Hybrid low noise power supply system with input voltage adaptive switching and state monitoring

Abstract

This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring, belonging to the field of power supply technology. It includes: a power input module, a transformer power supply module, a rectification and linear voltage regulation module, a microcontroller main control unit, a power soft-start module, a power status monitoring module, a ripple sampling hardware branch, and a communication interface module. It solves the defects of traditional power supplies, such as easy misoperation during manual voltage adjustment, power-on surge, large ripple noise, and lack of intelligent monitoring and protection. It is suitable for the low-noise and stable power supply requirements of high-fidelity audio and precision instruments.

Description

Technical Field

[0001] This invention relates to the field of power supply technology, and in particular to a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. Background Technology

[0002] High-performance audio equipment and precision electronic instruments have stringent requirements for power supply quality, which must have core characteristics such as low noise, low ripple, stable output voltage, and safe power-on start-up protection.

[0003] Existing traditional power supply systems suffer from several technical defects in practical applications: First, input voltage adaptation relies on manual operation. Traditional linear power supplies require manual methods such as toggling switch positions and changing jumpers to switch the series or parallel connection of the transformer primary winding to adapt to AC110V and AC220V mains voltages. Manual operation is prone to misjudgment and incorrect switching, which can easily damage the power supply and downstream load equipment, resulting in poor automation and operational safety. Second, there is the problem of surge current impact upon power-up. Conventional linear power supplies have a fast voltage build-up speed upon power-up, generating a large surge current. When directly powering downstream DAC circuits, analog audio amplifier circuits, and power amplifier circuits, this can easily impact delicate downstream circuits, accelerate component aging, and shorten the overall lifespan of the equipment. Third, there is a lack of... Real-time status monitoring and intelligent interaction capabilities are lacking. Traditional power supplies lack multi-node parameter detection circuits, making it impossible to collect real-time data on output voltage, output current, transformer operating temperature, and voltage regulator circuit status. The system's main controller cannot obtain the power supply's health status, and faults cannot be predicted in advance, making it difficult to guarantee the overall reliability of the system. Fourth, noise and ripple control are difficult to balance. While pure switching power supplies have high power supply efficiency, they have inherent switching interference noise and large output ripple, which cannot meet the ultra-low noise power supply requirements of high-fidelity audio equipment. While conventional linear power supplies have acceptable noise performance, they lack intelligent functions such as voltage adaptive switching, soft start, and status monitoring. A single power supply architecture cannot simultaneously meet the multiple application requirements of wide voltage adaptive adaptation, low noise output, safe start, and intelligent status monitoring.

[0004] Therefore, the industry urgently needs to develop a hybrid intelligent power supply system with automatic input voltage recognition and adaptive switching, soft-start protection against shocks, low noise and low ripple output, real-time status monitoring of multiple parameters, and fault reporting capabilities. Summary of the Invention

[0005] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring to solve the aforementioned technical problems.

[0006] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring, comprising: a power input module, a transformer power supply module, a rectification and linear voltage regulation module, a microcontroller main control unit, a power soft-start module, a power status monitoring module, a ripple sampling hardware branch, and a communication interface module.

[0007] The power input module is compatible with AC110V and AC220V mains power input and has a built-in 12V switching power supply circuit, which supplies power to the microcontroller main control unit and system control circuit.

[0008] The transformer power supply module adopts a toroidal transformer with two sets of windings of the same specification in the primary stage. The two sets of windings are connected to controlled relays K1 and K2 respectively. The relays are used to switch the two sets of windings to be connected in series or in parallel. When the system is powered on, the two sets of windings are initially connected in series by default.

[0009] The microcontroller main control unit samples the secondary rectified output voltage of the toroidal transformer and controls the relay to switch the winding connection mode according to the sampled voltage and the current winding connection mode, adapting to AC110V or AC220V mains power input.

[0010] The rectification and linear regulation module is provided with a bridge rectifier circuit, a filter capacitor and a composite low-noise linear regulation circuit in sequence. The composite low-noise linear regulation circuit is based on LM317 and integrates an operational amplifier error amplification structure, a BJT feedback control structure and a MOS power transistor output adjustment structure.

