, thus generating lower capacitance demand and therefore reducing manufacturing costs.

[0205] Generally, the time offset of the first channel binary control signal relative to the second channel binary control signal, the third channel binary control signal, and / or the other channel binary control signal may be less than the corresponding offset value of the mode duration of the first channel control signal.

[0206] For an LED driver 100 having three or more electrical driver outputs 182', 182'', 182''', each binary control signal in the corresponding binary control signals of different channels may be offset by a value less than the duration of the first channel control signal mode, and the corresponding time offsets of at least three control signal modes may be distributed across a predetermined and distributed set of time offsets across the first channel control signal mode. The distribution of the at least three control signal modes across the duration of the repeating control signal mode may be uniform or substantially uniform (taking into account, for example, the available granularity on the hardware used).

[0207] For example, bit inversion, similar to or corresponding to the manner described above for the distribution of pulses across repetitive control signal patterns, can be used to offset the corresponding binary control signal distribution for different channels. If It is a passage i The offset, then for the twelve-channel LED driver 100, the offset value is... This results in a cross-channel distribution that generally provides smoother current demand and therefore smoother voltage and lower capacitance demand. For multi-channel LED strips, this produces even less flicker because the instantaneous light intensity decreases simultaneously with increasing frequency.

[0208] Generally, the offset distribution across different channels can be determined by bit reversing the binary representation of a number, which in turn describes or corresponds to the channel indices of the electrical driver outputs 182', 182'', 182'''. This distribution can span the duration of the control signal mode used on the first channel.

[0209] The following is an example of a process of the type generally described above, which is designed to use a variable frequency to drive a constant voltage LED load 10 and avoid discontinuities, flickering and interference patterns.

[0210] It is the maximum number of pulses in a periodic signal (repetitive control signal mode). All pulses have the same threshold pulse width, measured in clock cycles. The threshold pulse width is equal to the total number of clock cycles in the periodic signal. of One-third.

[0211] It stores the counter value for each corresponding pulse width. One slot.

[0212] Expected strength It is a range The value inside.

[0213] The current task is to fill all slots, so that The slot is available. Figure 7 Fill in the flowchart shown.

[0214] The remaining task is to decide the order in which the new slots should be filled. If they are filled first... After that And so on, resulting in a signal with large low-frequency components. To minimize flicker, after the first slot has been filled, the next slot to be filled should be at the furthest distance from the first slot. This will ensure that the low-frequency content of the signal is always minimized. Therefore, if the first slot to be filled is... Then the next slot should be Then, it will continue in binary mode, inserting new pulses at the distance furthest from existing pulses. If For a certain integer It can be written as This type of index can be improved by using a linearly increasing index. of The calculation is performed by bit reversal.

[0215] Given the constraints of the microcontroller used in the inventor's setup, the high-resolution timer is limited to 16 bits. Therefore, this is merely an example. At the highest equivalent clock frequency of... In this case, when the timer period is at its maximum value, that is, equal to At that time, the lowest possible PWM frequency is achieved. Dividing the equivalent clock frequency by this number yields approximately This is too high for the current task. By setting each slot... by themselves It consists of three counter cycles in a clock cycle, with the maximum frequency changed to By order The lowest frequency is then... .

[0216] For any other set of minimum and maximum frequencies, the following procedure can be used: ● The given value is the clock frequency of the timer peripheral device. ,For example

[0217] ● Define the ratio between the highest PWM frequency and the lowest PWM frequency. ,Right now ,For example .

[0218] ● Define the target frequency of the lowest frequency content of the signal. ,For example .

[0219] ● In the formula Find the lowest integer in the middle. This causes the result to be lower than the maximum counter value. This gives... and Both, for example ,and .

[0220] ● will Truncated to integers, for example .

[0221] ● Example values ​​are given .

[0222] Defines the number of different pulse widths to be generated before the sequence restarts. Each pulse has... Clock cycles (e.g.) The maximum length of (one clock pulse). Defined as a pulse The pulse width, where The range is from arrive The index. Total open time is And the total time is ,For example One clock cycle. If the controller input signal is equal to the range... The average light intensity generated internally can then be defined as follows: .

[0223] If all If the values ​​are the same, the generated control signal will have a frequency. A pure PWM signal, and the duty cycle will be If all Value divided by an index Everything else will be zero, so the generated control signal will have a frequency. And it has a duty cycle A pure PWM signal. In general, the frequency content of the signal is mixed, depending on which bins are filled. However, the number of pulses per second is... Furthermore, the minimum frequency component of the generated signal will not be lower than [the minimum frequency component]. .

[0224] The algorithm can be summarized as follows: ● Configure the timer peripheral device to enable A periodic signal of pulses can be used to... and To generate. Being chosen to make for a certain , .

[0225] ● Determine the minimum pulse width to be used The lower this value, the faster the frequency will increase with increasing intensity. Become .

[0226] ● will Defined as at least using pulse width The number of filled slots. pass It is used to calculate, and then truncated to its integer part.

[0227] ● Define the array to be filled sequentially .

[0228] ● If , use value Fill The front of the middle One slot, and set the next slot as the residual. All other slots are set to zero.

[0229] ● Otherwise, if To fill all slots as evenly as possible: First, set all values ​​to... All values ​​are truncated to their integer parts. The residuals are then calculated. This is the range. The value within. (The last part is incomplete and likely refers to a previous value.) Each slot is increased by one to minimize the residual.

[0230] ● By setting To add random indexes, among which yes Bit reversal function.

[0231] Second aspect of the invention Turning now specifically to the second aspect of the invention, as understood above, the binary control signal is a digital control signal or an analog representation of a digital control signal, which is then used to turn the LED load 10 on and off. However, the drive circuit 140, the drive switch 180, and the LED load 10 themselves all possess dynamic electrical properties that will alter the resulting signal actually delivered to the LED load 10. For example, if there is a duty cycle... If the PWM signal is used as the control signal for driving the LED load 10, then the actual average current through the LED load 10 will be different. Multiply ,in When The current passing through LED load 10 at that time.

[0232] As an example, in the case where a raw or compensated binary control signal is provided to a MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) and this MOSFET is then used to output drive current to the LED load 10, the gate voltage... and its internal resistance The relationship between them is complex and depends on several factors. When the drive circuit 140 takes a non-zero time... Come to During charging, this complex relationship will cause Deterministic (but often unknown) changes during the charging and discharging process. Over time The length of time becomes so long that if a visible amount of light is emitted before charging is complete, this light will be emitted before the current through the LED load 10 reaches its maximum value. Similarly, the amount of light emitted is determined by a complex and often unknown relationship, determined by both the transistor itself and the driving circuit 140. It should be noted that the driving circuit 140 can be described below in conjunction with the third aspect, but any driving circuit 140 (such as using resistors in simple series connection) will introduce complex dynamic behavior to some extent. For example, resistors will be associated with specific precision and manufacturing variability, such that not every resistor of the same type will have exactly the same properties. Even if a resistor of higher quality is chosen, any transistor used in the driving circuit 140 will have dynamic properties, and so on.

[0233] The amount of light emitted from a light source such as an LED is directly proportional to the current. Therefore, by integrating the current over time, the total amount of emitted light can be measured. The intensity of light can be defined as... , in It is the period of the signal (the period of the repetitive control signal pattern). It is through the changing current of LED load 10, and It is the current achieved at 100% duty cycle (with the binary control signal at a constant high value).

[0234] Now, duty cycle It can be defined as the amount of time the currently used repetitive control signal pattern is at a high value divided by the period of the currently used repetitive control signal pattern. Ideally, However, due to the complex relationships described above, this equation will not hold. But measurable... and The relationship between them, such as by connecting a typical LED load 10 or another electrical load to the LED driver 100, and then measuring... .

[0235] Below, an example will be presented to illustrate this. Since changing the frequency of the output signal of the LED driver 100 typically adds additional complexity, the frequency is initially fixed at 250Hz for all intensities. Measurements are then taken... And in Figure 8 The results are presented in the image. The dashed lines represent... .

[0236] The inventor proposes to find the measured function The inverse function, i.e. , and then use To find the specific duty cycle of the binary control signal to be used given a desired light intensity.

[0237] To better understand how each type of PWM pulse is changed by the LED driver 100, let This is a system consisting of a drive circuit 140, a drive switch 180, and a typical LED load 10. Furthermore, let... For the binary control signals entering the system, and to make for Any single type of PWM pulse. Let To describe How by system The modified equation. In this equation, and It is the pulse width measured in seconds, where The effective pulse width can be calculated by integrating the pulse over time and dividing by its maximum value (thus producing a pulse with a constant maximum value). It is any set of external variables that may affect the system; these external variables can be ambient temperature, etc. Average value, maximum current The driving power supply voltage, etc., depend on... Scope and A pulse that can represent voltage and / or current.

[0238] Figure 4d A non-square pulse is illustrated as a further example of the concept of "effective pulse width" in this sense. It is a typical waveform of a pulsed current through an LED load. Since the light intensity is proportional to the current, the time integral of the waveform measures the amount of light emitted from a single pulse. If this time integral is set to be equal to the time integral of an ideal rectangular pulse with a maximum current Imax, then the rectangular pulse will represent the same light intensity and effective pulse width. Therefore, in illustrating the pulse... Figure 4d middle, It is the maximum (current) level of the pulse that usually appears on the upper platform of the pulse. It is the level as a function of time (in this case, current). It is the effective pulse width of the pulse. It should be recognized that... .

[0239] Then, relative strength It can be calculated as Divide by period time And then produce a value between 0 and 1.

[0240] For the system considered in this analysis, This will be large enough to generate an output signal. A continuous function whose derivatives are strictly positive and continuous for all values. Therefore, by the inverse function theorem, the inverse function... It exists. If If it is the required pulse, then Available Calculate given the known conditions, and if fed to Then it will produce .therefore, Together Together they will produce the identity function .

[0241] By targeting the full spectrum of relevant pulse widths Measurement of different values It can be applied to any set of external variables. Data collection points .

[0242] Parametric models can now be found. , making ,in yes The set of internal variables. For an ideal system, the input and output PWM-like pulse widths will be the same, i.e. Therefore, analyzing the difference between the ideal and the actual situation is more meaningful: , in This is another parameter model to be found. Since the human eye is highly sensitive at low light levels, the model error should be small for lower intensities. Furthermore, for low intensities, the PWM-like pulse width is shorter, and... The dynamics will affect This has a significant impact. Therefore, expanding the input values... In order to obtain More details about the smaller values ​​could be reasonable. This can be achieved through... Applying nonlinear functions (such as , or This is done using the logarithm of the intensity. In the following example, the logarithm of the intensity is used: , in It is another parametric model.

[0243] Since the behavior of the drive circuit 140 and the drive switch 180 typically involves an exponential relationship, curve fitting can be simplified by taking the logarithm of both sides and finding a model of the residual shape to remove this behavior: , in It is another parametric model.

[0244] This is compared with the collected data. Put them together, and you get

[0245] , And in finding Next, the parametric model It can be calculated as follows .

[0246] Then, what remains is the usage relationship. To find the parametric model For this measurement example, Figure 9 The relationship is illustrated in the diagram.

[0247] Given input variables There exists for the purpose of... Several methods can be used for modeling, and both parametric and nonparametric methods can be employed. For example, spline interpolation, polynomial approximation, lookup tables, kernel functions, and parametric models can all be used, each with its own advantages and disadvantages. One possible approach is to... Model it as two distinct low-degree polynomials, one of which has a maximum value. On the left, and on the right, a polynomial of this kind. Then, the two polynomials can be fused together, for example, using a sigmoid function.

[0248] Figure 10 The principle is illustrated in the diagram, where the black triangles represent measured data. The solid curves in the top chart that do not follow the black triangles represent the two polynomials and the sigmoid function. The sigmoid function itself is shown in the bottom chart. The solid curves that follow the black triangles represent the resulting model. .

[0249] The polynomials matching the curve to its left and right of its maximum value have degrees of 3 and 2, respectively. (Sigmoid function) By passing order and At the center point It consists of two exponential functions connected consecutively: ,in ,and .

[0250] and exist The continuity at a given point gives the parameters. and :

[0251] .

[0252] This gives when The asymmetric sigmoid function at time t.

[0253] Given two polynomials and a sigmoid function, then the total... Approximate parameters ,Right now ,in These are the parameters of the two polynomials used.

[0254] These unknown parameters can be found using any suitable type of optimization algorithm. One possible approach is to minimize the system. The previous squared error of the PWM pulse width, for example, using The difference: , or the logarithm of those values. However, due to the quantity express It has been approved. The subsequent actual PWM pulse width, therefore a better approach would be based on We calculate the error, since that's the quantity of interest. For any calculated value... You can use the sampled dataset The closest value in To search This makes the difference computable. And use it as an error.

[0255] Generally, the method for configuring the LED driver 100 for subsequent dimming of the LED load 10 according to the second aspect of the present invention is as follows: Figure 11 The example illustrates, and relates to, the use of at least one transistor of LED driver 100 to modulate the variable current supplied by LED driver 100 to LED load 10 based on the binary control signal described above.

[0256] As described above, the binary control signal includes or is composed of repetitive control signal patterns, which in turn have an instantaneous time average at each time point. This instantaneous time average is calculated over time over the control signal pattern used at the time in question. This instantaneous time average is represented as the "duty cycle" of the aforementioned binary control signal.

[0257] In the first method step 1101, the method begins.

[0258] In subsequent step 1102, LED driver 100 is connected to LED load 10, or alternatively to a different load 20. See also Figure 12 It shows an LED driver 100 driving an LED load 10, which in turn radiates light, which is detected by a light sensor 11; and the LED driver 100 optionally drives different electrical loads 20, the current through which is then measured by a current measuring sensor 21. Of course, the current through the electrical load 20 or the light emitted by the LED load 20 can be measured in many different direct or indirect ways, and Figure 12 This is just an example. For instance, a current clamp can be used to measure the current. Alternatively, instead of measuring the light emitted from the LED load 10, the current passing through the LED load 10 can be measured.

[0259] In subsequent step 1103, the LED driver 100 is operated to provide a variable current to the LED load 10 or another load 20 in the general manner described herein using several different instantaneous time-averaged values ​​of the binary control signal. It should be noted that since a single operation of the control signal mode may be sufficient, in some embodiments, it is recommended to use several repetitive control signal modes one after another, as is the case during continuous operation of the LED driver 100 at dimming values ​​corresponding to the control signal modes discussed. In other embodiments, multiple measurements can be performed continuously for different controlled duty cycle values. For example, slow brightening or dimming can be used by gradually changing the duty cycle of the repetitive control signal mode after each or several control signal modes have been run, and measuring current or light as described, thereby providing multiple measurements involving only the execution of dimming actions and the taking of measurements.

[0260] In subsequent step 1104, for each of several different instantaneous time averages, a corresponding measurement is obtained, such as by reading from sensor 11 or 21, of a quantity such as emitted light or current flowing through load 20. This quantity can then be, or explicitly related to, at least one of the following: first, the average current flowing through LED 10 or load 20, and second, the average light intensity emitted by LED 10. Thus, a first relationship is obtained, a measurement of the absolute or relative current / light intensity change according to the instantaneous time average. It should be understood that steps 1103 and 1104 can be performed iteratively, wherein a first instantaneous time average is used and a first measurement is performed, followed by a second instantaneous time average and a second measurement, and so on, for example, using a dimming control scheme as mentioned above. It should be understood that step 1104, obtaining the measured values ​​and the first relationship, can also be considered as a step of automatically determining the first relationship based on the obtained measured values.

[0261] The first relationship can be a value-for-value relationship between each absolute or relative current / light intensity and each corresponding instantaneous time average. This contrasts with the second relationship described below, which defines the nature of the absolute or relative current / light intensity as a function of the instantaneous time average. It should be understood that the configuration information generated by the currently described method may simply include a set of absolute or relative current / light intensities and a table or similar data structure containing the corresponding instantaneous time averages, used to determine a specific instantaneous time average (and the corresponding determined repetitive control signal pattern) based on a simple lookup of the desired absolute or relative current / light intensity. Alternatively, the configuration information may include information about the second relationship, thereby providing the possibility of calculating a specific instantaneous time average based on the configuration information given the desired absolute or relative current / light intensity.

[0262] Generally, the absolute value of current or light intensity can be measured, or a relative value of the same metric. For relative values, the metric can be correlated with the measured current or light intensity at the maximum dimming value, and then any measured value at a lower dimming value can be proportional to the value at the maximum dimming value. To illustrate this, consider a PWM with a fixed frequency. and duty cycle (as defined above) binary control signals, where Indicate the i-th measurement sample. Then, the duty cycle... Generate the maximum current through load 20 or the maximum light intensity through LED load 10. First, by setting... To measure the maximum current or light intensity, and the resulting current is expressed as... (And in the case of measuring light intensity instead, corresponding to the obtained light intensity). Now, for any given duty cycle It can measure current. (or correspondingly, light intensity), and by utilizing Normalizing this quantity allows us to calculate the equivalent duty cycle. In the following calculations used to determine the parametric model, only relative values ​​are used, not absolute values. More generally, relative current and / or light intensity can be relative to the maximum measured current or light intensity, respectively.