[0011] The secondary output of the toroidal transformer is connected to a composite low-noise linear voltage regulator circuit after passing through a bridge rectifier circuit and a filter capacitor.

[0012] The power soft-start module is electrically connected to the microcontroller main control unit. By controlling the reference input voltage of the operational amplifier, the output voltage is raised in a controlled and gradual manner after the transformer winding connection is configured, so as to achieve 0V start-up adjustment.

[0013] The power status monitoring module is deployed at key working nodes of the power system to collect output voltage, output current, transformer temperature and voltage regulator circuit operating status parameters. The monitoring signal is connected to the ADC interface of the microcontroller main control unit via an analog sampling circuit.

[0014] The ripple sampling hardware branch is connected to the output of the composite low-noise linear voltage regulator circuit to collect the effective value of the output ripple in real time.

[0015] The microcontroller main control unit has built-in dual interlocking logic of winding switching software timing interlock and hardware physical interlock, adaptive step soft start control logic of current hysteresis constraint, and adaptive ripple suppression logic based on two-level discrete clustering fitting.

[0016] The communication interface module is a serial communication interface. The microcontroller main control unit establishes communication with the system main control SOC through the serial communication interface and transmits output voltage, output current, transformer temperature, power supply operating status, fault alarm information and parameter self-calibration records to the system main control SOC.

[0017] Preferably, the microcontroller main control unit has built-in mains voltage hysteresis judgment and control logic, and sets a voltage judgment constraint formula:

[0018] In series mode:

[0019] The input is determined to be AC110V, and the system switches to parallel mode.

[0020] It is determined to be AC220V input and remains in series mode;

[0021] In parallel mode:

[0022] It is determined to be AC110V input, and the parallel connection mode is maintained.

[0023] The input is determined to be AC220V, and the system switches to series mode.

[0024] in, Real-time sampling voltage of the transformer secondary rectifier; This is the upper limit threshold of the secondary voltage corresponding to the 110V input in series mode; This is the lower limit threshold of the secondary voltage corresponding to a 220V input in series mode; This is the upper limit threshold of the secondary voltage corresponding to a 110V input in parallel mode. This is the lower limit threshold of the secondary voltage corresponding to a 220V input in parallel mode.

[0025] When the sampled voltage is within the hysteresis range between the two thresholds of the corresponding mode The microcontroller main control unit maintains the current transformer winding connection state unchanged;

[0026] In parallel mode, if Immediately shut down the power output and trigger an overvoltage fault. This is the overvoltage threshold.

[0027] Preferably, the microcontroller main control unit has built-in winding switching electrical interlock timing logic. After the voltage type is determined, it controls the currently working relay to disconnect and maintain a static delay. ;

[0028] The relay in the target connection mode is activated, and the relay contact level status is synchronously collected for readback verification.

[0029] If the verification result is consistent with the preset winding topology, the switching is determined to be complete. If the verification is inconsistent, a single switching retry is automatically initiated. If the retry fails, a status fault flag is triggered and stored in the fault register.

[0030] In this circuit, a hardware interlock is set up in the relay drive circuit: the normally open auxiliary contact of K1 is connected in series in the series drive circuit of K2, and the normally open auxiliary contact of K2 is connected in series in the series drive circuit of K1. This prevents the winding short circuit caused by one relay being engaged and the other being released, and allows the two relays to be engaged or released at the same time.

[0031] Preferably, the microcontroller main control unit has built-in adaptive stepped soft-start control logic with current hysteresis constraint, and sets a dynamic boost step size calculation formula:

[0032] ,in, This represents the real-time boost step size for the nth step. As the initial reference boost step size, This is the step size decay coefficient. This is the output voltage for the current gear. This is the rated output voltage of the power supply;

[0033] Simultaneously, the current hysteresis recovery judgment condition is set: ,in, To output instantaneous current, The overcurrent threshold, This is the current hysteresis margin;

[0034] The microcontroller main control unit gradually increases the output voltage according to the dynamic boost step size, wherein the instantaneous output current is sampled in real time during the boost process. ,when At this time, pause the stepped voltage boost and lock the current voltage level until the desired voltage is met. When the hysteresis recovery condition is met, the boost process for the subsequent gears continues according to the dynamic boost step size.