[0263] For an ideal drive circuit, the equivalent PWM duty cycle is equal to the generated PWM duty cycle, i.e. However, due to the nonlinear effects in the driving circuit, therefore .

[0264] By Set as and make exist and Changes between them allow for the collection of data points. The set is used to construct a model corresponding to the lowest class PWM frequency. Similarly, by... Set as It can collect data points corresponding to the model with the highest class of PWM frequency.

[0265] These models are respectively called and These models take a middle ground. and The number between them is used as the corresponding The input, and the output is between and The correspondence between Compensation value:

[0266]

[0267] A compensation term is applied to each class of PWM pulse. Given a normalized light intensity... and fixed-frequency PWM This generates a control signal pattern with the required pulse width. The model is applied to each corresponding class of PWM pulse in the control signal pattern. If such a pulse is called... Then the pulse passes through the equation Related to relative light intensity. Similarly, the compensation term is related to the PWM-like duty cycle as follows: utilizing... Scaling: Putting these together, we get... .therefore: ;

[0268] To mitigate the simplifications made, it is now available. and To construct the final compensation term.

[0269] Multiple PWM-like pulses with varying repetitive control signal patterns cause the PWM-like frequency to change from... Smoothly change to In this case, It can be defined in which the PWM frequency ranges from Change to Within the range of Change to The scaling factor. A good weighting factor is found empirically. It is used for the final compensation item: .

[0270] In subsequent step 1105, the configuration information is determined based on the first relationship, specifying the mapping of each current / light intensity value among several different absolute or relative such current / light intensity values ​​to its corresponding instantaneous time average. Therefore, the mapping can be direct, via a lookup table or the like, or it can be indirect, such as via a parametric model. In other words, the configuration information determined based on the first relationship can also be configuration information in the form of information about the second relationship. Specifically, the configuration information can be, for example, parameters of a parametric model of the general type described above.

[0271] In some embodiments, the method may further include: in step 1106, identifying a parametric model of the second relationship. This parametric model may then include one or more at least piecewise continuous functions determined by a set of parameters. Therefore, the functions may be analytic functions, rather than lookup tables or the like. The parametric model is generally used to determine the mapping of each of several different absolute or relative current / light intensity values ​​to the corresponding instantaneous time average. It should be understood that the identification of the parametric model may include using a predetermined parametric model or selecting one from several possible parametric models depending on the type of LED driver 100 used, or the like. The identification of the parametric model itself does not necessarily require determining or calculating the values ​​of the parameters included or used in the identified parametric model.

[0272] That is, in subsequent step 1107, corresponding values ​​of the parameters of the identified parametric model can be determined, providing a fit of at least one function to the second relation, such as the best fit determined according to a suitable optimization algorithm. Any suitable method can be used to determine the fit, wherein the second relation is evaluated for each measurement and its corresponding instantaneous time average. The configuration information will then include the parameter values.

[0273] In some implementation schemes, the second relationship includes business ,in It is a function. It is the instantaneous time average (of the repetitive control signal pattern), and It is an absolute or relative current / light intensity. For example, Producer In another example, Producer . These two formulas have proven useful in practical implementations.

[0274] Therefore, the function It can be approximated as a measurement point This provides an approximate relationship. Solve this problem. ,get .in other words, It is used during the calculation of configuration information, and is used once the configuration information is used to operate the LED driver 100. and .

[0275] Generally, for any bijective function whose pulse width is continuous for all classes of PWM pulses that are strictly greater than zero, ... It can be used in the following ways: when the parameter values ​​of the parametric model are found, use the amount And when applying a parametric model with its parameter values, use and its inverse function Both.

[0276] The parametric model may include one or more polynomials and / or other primitive functions. Generally, the one or more functions may include at least one first-order or higher-order polynomial, such as a second-order or higher-order polynomial, such as a third-order or higher-order polynomial. In some embodiments, the one or more functions may include two first-order or higher-order polynomials and / or other primitive functions, which are then interconnected by multiplication using a sigmoid function. Any two polynomials may be chosen to approximate the overall shape of the second relation on either side of a local minimum or maximum value of the second relation by selecting the parameter values ​​of the polynomials.

[0277] One or more functions can generally be based on the function To define, where about The absolute value of the first derivative (in other words) ) is decreasing, such as across The entire range of its use decreases monotonically. It is an absolute or relative current / light intensity. For example, It can be a logarithmic or square root function. From a general perspective, the parametric model M can include differentiable functions. The differentiable function takes system variables. As input and having the following properties: for in All within the defined scope and , And among them Indicates ratio Short pulse width, and yes about The derivative. In some implementations, to Input parameters Only through To use it, so that for a certain parameter model , Can be written in form .

[0278] Generally, the sigmoid function can be derived from two functions. and The two functions are defined such that they produce continuous functions at the expected local minimum or maximum of the absolute or relative current / light intensity in the second relationship, and that the derivative of the continuous function does not change its sign.

[0279] This sigmoid function can be of the general type described above, using one or more exponential functions. However, other types of sigmoid functions can also be used.

[0280] For example, by Replace with However, using the same parametric model construction method outlined above in other aspects, a simpler sigmoid function can be created as follows:

[0281]

[0282] make exist Use and exist When to use:

[0283] Example configuration parameter value , , ,and . Here, the degree of the sigmoid function is the polynomial degree, and it should be chosen such that the degree is higher than that of the polynomial it needs to suppress. For example, in a parametric model using two polynomials of degree two and three respectively, the sigmoid function should have a degree of at least four. It is the center point of the sigmoid function, and it is defined as the point where the two functions constituting the sigmoid function are joined together. and Define functions separately in The variability on the left and right sides.

[0284] Using the continuity of functions and their properties at points The derivative in and The calculation can be performed as follows: ,and ,and ,in , ,and .

[0285] The example parameter values ​​provided above produce the following: Figure 13 The sigmoid function is shown. Generally, it is defined by... and , and The above formula can be used to calculate and obtain a continuous and asymmetric sigmoid function, where, in special cases... The following implements a symmetric S-shaped function.

[0286] The form of a function expression can generally be written as: ,in It is a polynomial. The shape can generally be written as ,in It is another polynomial.

[0287] In addition to direct or indirect information about absolute or relative current / light intensity and corresponding instantaneous time averages, the configuration information may also include additional information. This information may be determined automatically by the LED driver 100 or manually in conjunction with the configuration of the LED driver 100, and may include, for example, information about: the number of pulses in the control signal mode; the temperature of the LED driver 100 when the measurement is obtained; and / or the maximum current supplied to the LED driver 100. In some embodiments, measurements of absolute or relative current / light intensity for various instantaneous time averages may be performed for several sets of such measurements, and each set of information regarding the absolute or relative current / light intensity for various instantaneous time averages may be stored together with or in association with this additional configuration information as part of the configuration information. In other words, configuration information about a first relationship and / or a second relationship may be stored together with corresponding information about the number of pulses in the control signal mode, temperature, and / or the maximum current supplied. Once this information is stored, it can be used for interpolation when driving the LED driver 100 between measured states spanned by these parameters (i.e., the number of pulses in the control signal mode; temperature; and / or the maximum current supplied), as illustrated below. Specifically, the current operating state defined by these parameters can be used to determine the parameter values ​​of the parameter model as corresponding interpolations between the given parameter values ​​of the known operating states.

[0288] In some implementations, the configuration information may therefore include: at least two different pulse counts for the control signal mode; at least two different temperatures of the LED driver 100 when the measurement is obtained; and / or at least two different maximum allowable currents to be supplied to the LED driver 100.

[0289] In subsequent step 1108, configuration information is stored for future use. This storage may be in RAM region 134, elsewhere within the LED driver 100, or in external storage units. For example, the configuration information may be stored in non-volatile memory such as ROM or flash memory for use across continuous power cycles of the LED driver 100. Once the configuration information is stored, it can be used for the operation of the LED driver 100, whereby the configuration information is used to implement binary control signals suitable for providing compensation for the desired light intensity of the LED driver 100, as will be explained further below.

[0290] In subsequent step 1109, the method ends.

[0291] Another load 20 may correspond in electrical properties, such as impedance, to a specific LED load 10 to be used subsequently with the LED driver 100. Similarly, the LED load 10 used for configuration does not need to be the same LED load 10 driven later by the LED driver 100 during operation, but it is sufficient that the load used for configuration is similar to the LED load 10 to be used during lighting operation or that the configuration information is known in a clear way regarding any differences in electrical properties between the load used for configuration and the LED load 10 driven later by the LED driver 100.

[0292] Figure 14 Another method for operating the LED driver 100 is illustrated.

[0293] In the first method step 1401, the method begins.

[0294] In subsequent step 1402, use Figure 11 The illustrated method configures the LED driver 100 as described above. As recalled, at the end of this configuration, there is stored configuration information that can be used to generate compensated binary control signals given the desired light intensity of the driven LED load 10.

[0295] In a subsequent step 1403, which may be performed at a later time, the LED driver 100 is then operated by determining a binary control signal corresponding to a specific desired current or light intensity, the determination of which is performed using stored configuration information. The binary control signal is then used to provide a variable current to the driver output 182 of the LED driver 100. More specifically, the variable current is provided based on the determined binary control signal in any manner generally described herein. Preferably, the same LED driver 100 used in the calibration method is also used in the operation method. However, it should be recognized that, as described herein, calibration may take several different aspects into account, and specifically, calibration information may include information that is not directly and specifically designed to compensate for different electrical properties in the components of the LED driver 100. Therefore, the calibration method for a standard LED driver 100 may also be performed at least partially, and the resulting calibration information may then be used when operating different LED drivers 100 of the same standard type (such as several different LED drivers 100 of the same type).

[0296] For example, as described above in conjunction with the first aspect, determining the binary control signal may include adjusting (compensating) the pulse width of one or more pulses within a repeating binary signal pattern and / or adjusting the number of pulses within the repeating binary signal pattern; and the pulses within the repeating binary signal pattern may be adjusted to be distributed across a predetermined set of available pulse positions across the repeating binary signal pattern (such as a predetermined set of more or less uniformly distributed available pulse positions). The pulse compensator 137 may be arranged to perform the adjustment / compensation based on previously determined and stored configuration information (such as a parameter model of one of the various types described herein).

[0297] As discussed above in conjunction with the first aspect, in embodiments of the present invention, the frequency of the binary control signal can be constant in the sense that the repetitive control signal pattern has a constant set period length. In one example provided, the frequency... At low light levels, the repetitive control signal pattern can contain only a single pulse, meaning the binary control signal is a 250Hz PWM signal. Utilizing the cycle time... By scaling the desired current or light intensity, the determined parametric model can be used to calculate the compensated pulse width. The premise is the desired pulse width Since only a single pulse exists in the control signal mode, a compensated duty cycle can also be used. 1. Use the desired light intensity I to express it explicitly:

[0298]

[0299]

[0300] .

[0301] Given this relation, This can now be applied to every type of PWM pulse, even if the frequency changes when a new pulse is added to the repetitive control signal pattern. However, this presents a problem when the drive circuit 140 depends on the instantaneous time-averaged value: if it is assumed that there is only one pulse during the 250Hz cycle, then... It is applied only once, and since the signal is the same as when calibration was performed, a perfect inverse model is achieved. Alternatively, if we assume that the average light intensity or current is the same, but the pulse is divided into two separate and uniformly distributed pulses within a 250Hz cycle, effectively forming a 500Hz PWM signal, then... This requires applying the control signal twice. This results in slightly different signals entering the drive circuit 140, which in turn causes the external variable to deviate from the value used during calibration (the instantaneous average value is different). If more pulses are added to the repetitive control signal pattern, the deviation will increase even at a constant total duty cycle in the control signal pattern. A maximum frequency of 32kHz is reached with a maximum pulse count of 128 in a 250Hz cycle.

[0302] To solve this problem, two parametric models can be constructed: one is... It is optimized for 250Hz binary control signals, and another one is... It is optimized for 32kHz binary control signals. For intermediate frequencies, Then the interpolation between these two extreme point models can be used to calculate. Then, let For values ​​between 0 and 1, such that The value is 0 until the first pulse has reached the threshold pulse width (as described in conjunction with the example in the first aspect). When the second pulse is added and begins to grow, And thereafter it increases linearly until it reaches the point of maximum pulse count, at which point, .Then, This can be determined based on the following empirical expansion of the equation that provides a good fit under frequency changes occurring in this example and similar examples: .

[0303] Finally, given A model has been developed that takes into account component variations, the electrical properties of the drive circuit 140 and the drive switch 180, and, in particular, the behavior caused by variations in the transistors used and the signal frequency.

[0304] Generally, in some implementations, determining the binary control signal in step 1403 may include: determining the number of pulses within a control signal pattern. The binary control signal can then be selected to fit the desired average current or light intensity based on an interpolation between corresponding first and second expressions of the binary control signal. The first expression may be determined based on configuration parameters of the first number of pulses within the control signal pattern, and the second expression may be determined based on configuration parameters of the second number of pulses within the control signal pattern. Interpolation may be performed for the current number of pulses within the control signal pattern, which is between the first and second number of pulses. Several different interpolations may therefore be used when dimming is performed between a first dimming value and different second dimming values. For example, the interpolation from the first dimming value to the second dimming value may be continuous.

[0305] More generally, the first and second expressions can be determined based on the corresponding pulse widths within the control signal pattern.

[0306] More generally, interpolation can be performed on the first and second expressions using the square root of the normalized value of the current number of pulses as a weighting factor.

[0307] Another example considers that the drive circuit 140 itself may depend on binary control signals. The fact of the average value. This fact will be discussed in detail below in conjunction with the third aspect. More specifically, in some implementation schemes, The average value can be changed The rise and fall times. To account for this, two models are used for two extreme cases: one for when the rise and fall times are at their slowest rate of change, and one for when they are at their highest rate of change. These are given separately for each model. and These two models will then use weight functions. They merged together.

[0308] , in All external sets , and Include variables ,and and Indicates that they are created respectively and The conditions used at the time. Generally, the methods described above can be used for any other variables that may affect the characteristics of the power transistor 181 or the drive circuit 140. For example, the external set may contain information about the system temperature or the load size of the LED load 10. Suppose, for example, to model temperature dependence, two models can be constructed. and ,in and This corresponds to the different temperatures used during calibration. When these models are applied, the current temperature is then measured (and included in the variables under discussion). (in the middle), and then another weight function can be used. To perform similar interpolation.

[0309] Furthermore, in a more general sense, the model can be written as a weighted sum of its sub-models, i.e. ,in It is a weight function that depends on external variables.

[0310] In subsequent step 1404, the method ends.

[0311] It should be recognized that the binary control signal selected using configuration information to match the desired light intensity or current as described above can be considered to constitute a “compensated” binary control signal in the following sense: the binary control signal actually used (such as being fed to the drive circuit 140) is not a naive binary control signal designed using a specific control signal mode duty cycle corresponding to the desired light intensity or current, but is compensated relative to such a naive binary control signal to take into account the specific electrical properties of the drive circuit 140, drive switch 180, etc., so as to actually generate the desired waveform corresponding to the naive binary control signal at the output 182 of the LED driver 100 and generate the desired waveform in the LED load 10.

[0312] It should be recognized that, depending on the specific needs of the current implementation plan, various combinations of methods and approaches can be used to determine the configuration information presented and illustrated herein.

[0313] Specifically, given the measured current or light intensity and the set duty cycle (binary control signal), there are many ways to identify and determine the parameterized model to realize the inverse model. The expected accuracy. Inverse model The specific implementation will largely depend on how the drive circuit 140 and drive switch 180 are implemented. For example, the parameterization model may differ depending on whether the drive circuit 140 uses a variable resistor driven by a BJT transistor or a DAC (digital-to-analog converter) (see below for the third aspect). Another example of a parameterization model already used by the inventors is where... It is calculated as the following parameterized model. , in Other parameters are constants in the model adapted after measuring the data, and these parameters are determined using some appropriate and conventional optimization method.

[0314] Third aspect of the invention Turning specifically to the third aspect, the LED driver 100 includes a driving circuit 140 and a driving switch 180 as described above. The driving switch 180 is arranged to provide a variable current or voltage to the LED load 10 as a switching electrical signal. The microcontroller 130 is arranged to provide a control signal to the driving circuit 140, such as a compensated binary control signal as described in the second aspect, to provide a variable current or voltage to the LED load 10, such as via the driving switch 180.

[0315] In practice, the third aspect presupposes a binary control signal source arranged to provide a binary control signal as a square wave, the binary control signal having a control signal pattern that repeats over time to form the binary control signal as described above. Also as described, the control signal pattern has an instantaneous time average at each time point, which is or corresponds to the time average of the control signal pattern used by the signal at the time in question. The binary control signal source may be the microcontroller 130 itself, but it should be recognized that the third aspect of the invention also, in principle, works with various internal or external binary control signal sources, such as a separate piece of hardware arranged to provide binary control signals to the drive circuitry 140.

[0316] The LED driver 100 (and specifically, one or more drive switches 180) also includes one or more electrical driver outputs 182 as described above, each of which is arranged to provide drive current to the LED load 10 based on a binary control signal to drive one or more LEDs of the LED load 10.