[0035] Preferably, the microcontroller main control unit is equipped with a pre-established adaptive calculation formula for the sampling period: ,in, For real-time sampling period, As the reference sampling period, This is the load adjustment coefficient. This is the temperature regulation coefficient. This is the ripple adjustment coefficient. This refers to the real-time output power of the power supply. This refers to the rated output power of the power supply. This refers to the real-time temperature of the transformer. The reference temperature is room temperature. The over-temperature protection threshold; To output the RMS value of the ripple; The preset reference ripple value;

[0036] The ADC sampling interval for voltage, current, and temperature parameters is dynamically adjusted based on the aforementioned adaptive calculation formula for the sampling period.

[0037] Preferably, the microcontroller main control unit is equipped with a pre-established formula for calculating the maximum output power under transformer temperature constraints:

[0038] ,in, This represents the maximum allowable output power at the current temperature. The power attenuation coefficient;

[0039] The real-time clamping power supply output power is calculated based on the maximum output power calculation formula.

[0040] Preferably, the microcontroller main control unit acquires the effective value of the ripple output by the linear voltage regulator module in real time through the ripple sampling hardware branch. ;

[0041] Will Compared with the preset reference ripple value The real-time ripple deviation is obtained by performing a difference calculation.

[0042] The threshold range to which the real-time ripple deviation belongs is detected, and the corresponding error amplification gain level is matched and called to complete the coarse adjustment of the graded level. The coarse adjustment gain level has cross-range switching and locking logic.

[0043] The microcontroller uses fixed sampling points as the timing window to continuously sample the residual ripple deviation after coarse adjustment for multiple cycles, thus constructing a discrete deviation timing sampling sequence.

[0044] Using a preset baseline ripple value as a zero-deviation baseline, the positive and negative deviation polarities are determined by comparing the sampling sequences.

[0045] Perform differential operations on adjacent sampled values ​​to calculate the real-time rate of change of the deviation amplitude and its positive or negative polarity;

[0046] The sampling sequence is divided into low-amplitude, medium-amplitude, and high-amplitude intervals based on the magnitude of the deviation sampling values. Sampling points in the same interval are grouped together, and a linear distribution is determined by piecewise linear fitting using the least squares method with the minimum value of each group as the benchmark.

[0047] The sampled sequence is divided into secondary discrete clusters according to the positive and negative polarity and magnitude of the deviation change rate to determine the secondary linear distribution;

[0048] By comparing the first-order and second-order linear distributions, the maximum overlapping layer is determined and the independent overlapping range within the layer is extracted to determine the deviation disturbance attribute.

[0049] The BJT feedback bias parameter adjustment direction is uniquely matched based on the deviation polarity, and the adjustment rate level is matched based on the deviation amplitude and deviation attribute.

[0050] Set a constraint on the maximum adjustment step size of the BJT feedback bias fine-tuning in a single operation. After each fine-tuning is completed, resample and update the deviation time series to dynamically iterate and adjust the adjustment direction and rate.

[0051] When the residual ripple deviation of multiple consecutive samplings falls within the preset allowable range, the current BJT feedback bias parameter is locked until a sudden load change or environmental temperature disturbance is detected that causes the ripple deviation to exceed the range, at which point the gain coarse adjustment and bias fine adjustment are restarted.

[0052] Preferably, the microcontroller main control unit is also used to classify faults into three levels: early warning type, self-healing minor faults, and non-self-healing severe faults;

[0053] When it is a warning type fault, the warning information is only reported through the communication interface, and no reset or shutdown operation is triggered.

[0054] When it is a minor fault, a delay is triggered. It will then automatically reset and restart.