[0317] An exemplary LED driver 100 is shown connected to the LED load 10 for driving the LED load 10. Figure 15 As illustrated, the LED driver 100 includes a power transistor 181. In Figure 15In the example shown, power transistor 181 actually constitutes switch 180. Power transistor 181 is arranged to power LED load 10 via driver output 182 and has a power transistor control terminal 181a and a pair of different terminals 181b, 181c. Control terminal 181a is arranged to receive a voltage or current signal arranged to modulate power transistor 181 from driver circuit 140. One of the other terminals 181b, 181c (in...) Figure 15 In the example shown, terminal 181c) is connected to driver output 182. Power transistor 181 can be a FET (field-effect transistor), such as a MOSFET.

[0318] The drive circuit 140 is connected to the power transistor control terminal 181a.

[0319] Figure 15 The exemplary LED driver in this example is a single-channel LED driver 100. It should be understood that in the case of a multi-channel LED driver, the switch 180 and other related components will be replicated and one will be provided for each channel.

[0320] exist Figure 15 In this configuration, the positive terminal of the LED load 10 is connected to the feed voltage, and the negative terminal is connected to the drain terminal of an N-channel MOSFET transistor (power transistor 181). The source terminal of transistor 181 is connected to ground, and its gate is connected to the drive circuit 140. To handle the parasitic inductance from the long cable to the LED load 10 that causes slow current decay, a Schottky diode 190 is provided, arranged to provide a return path for the current. It should be understood that any such Schottky diode 190 may also form part of the LED driver 100. It should be understood that other types of diodes may be used as alternatives to the Schottky diode 190, such as standard silicon diodes.

[0321] It should be recognized that switch 180 can generally be provided in other ways. For example, a GaN transistor can be used instead of a MOSFET. A P-channel MOSFET can be used, and the transistor's connection point can be directed to VDC.

[0322] As described above, the binary control signal can be generated directly by the binary control signal generator 136 of the microcontroller 130, or provided from an external source.

[0323] From a general perspective, MOSFET transistors are well-suited for applications where they are either fully on or fully off. However, power losses occur at the transition points between these on and off states. Therefore, shortening this transition time reduces these losses. This can be achieved by designing the drive circuit 140 to be fast, but as discussed above, this leads to other issues related to non-uniform light at low light levels.

[0324] A common way to limit the speed of a drive circuit is to connect a resistor in series with the gate drive logic. However, this method will limit the maximum PWM-like frequency of the LED driver 100 because power loss is proportional to the operating frequency.

[0325] However, the inventors have recognized that the binary control signal can be designed such that the PWM-like frequency is low at low dimming levels and high at high dimming levels. In fact, this is the case when operating according to the principles described herein using many reasonable dimming methods, such as the use of the PWM-like pulses described above in conjunction with the first aspect. Furthermore, the driver circuit 140 can be designed to be slow at relatively low dimming levels and fast at relatively high dimming levels. This achieves both uniform illumination at low dimming levels and high efficiency at high dimming levels, despite the higher PWM-like frequency. This will now be described and illustrated.

[0326] The drive circuit 140 is arranged to construct the electrical control signal required to operate the switch 180, which in turn supplies power to the LED load 10. Given the binary control signal used to drive the LED load 10 to turn on and off, the LED driver 100 should generate a drive signal with the correct voltage / current level, and the drive circuit 140 should achieve this end goal while providing a control signal with appropriate rise and fall times for the present specific embodiment. Specifically, when processing the switch 180 based on an n-channel FET transistor, a suitable positive voltage level should be generated at the gate terminal 181a of the power transistor 181 relative to its source terminal 181b.

[0327] Generally, for FET transistors, the main contributors to power loss are switching losses and conduction losses. To minimize losses without using unnecessary expensive transistors, the two loss types are typically balanced so that they contribute to the loss at roughly the same order of magnitude.

[0328] However, for a constant voltage LED load 10, the current through the LED load 10 is strongly correlated with the control signal entering the switch 180, which means that additional factors must be considered. First, the generated pulse will have a certain physical length within the electrical trace across the LED load 10. As discussed above, for a 5m physical length LED strip, and assuming that the speed of light inside copper is approximately half the speed of light in a vacuum, the signal travels approximately [missing information - likely a timeframe or length]. Upon reaching the other end of the LED strip, the wave will be reflected and travel in the opposite direction. If the wave length is on the same order of magnitude as the LED strip, a visible interference pattern can be emitted. This is usually not a problem for conventional LED drivers because at their lowest possible dimming levels (such as...), Under these conditions, the pulse is long enough to make the resulting interference pattern invisible. Substituting this into mathematical rigor, if using... PWM frequency and At the dimming level, the generated pulse is , equivalent to length The pulse.

[0329] Remember that the generated pulse has a rectangular shape, compared to the fundamental frequency (e.g.) High frequencies are contained in the signal (harmonics). To avoid any interference patterns, frequencies up to at least 10 times the fundamental frequency should be considered. For conventional LED drivers, this is still not a problem because the equivalent pulse length is 65.2m, which is still much longer than that of a normal LED strip. However, for dimming to lower levels, such as... The equivalent length is further reduced to This is comparable in length to a conventional LED light strip, and this increases the risk of interference patterns becoming visible.

[0330] Therefore, it is assumed that the dimming level must be lower than Even at low PWM frequencies, rectangular pulses cannot be used to dim to darkness with high light quality.

[0331] When high-frequency currents flow through a conductor, additional problems arise in the form of crosstalk. If the high-frequency content of the signal has a corresponding wavelength comparable to the length of the conductor, the conductor will begin to act as an antenna, potentially transferring energy to nearby conductors. This can be problematic for multi-channel LED strips, as it may generate light from channels that should be off, potentially causing issues such as color errors.

[0332] Therefore, the frequency of the drive signal to the LED load 10 should be relatively low at relatively low dimming levels, but relatively high at relatively high dimming levels.

[0333] One way to reduce the frequency of the drive signal is to reduce the frequency of the control signal going to switch 180. For MOSFETs, this can be done by slowing down the charging and discharging of the gate capacitance, such as by increasing the series resistance between drive circuit 140 and the gate 181a of transistor 181. This results in longer rise and fall times.

[0334] However, doing so increases switching losses, which limits the feasible PWM frequency for the binary control signal, potentially leading to flickering. Therefore, to increase the frequency as intensity increases, while remaining efficient and capable of dimming all the way to complete darkness, slow rise and fall times can be used for low light intensities, while fast rise and fall times can be used for higher light intensities.

[0335] As discussed herein, the LED driver 100 can be designed such that the average light intensity of the connected LED load 10 is positively correlated with the number of pulses generated per unit time. Therefore, the average intensity can be used as a control signal to the drive circuit.

[0336] Figure 16 A block diagram of an exemplary drive circuit 140 is shown. Binary control signals are delivered by an (internal or external) binary control signal generator 136 to a signal averaging section 170, which in turn provides a time-averaged signal value representing a positive and / or negative value of a general signal level currently in use, which may be associated with a currently used repetitive control signal pattern. Generally, the drive circuit 140 is arranged to receive or generate one or more level signals (such as the time-averaged signal value), each of which is provided as an incrementing or decrementing function of an increasing instantaneous time-averaged value.

[0337] The corresponding time-averaged signal value is then used to control the first variable current control resistor and the second variable current control resistor 161 (the first current throttling portion and the second current throttling portion), each of which is then connected in series between the first corresponding first switching transistor 151 or the second switching transistor 152 (the switching transistors 151 and 152 together form the switch 150 of the drive circuit 140) and the second switch 180 (power transistor 181). More specifically, the control terminal 181a of the power transistor 181 is connected in series in the manner discussed herein. The first switching transistor 151 is connected to a high voltage 154 (such as the drive power supply 123 of the power supply 120) and is arranged to supply current to the switch 180 (control terminal 181a) via the first variable resistor 161 according to the provided binary control signal from the binary control signal generator 136. Correspondingly, the second switching transistor 152 is connected to a low voltage 153 (ground in this example) and is arranged to provide current from the switch 180 (control terminal 181a) via a second variable resistor 161 according to the provided binary control signal from the binary control signal generator 136. Figure 16The exemplary circuit shown is a push-pull circuit employing two such current throttling sections 161, one of which is arranged to throttle the current to the gate terminal 181a, and the other is arranged to throttle the current from the gate terminal 181a. In other configurations, only one current throttling section 161 may be used.

[0338] Generally, the drive circuit 140 may include a drive circuit switch 150, which is arranged to selectively provide electrical coupling between the control terminal 181a of the power transistor 181 and either the low voltage 153 or the high voltage 154 based on a momentary low or high value of a binary control signal. It should be understood that this coupling can be regulated by a variable resistor 161, but may also be regulated by additional electrical components. The coupling can also be through and via a direct connection to the variable resistor 161.