[0055] When a severe fault occurs, the power output channel is immediately blocked. At this time, when a fault event is triggered, the composite information of the fault event is split and the indivisible fault information atomic units are extracted. The fault information atomic units include a fault level identifier unit, a fault type identifier unit, and an associated operating condition parameter unit.

[0056] Each fault information atomic unit is mapped to a preset standardized fault coding meta-set to generate a basic coding value corresponding to each atomic unit. The standardized fault coding meta-set contains standardized coding rules corresponding to each fault information atomic unit.

[0057] The real-time operating parameters corresponding to the associated operating condition parameter unit are quantized. Based on the preset parameter threshold and encoding length constraint, the quantized parameter values ​​are range-shrinked to obtain the target quantized value that adapts to the encoding field length.

[0058] For the basic coding value corresponding to each fault information atomic unit, corresponding correction factors are introduced and combined to generate the intermediate coding value corresponding to each atomic unit. The correction factors include fault level correction factor, fault type correction factor and operating condition parameter correction factor, which are used to correct the deviation of the basic coding value.

[0059] The intermediate encoded values ​​are iteratively intersected and checked. If there are field conflicts, the encoded values ​​are adjusted based on the preset priority rules to ensure that there are no conflicts in all fields.

[0060] The verified intermediate code values ​​are coupled and fused together, and a verification field generated based on all the intermediate code values ​​is added to form a fixed-length hierarchical standardized fault code.

[0061] The standardized fault code and fault occurrence timestamp are synchronously packaged and uploaded to the system main control SOC through the communication interface module. The system main control SOC restores the fault level, fault type, associated operating condition parameters and verification information from the standardized fault code based on the preset encoding parsing rules.

[0062] Compared with the prior art, the beneficial effects of this application are as follows:

[0063] It achieves automatic AC110V / 220V mains power adaptation, and with safety interlocks and hysteresis control, it solves the problems of manual switching errors and power grid fluctuation interference. The composite voltage regulation architecture combined with adaptive ripple suppression balances high efficiency and ultra-low noise, and the ripple and efficiency indicators are better than those of traditional linear power supplies. Through intelligent soft start, multi-parameter monitoring and hierarchical fault handling, it effectively suppresses surges and improves the reliability of system operation. The standardized communication interface supports the uploading of operating condition and fault data, which facilitates integration and management with the upper-level system.

[0064] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0065] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0066] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0067] Figure 1 This is a structural diagram of a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring, according to an embodiment of the present invention.

[0068] Figure 2 This is a schematic diagram of the composite voltage regulator and soft-start circuit in an embodiment of the present invention. Detailed Implementation

[0069] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0070] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring, comprising: a power input module, a transformer power supply module, a rectification and linear voltage regulation module, a microcontroller main control unit, a power soft-start module, a power status monitoring module, a ripple sampling hardware branch, and a communication interface module.

[0071] The power input module is compatible with AC110V and AC220V mains power input and has a built-in 12V switching power supply circuit, which supplies power to the microcontroller main control unit and system control circuit.

[0072] The transformer power supply module adopts a toroidal transformer with two sets of windings of the same specification in the primary stage. The two sets of windings are connected to controlled relays K1 and K2 respectively. The relays are used to switch the two sets of windings to be connected in series or in parallel. When the system is powered on, the two sets of windings are initially connected in series by default.

[0073] The microcontroller main control unit samples the secondary rectified output voltage of the toroidal transformer and controls the relay to switch the winding connection mode according to the sampled voltage and the current winding connection mode, adapting to AC110V or AC220V mains power input.

[0074] The rectification and linear regulation module is provided with a bridge rectifier circuit, a filter capacitor and a composite low-noise linear regulation circuit in sequence. The composite low-noise linear regulation circuit is based on LM317 and integrates an operational amplifier error amplification structure, a BJT feedback control structure and a MOS power transistor output adjustment structure.

[0075] The secondary output of the toroidal transformer is connected to a composite low-noise linear voltage regulator circuit after passing through a bridge rectifier circuit and a filter capacitor.

[0076] The power soft-start module is electrically connected to the microcontroller main control unit. By controlling the reference input voltage of the operational amplifier, the output voltage is raised in a controlled and gradual manner after the transformer winding connection is configured, so as to achieve 0V start-up adjustment.

[0077] The power status monitoring module is deployed at key working nodes of the power system to collect output voltage, output current, transformer temperature and voltage regulator circuit operating status parameters. The monitoring signal is connected to the ADC interface of the microcontroller main control unit via an analog sampling circuit.

[0078] The ripple sampling hardware branch is connected to the output of the composite low-noise linear voltage regulator circuit to collect the effective value of the output ripple in real time.

[0079] The microcontroller main control unit has built-in dual interlocking logic of winding switching software timing interlock and hardware physical interlock, adaptive step soft start control logic of current hysteresis constraint, and adaptive ripple suppression logic based on two-level discrete clustering fitting.

[0080] The communication interface module is a serial communication interface. The microcontroller main control unit establishes communication with the system main control SOC through the serial communication interface and transmits output voltage, output current, transformer temperature, power supply operating status, fault alarm information and parameter self-calibration records to the system main control SOC.

[0081] The power input module includes an EMI filter circuit, a surge protection circuit, and a 12V isolated flyback switching power supply circuit. The EMI filter circuit employs a two-stage common-mode inductor and differential-mode capacitor structure, with a common-mode inductor inductance of 10mH and a differential-mode capacitor capacitance of 0.1mH. / 275VAC; the surge protection circuit uses a 14D471K varistor in series with a gas discharge tube. The 12V flyback switching power supply uses a UC3843 as the control chip, the transformer uses an EE25 magnetic core, the primary winding has 30 turns, the secondary winding has 6 turns, and the auxiliary winding has 8 turns; the switching transistor is an N-channel MOSFET IRF840, the output rectifier diode is a Schottky diode MBR20100, and the output filter capacitor is a 2200μF / 25V electrolytic capacitor connected in parallel with a 0.1 Ceramic capacitors. This switching power supply provides a stable 12V DC power supply for the microcontroller main control unit, relay drive circuit, and status monitoring circuit, with an output voltage accuracy of ±0.5V and an output ripple of <50mV.

[0082] The transformer power supply module uses a primary double-winding toroidal transformer with a magnetic core model of H50×30×20. The two primary windings have identical parameters: 110 turns, 0.8mm wire diameter, and 1.2H inductance. The secondary single winding has 24 turns, 1.5mm wire diameter, and 50mH inductance. The two primary windings are connected to two double-pole double-throw relays K1 and K2, respectively. The relays are Omron G2R-1-12VDC, with a rated contact current of 10A / 250VAC. The winding connection topology is as follows:

[0083] Series mode (default power-on state): K1 normally closed contact is closed, K2 normally closed contact is closed, the tail end of winding 1 is connected to the head end of winding 2, the total number of turns is 220, and it is compatible with AC220V input.

[0084] Parallel mode: K1 normally open contact is closed, K2 normally open contact is closed, winding 1 and winding 2 are connected at their beginning and end, with a total of 110 turns, compatible with AC110V input. The relay drive circuit uses an NPN transistor S8050 as the switching transistor, with a 1kΩ current-limiting resistor connected in series at the base to the microcontroller's I / O port, and the collector connected in series with the relay coil to a 12V power supply. A 1N4007 freewheeling diode is connected in reverse parallel across the relay coil.

[0085] The rectifier and linear regulator module consists of a bridge rectifier circuit, a Π-type filter circuit, and a composite low-noise linear regulator circuit. The bridge rectifier circuit uses a KBPC1010 rectifier bridge with a rated current of 10A and a reverse withstand voltage of 1000V; the Π-type filter circuit consists of two 2200Ω... Composed of a 50V electrolytic capacitor and a 10mH differential-mode inductor, the filtered DC voltage ripple is <100mV. Figure 2 The diagram shows the schematic of a composite voltage regulator and soft-start circuit. On the left is an LM317 reference source circuit, providing a stable 1.25V reference voltage. In the middle is an OP07 error amplifier circuit, with its inverting input connected to the output voltage divider sampling and its non-inverting input connected to the microcontroller's DAC control signal. On the right is an IRF9540 power regulator and output filter circuit, achieving 0V-start soft-start and low-ripple voltage regulation.