[0339] Further generally, the drive circuit 140 may include one or more such variable resistors 161, which are arranged to provide a corresponding variable resistance along the electrical coupling. The variable resistance of each of the variable resistors 161 is arranged to vary as a function of a level signal (such as an incrementing or decrementing function, depending on how the level signal is defined and generated), resulting in an increase in the instantaneous time-averaged value leading to an increase in the instantaneous current flowing to or out of the control terminal 181a of the power transistor 181. For Figure 16 In the upper part of the circuit, the inverted binary control signal relative to the high voltage Vdrive is averaged and used as a level signal, so that a larger negative signal is converted into a larger current flowing to the control terminal 181a. In the lower part, the non-inverted binary control signal relative to the low voltage GND is averaged and used as a level signal, so that a larger positive signal is converted into a larger current flowing out of the control terminal 181a.

[0340] To explain this setup more closely, we can first consider... Figure 17 The circuit shown is a conventional push-pull drive circuit. Figure 17 The push-pull drive circuit is based on a pair of MOSFET transistors, one n-channel and one p-channel. The circuit uses separate resistors to individually tune the rise and fall times.

[0341] exist Figure 17In the diagram shown, ~PWM is the inverting input control signal. When the binary control signal (represented as "PWM" in these examples) is low, ~PWM is high, turning on Q6A and turning off Q6B. This discharges the gate of the power transistor connected to GATE_DRIVE through R34. When PWM is high, ~PWM is low, turning off Q6A and turning on Q6B. This charges the gate of the power transistor through another resistor, R33.

[0342] Now consider Figure 18 The more complex circuit shown incorporates resistors 163 and 165, wherein the combination of the first current-controlling transistor 162 and resistor 163 constitutes a first current-throttling mechanism 161, and the second current-controlling transistor 164, together with resistor 165, constitutes a second current-throttling mechanism 161. The two current-throttling mechanisms 161 depend on the average value of the binary control signal PWM in a manner corresponding to what has been discussed above.

[0343] exist Figure 18 In the middle, the first current-controlled PNP transistor 162 Q2B and the second current-controlled NPN transistor Q2A 164 are... Figure 17 The push-pull circuit shown is connected in series. In the example shown, current-controlled transistors 162 and 164 are BJT transistors, which include: corresponding base terminals 162a and 164a; corresponding emitter terminals 162b and 164b; and corresponding collector terminals 162c and 164c. In other embodiments, current-controlled transistors 162 and 164 may be MOSFET transistors having corresponding gate, source, and drain terminals.

[0344] Generally, the variable resistor section 160 may include one or more current throttling sections 161, in which case the current control transistors 162, 164, combined with resistors 163, 165, are arranged to throttle the current that can be electrically coupled, and thus limit the rise time and / or fall time of the electrical signal supplied to the control terminal 181a of the power transistor 181, thereby controlling the rise time and / or fall time of the current through the drive output section 182 when the switch 150 switches the electrical coupling between a low voltage 153 and a high voltage 154. As mentioned, the variable resistor section 160 may be arranged to vary the resistance as a function of a level signal (such as a decreasing function). This function may be a monotonic function.

[0345] Further generally, the variable resistor section 160 may include one or more current-controlled transistors (in this case, also 162 and 164), which are configured such that current flowing through the electrically coupled transistors flows across the current-controlled source / emitter terminals 162b, 164b and the current-controlled drain / collector terminals 162c, 164c. Level signals may be coupled to the current-controlled gate / base terminals 162a, 164a, such that an increase in the instantaneous time average results in a higher instantaneous current across the current-controlled source / emitter terminals 162b, 164b and the current-controlled drain / collector terminals 162c, 164c. Thus, for higher dimming values, a faster drive circuit 140 is achieved for both the rising and falling edges of the current waveform output to the LED load 10, and vice versa.

[0346] Then, each of the current-controlled transistors 162 and 164 can be connected in series along the electrical coupling to a corresponding resistor 163 and 165 in order to limit the current to a desired value.

[0347] Furthermore, switch 150 may include one or more switching transistors 151, 152, each switching transistor having a corresponding gate / base terminal 151a, 152a, a corresponding source / emitter terminal 151b, 152b, and a corresponding drain / collector terminal 151c. Generally, the gate / base terminals 151a, 152a of each such switching transistor 151, 152 may be connected to a positive or negative binary control signal in the form of a voltage or current that varies according to the value of a binary control signal. Figure 18 In the illustrated case, two switching transistors 151 and 152 are formed together to provide a push-pull circuit, and the two switching transistors 151 and 152 are arranged between two different voltage potentials in the form of a low voltage 153 and a high voltage 154. Furthermore, the control terminal 181a of the power transistor 181 is coupled to a point between the two switching transistors 151 and 152. In the case of using BJT transistors, they can be connected as a so-called "common-emitter" (emitter terminal interconnect) or "common-collector" (collector terminal interconnect). In the case of using MOSFET transistors, the drain terminals can be interconnected.

[0348] exist Figure 18 In the original circuit, switching transistors 151 and 152 are each FETs. However, by modifying the circuit layout, BJT transistors can be used instead as switching transistors 151 and 152.

[0349] Similarly, Figure 18For specific examples, the variable resistor section 160 may include both a first current control transistor 162 and a second current control transistor 164. The first current control transistor is configured such that current electrically coupled between the first switching transistor 151 and the control terminal 181a of the power transistor 181 passes through the first current control transistor 162. The second current control transistor is configured such that current electrically coupled between the second switching transistor 152 and the control terminal 181a of the power transistor passes through the second current control transistor 164. In other words, the electrical coupling discussed herein may consist of two separate connection lines, both terminating in providing or absorbing current to the control terminal 181a, but each connection line is operable only during the rise or fall of the pulse.

[0350] Regarding the signal averaging section 170, it may be arranged to provide a level signal as or as a response to an instantaneous time average or its reciprocal, as discussed above. Specifically, the signal averaging section 170 may include a low-pass filter or digital-to-analog converter arranged to provide the level signal as a smoothed or time-averaged positive or negative version of a binary control signal. The level signal itself may be a current or voltage corresponding to an instantaneous time average in any of the aforementioned manner.

[0351] One objective of the drive circuit 140 is to minimize switching losses in the power transistor 181 and to provide slow switching characteristics for low light intensity. This power transistor is driven by a control signal designed to use a low class PWM frequency for low light intensity and a higher class PWM frequency for higher light intensity. In other words, the rise time and / or fall time at low light intensity should or could be so long that an unrealistic risk of heat generation would occur if the intensity and class PWM frequency increase. The time average of the binary control signal can be used as a level signal to control the variable resistor when the time average of the binary control signal is positively correlated with the number of pulses per second (class PWM frequency). On the other hand, this also means that the inverse signal of the time-averaged binary control signal is inversely correlated with the number of pulses per second and can also be used to control the variable resistor. Therefore, the level signal can be a time-averaged binary control mode or related to a time-averaged binary control mode, or a negative time-averaged binary control mode or related to a negative time-averaged binary control mode. Regardless of the approach, the objective remains to reduce the variable resistor as light intensity increases, and the variable resistor will be configured to change with the level signal to achieve this objective.

[0352] In the examples provided in this article, the average value of the binary control signal is used to control the NPN BJT transistor, while the average value of the inverse binary control signal is used to control the PNP BJT transistor. Similarly, if the current-limiting transistor is a MOSFET, the non-inverting signal will typically be used for the N-channel MOSFET, while the inverting signal will be used for the P-channel MOSFET.

[0353] Therefore, assuming that the PWM-like frequency increases as the time average of the binary control signal increases is equivalent to assuming that the PWM-like frequency increases as the light intensity emitted from LED load 10 increases. A level signal that increases or decreases with light intensity can now be constructed, depending on whether it is based on a non-inverting or inverting binary control signal. This level signal is then used to control a variable resistor such that the variable resistor decreases as light intensity increases.

[0354] As noted above, the level signal can be provided as a function of the instantaneous time average. However, it should be noted that in some implementations, the level signal is not directly calculated as a function of the instantaneous time average, but may instead be calculated to correlate with the instantaneous time average in some other way. For example, the level signal may be calculated based on control input 110 and (in a manner corresponding to the above) calculated or determined as an increasing or decreasing function of such control signal, which is then correlated with the instantaneous time average.

[0355] Return to Figure 18 The added components in the circuit shown can be observed first. Figure 17 and Figure 18 The differences between them, and should be noted, Figure 17 Resistors R11 and R12, as well as transistors Q2A and Q2B, are not installed, but resistors R33 and R34 are. In this case, a conventional push-pull drive circuit is implemented based on two MOSFET transistors Q6A and Q6B. The inverted signal ~PWM is used to control the gates of Q6A and Q6B, and the high voltage value of ~PWM is the same as the voltage Vdrive. R33 controls how fast the gate connected to GATE_DRIVE charges, and R34 controls how fast it discharges. This provides the same rise and fall times for the current to the LED load 10, regardless of the duty cycle of the repetitive control signal mode PWM (as defined above).

[0356] Now, we can examine the case where R11, R12, Q2A, and Q2B are used instead of R33 and R34, such as... Figure 18 As in the case described above. The bias currents to BJT transistors Q2A and Q2B will now determine the effective resistance to GATE_DRIVE. Looking at the lower part of the circuit, the PWM signal first passes through a low-pass filter. and These components are used to create a constant voltage level proportional to the duty cycle of the PWM signal. The low-pass filtered signal is now connected to a resistor divider. and This combination allows for the creation of a suitable, selected mapping between the duty cycle and the bias voltage. .Then, This is used to create the bias current to the BJT transistor Q2A, which is then amplified by the transistor Q2A's amplification factor. Magnification. Therefore, through... The current is determined by the duty cycle, as long as the transistor is operating in its active region. Once it is saturated, the current flows to ground instead. limit.

[0357] To analyze biased networks, one can observe Figure 19 The circuit in the middle. When the PWM signal is fed to... When the signal is generated, The signal will and The values ​​change linearly. When the duty cycle is 0%, And when the duty cycle is 100%, Therefore, the circuit can be as follows: Figure 19 The drawing was redrawn as shown, making it exportable. and The parsing expression.

[0358] By using the principles of resistance laws (Ohm's law and Kirchhoff's current law), and Can be given Calculate under the following circumstances:

[0359] because and Together they are used as a low-pass RC filter, so values ​​are assigned to them so that they together achieve the desired suppression of changing signals, thereby creating a DC voltage. The filter's cutoff frequency... (That is, the frequency at which the filter begins to attenuate the signal) thus becomes a design parameter and is defined as follows and

[0360] .

[0361] Now all that remains is to operate at a given voltage. The resistor R16 is determined under the following conditions. When Q1A is closed, the circuit is based on... Figure 20 To simplify. Due to the base-emitter voltage in an NPN transistor. It can be modeled as a forward-biased diode, therefore the Shockley diode equation can also be used here, as a function of voltage. current This will follow a curve similar to the one described above for an LED connected in series with a resistor. Therefore, the Shockley diode equation is generated: and

[0362] in It is thermal voltage. It is an ideal factor, and It is the scaling factor current.

[0363] Now, It can be described as Functions: .

[0364] silicon-based diodes and Example value and And in Figure 21 and Figure 22 Used in the chart shown.

[0365] For a given voltage Now, the desired effective resistance The expression makes the resistance ratio calculable. Since the switching losses in power transistor 181 are more or less proportional to the rise and fall times of the electrical signal delivered to the control terminal 181a of power transistor 181, twice the rise and fall times will result in half the power loss. Therefore, if Then, in relation to Having the same value but with Compared to a circuit with a resistor ratio (corresponding to the case using a constant, non-variable resistor), for a binary control signal with a sufficiently high average value, only half the power loss will be generated. Therefore, the resistor ratio... Measure the potential efficiency gain compared to conventional circuits.

[0366] Firstly, by giving the above values ​​and ( Figure 21 In the case of drawing as The function This leads to a simplified analysis. In this graph, it can be seen that the voltage... It is fairly stable in the linear region, but it is temperature-dependent. It also depends on Choice ( Figure 22 However, it remains quite stable for various resistance values. Therefore, one could try... Treating it as constant, and further assuming that in Figure 18 The intermediate voltage GATE_DRIVE equals This allows the effective resistance to be calculated at the moment when the MOSFET gate just begins to charge. Then, the resistance is connected in series with the transistor... The equivalent resistance can be defined as Ohm's law, along with the amplification factor of transistors. Together they can be used to calculate while the transistor is operating in its active region. : .

[0367] Now, It can be determined as and Parallel connection. Let and .but:

[0368]

[0369]

[0370]

[0371] ;and .

[0372] Due to higher voltage To produce a lower resistance value Therefore, obtain .

[0373] Therefore, in order to maximize the resistance ratio The second term in the second factor needs to be minimized, which is achieved by allowing Approaching infinity makes the term become zero. In other words, when targeting... When optimizing for high values, R14 is not needed, and the expression is then simplified to .

[0374] distance The closer, It can grow larger, but on the other hand, it also becomes more temperature sensitive because... It depends on the temperature. Therefore, It needs to be set as low as possible, but not lower. This can be achieved by targeting... Different values ​​were experimentally determined by applying temperature variations to the circuit. Allowed The lower the value, It can be larger, and The higher the value, the lower the power loss obtained from high-modulation brightness. According to experiments conducted by the inventors, approximately 10% has been achieved. Value. If the MOSFET gate is modeled as a capacitor. The capacitor requires a resistor. In voltage During charging, the voltage across the gate... Become .

[0375] It is the circuit's time constant, which indicates how quickly the capacitor is charged. Therefore, 10 The value means that the time constant is reduced by a factor of 10 at high dimming values, and these power losses can be significantly reduced due to power losses that occur during the transition when the transistor is turned on and off.

[0376] about Figure 18 The upper part of the circuit used for gate drive logic is a direct mirror image of the lower part, and the same reasoning can be used to derive the same findings.

[0377] With Figure 18 Another possible solution for features similar to those provided in the example is to separate the gate driver circuitry from the variable resistor. By doing so, a conventional gate driver IC can be used in series with a variable resistor, such as... Figure 23 (Block diagram) and Figure 24 (Detailed circuit example) is shown.

[0378] exist Figure 24 In the diagram, the left-hand portion of the circuit represents a universal gate driver, which is arranged to deliver voltage Vdrive when the binary control signal PWM is high and 0V when PWM is low. Figure 18 Compared to the circuit shown, Figure 24 The circuits in them are different, but they perform essentially the same things, as will be explained now.

[0379] Vout is the voltage from the general-purpose gate driver. Figure 24 In the example shown, the general-purpose gate driver includes push-pull transistor pairs 151 and 153 (note the difference between them). Figure 18(The correspondence between 151 and 153 in the text), but in other implementations, it can be replaced by circuits based on: for example, BJT transistor pairs in a common-collector or common-emitter configuration, optocouplers, or gate driver integrated circuits.

[0380] By looking in the direction of the arrow pointing to the BJT transistors, you can see the current path through transistors Q8A and / or Q8B. When Vout is high, the current takes the path through Q8B, and when Vout is low, the current takes the path through Q8A.

[0381] VHighBias is the potential at the point between resistors R10 and R32, and VLowBias is the potential at the point between R37 and R40. Figure 18 The comparisons were made, and the corresponding VHighBias and VLowBias potentials were also marked.

[0382] When Vout is high, the current through the base terminal of Q8B is determined by the difference between Vout (which can be, for example, 5V) and VHighBias. The higher this difference, the higher the current going to GATE_DRIVE. A higher difference is achieved by allowing VHighBias to be lower. (The last sentence appears to be incomplete and possibly refers to a different topic.) Figure 24 As can be seen, VHighBias is connected to ~PWM via a low-pass filter. Therefore, the higher the average value of the binary control signal PWM, the lower VHighBias will be, and the more current will go to GATE_DRIVE.

[0383] When Vout is low, the current through the base of Q8A is determined by the difference between Vout (which can be, for example, 0V) and VLowBias. The larger this difference, the higher the current from GATE_DRIVE. A higher difference is achieved by allowing VLowBias to be higher. As shown in the figure, VLowBias is connected to the binary control signal PWM via a low-pass filter, so the higher the average value of the PWM signal, the higher VLowBias will be, and the higher the current from GATE_DRIVE will be.

[0384] For example in comparison Figure 18 and Figure 24 As can be seen from the circuit diagram, various parts such as 150, 160, and 170 are arranged in different ways, but each has a corresponding function.

[0385] As seen in the example above, the drive circuit 140 may include a MOSFET-based push-pull drive circuit. However, as... Figure 25 As seen in the example, the corresponding push-pull circuit can be designed alternatively based on BJTs. Figure 25In this context, V3 and V4 are bias voltages generated by low-pass filtering the binary control signal PWM. When V3 = 0V and V4 = 5V, the fastest possible control is achieved, such as... Figure 26a As shown. When V3=5V and V4=0V, the slowest possible control is achieved, as follows. Figure 26b As shown.

[0386] If V3 and V4 are generated by low-pass filtering the inverted binary control signal PWM and the non-inverted binary control signal PWM, respectively, then the switching speed can be implemented as a function of the duty cycle (as defined above). Therefore, the following is achieved: Figure 18 and Figure 23 The corresponding functionality in [the context].

[0387] Other possible methods for generating V3 and V4 include: the microcontroller 130 includes a DAC (digital-to-analog converter) whose output is used to achieve better control of the control signals. A dedicated 2-channel DAC can also be used for each output channel.

[0388] Figure 27 An example is shown of a method for driving an LED load 10 using an LED driver 100.

[0389] In the first step 2701, the method begins.

[0390] In subsequent step 2702, the driver output 182' of the LED driver 100 is connected to the LED load 10.

[0391] In subsequent step 2703, power is supplied to the LED driver 100 via the DC bus 121, and the binary control source 136 provides a binary control signal, which in turn causes the power transistor 181 to provide drive current to the driver output 182' as described above, thereby resulting in a variable current being supplied to the LED load 10, as generally described herein.

[0392] In subsequent step 2704, the binary control signal can be adjusted to reflect different dimming values, thereby modifying the instantaneous time average. This step may require mapping the dimming state adjustment to a corresponding adjustment of the binary control signal.

[0393] In subsequent step 2705, step 2703 can be repeated, but with updated binary control signals.

[0394] The method ends in subsequent step 2706.

[0395] The preferred embodiments have been described above. However, it will be apparent to those skilled in the art that many modifications can be made to the disclosed embodiments without departing from the basic spirit of the invention.

[0396] For example, in addition to the components described and illustrated herein, the LED driver 100 may also include various types of additional components. It should also be appreciated that, for clarity, the examples provided herein are kept simple. In various embodiments of the invention, different voltages and currents may be used and applied to the LED load 10.

[0397] Generally, the various principles and exemplary methods described herein, as well as all aspects of the LED driver 100, can be freely combined for compatibility. Furthermore, one or more aspects of both or all of the first, second, and third aspects described herein can be freely combined in the same LED driver or method for operating such an LED driver.

[0398] Therefore, the present invention is not limited to the described embodiments, but may vary within the scope of the appended claims.

Claims

1. An LED driver (100), the LED driver comprising: The first channel electric driver output (182'), wherein the LED driver (100) is arranged to provide current to the first channel electric driver output (182') according to a first channel binary control signal; and The first channel timer function (132') is configured to update the first channel timer counter variable at a first channel timer frequency. The first channel timer function (132') is configured to compare the first channel timer counter variable with a set first channel comparison value for each update. Furthermore, the first channel timer function (132') is configured to switch the first channel binary control signal between a first binary state and different second binary states when the first channel timer counter variable has reached or exceeded the set first channel comparison value. The duration of the repeated first channel control signal pattern of the first channel binary control signal is longer than the time taken for the first channel timer / counter variable to complete a full cycle. The first channel control signal pattern includes one or more pulses, and wherein... The LED driver (100) is configured to update the set first channel comparison value at least twice during each individual first channel control signal mode.

2. The LED driver (100) according to claim 1, wherein The frequency of the first channel timer is at least as high as or higher than the predetermined clock frequency of the CPU central processing unit included in the LED driver (100).

3. The LED driver (100) according to claim 1 or 2, wherein The first channel timer function (132') uses a DLL delay phase-locked loop function (133'), which is configured to divide the timer input clock period into several fractional steps.

4. The LED driver (100) according to any of the preceding claims, wherein The first channel timer function (132') is cyclic.

5. The LED driver (100) according to any of the preceding claims, wherein the LED driver further comprises: RAM random access memory area (134). and The comparison value copy function (135) is operable to update the first channel timer function (132') using the updated set first channel comparison value from the RAM area (134).

6. The LED driver (100) according to claim 5, wherein The comparison value copy function (135) is a DMA direct memory access channel or an ISR interrupt service routine executed in response to an IRQ interrupt request.

7. The LED driver (100) according to claim 5 or 6, wherein The comparison value copy function (135) is configured to respond to an interrupt that occurs when the first channel timer counter value reaches its full cycle value, the response including copying a new first channel comparison value from the next or subsequent memory location of the RAM region (134).

8. The LED driver (100) according to claim 7, wherein The DMA channel is configured as a ring, such that after it has reached the end of the RAM region (134), it resumes copying from the beginning of the RAM region (134).

9. The LED driver (100) according to claim 8, wherein The size of the RAM region (134) is at least 4 comparison values, such as at least 8 comparison values, such as at least 16 comparison values.

10. The LED driver (100) according to any one of claims 5 to 9, wherein The CPU (131) is configured to update the RAM region (134) in response to an updated dimmer setting received by the LED driver (100) using an updated set of first channel comparison values ​​specifically selected to reflect the updated dimmer setting.

11. The LED driver (100) according to any of the preceding claims, wherein The individual pulse of the binary control signal is longer than the time it takes for the first channel timer counter variable to complete a full cycle.

12. The LED driver (100) according to any of the preceding claims, wherein The LED driver (100) is arranged to vary the number of pulses provided across the first channel control signal mode in order to achieve different desired instantaneous time averages of the first channel control signal mode.

13. The LED driver (100) according to any of the preceding claims, wherein The LED driver (100) is arranged to vary the pulse width of one or more pulses in a set of one or more pulses provided across the first channel control signal mode in order to achieve different desired time-averaged currents provided to the output section (182') of the first channel electric driver.

14. The LED driver (100) according to claims 12 and 13, wherein The LED driver (100) is arranged to gradually increase the time-averaged current applied to the output of the first channel electric driver (182') by first increasing the pulse width of one or more pulses provided across the first channel control signal mode and then increasing the number of pulses provided across the first channel control signal mode.

15. The LED driver (100) according to any one of claims 12 to 14, wherein The LED driver (100) is arranged to distribute at least one pulse across the first channel control signal pattern onto a set of predetermined and distributed time pulse positions across the first channel control signal pattern.

16. The LED driver (100) according to claim 15, wherein The distribution is based on the bit reversal of the binary representation of a number, which in turn describes or corresponds to the index of the pulse within the first channel control signal pattern.

17. The LED driver (100) according to any of the preceding claims, wherein The duration of the first channel control signal mode is constant across different desired times provided to the first channel electric driver output (182') with an average current that is constant.

18. The LED driver (100) according to any of the preceding claims, wherein In addition to the first channel electrical driver output (182'), the LED driver (100) also includes: The second channel electric driver output (182'') is configured to supply current to the second channel driver output (182'') according to the second channel binary control signal; and The second channel timer function (132'') is configured to update the second channel timer counter variable at the second channel timer frequency. The second channel timer function (132'') is configured to compare the second channel timer counter variable with a set second channel comparison value for each update. Furthermore, the second channel timer function (132'') is configured to switch the second channel binary control signal between a third binary state and different fourth binary states if the second channel timer counter variable has reached or exceeded the set second channel comparison value. The duration of the repeated second-channel control signal pattern is longer than the time taken for the second-channel timer / counter variable to complete a full cycle. The second-channel control signal pattern includes one or more pulses, and wherein... The LED driver (100) is configured to update the set second channel comparison value at least twice during each individual second channel control signal mode.

19. The LED driver (100) according to claim 18, wherein The duration of the first channel control signal mode is an integer multiple of the duration of the second channel control signal mode, such as being the same as the duration of the second channel control signal mode, and wherein... The time offset of the first channel binary control signal relative to the second channel binary control signal is less than the offset value of the first channel control signal mode.

20. The LED driver (100) according to claim 19, wherein The LED driver (100) includes three or more electrical driver outputs (182', 182'', 182'''), and the LED driver (100) is arranged to apply a corresponding current to each of these driver outputs (182', 182'', 182''') according to a corresponding binary control signal, wherein Each of the binary control signals has a corresponding repeating control signal pattern, the duration of which is an integer multiple or fraction of the duration of each repeating control signal pattern of the other binary control signals, such as being the same as the duration of each repeating control signal pattern of the other binary control signals. The LED driver (100) is arranged to time-offset each of the binary control signals below the duration of the first channel control signal pattern, and wherein... The corresponding time offsets of at least three control signal modes are distributed across a set of predetermined and distributed time offsets of the control signal modes across the first channel.

21. The LED driver (100) according to claim 20, wherein The corresponding time offsets of the at least three control signal modes are different.

22. The LED driver (100) according to any of the preceding claims, wherein The clock frequency of the first channel timer function (132') is at least 50MHz, such as at least 100MHz, such as at least 150MHz.

23. The LED driver (100) according to any of the preceding claims, wherein The LED driver (100) is a constant voltage LED driver.

24. The LED driver (100) according to any of the preceding claims, wherein The first channel binary control signal is configured to switch between the first binary state and the second binary state at a frequency of at least 20 kHz, preferably at least 25 kHz, and then switch back to the first binary state.

25. The LED driver (100) according to claims 23 and 24, further comprising: DC bus (121), which includes a ceramic capacitor (122).

Images