[0086] Composite low-noise linear voltage regulator circuit (revised): LM317 serves as the reference source, with its output connected to a 10kΩ resistor to ground, providing a stable 1.25V reference voltage; the inverting input of operational amplifier OP07 is connected to the output voltage sampling voltage divider resistors (R1=1kΩ, R2=24kΩ), and the non-inverting input is connected to the output of the microcontroller DAC; the output of OP07 is connected to the base of NPN BJT transistor 2N3904, the collector of 2N3904 is connected to the gate of P-channel MOSFET IRF9540, and the emitter is grounded; the source of IRF9540 is connected to the output of a Π-type filter circuit, and the drain is connected to the power supply output; a 10kΩ resistor is connected in series between the gate and the source to provide gate bias; a 1000μF / 35V electrolytic capacitor and a 0.1μF / 35V electrolytic capacitor are connected in parallel at the output. Ceramic capacitors and 10 Tantalum capacitors further suppress high-frequency ripple. Working principle: The microcontroller DAC outputs a 0~1.25V voltage as a reference. OP07 compares and amplifies the sampled output voltage with the reference voltage, and adjusts the gate voltage of IRF9540 through 2N3904. When the output voltage increases, the output of OP07 increases, the conduction of 2N3904 increases, the gate voltage of IRF9540 decreases, the on-resistance increases, and the output voltage decreases; conversely, the opposite occurs, achieving negative feedback voltage regulation. A P-channel MOSFET serves as the main regulating transistor, with an on-resistance of only 20mΩ, significantly reducing the voltage drop and power consumption of the voltage regulator circuit; the BJT feedback structure effectively suppresses high-frequency ripple above 10kHz.

[0087] The power supply soft-start module utilizes the built-in 12-bit DAC output control signal of the microcontroller. The DAC model is an STM32F103C8T6 built-in DAC with an output voltage range of 0-3.3V. After a 10kΩ + 2.2kΩ voltage divider, a reference voltage of 0-1.25V is obtained and connected to the non-inverting input of the OP07. After the transformer winding switching is completed, the microcontroller controls the DAC output voltage to gradually increase from 0V, and the voltage at the non-inverting input of the OP07 increases accordingly. The output voltage smoothly rises from 0V to the rated value of 24V, achieving true 0V start-up. During the soft-start process, the microcontroller samples the output current in real time. When the current exceeds the overcurrent threshold, the DAC output boost is paused until the current returns to normal.

[0088] The power status monitoring module is deployed at the following key operational nodes:

[0089] Output voltage sampling: The sampling point is located at the output terminal of the composite linear regulator circuit. A 1kΩ + 24kΩ voltage divider resistor is used to convert the 24V output voltage to 0.96V, which is then filtered by an RC filter circuit (1kΩ + 0.1kΩ). Then connect it to the ADC1 channel of the microcontroller;

[0090] Output current sampling: A 0.01Ω / 5W high-precision sampling resistor is connected in series in the output circuit. The sampled voltage is amplified 100 times by the operational amplifier INA128 and then connected to the microcontroller's ADC2 channel.

[0091] Transformer temperature sampling: An NTC thermistor MF52-103F3950 is used and installed on the surface of the toroidal transformer core. A voltage divider circuit is formed by a 10kΩ pull-up resistor, and the sampling voltage is connected to the ADC3 channel of the microcontroller.

[0092] Voltage regulator circuit status sampling: Sample the input-output voltage difference of LM317 and the source-drain voltage of MOS transistor, and connect them to the microcontroller's ADC4 and ADC5 channels respectively to determine whether the voltage regulator circuit is working properly.

[0093] Ripple sampling hardware branch: adopts AC coupling, through 1 Ceramic capacitors isolate the DC component, allowing only AC ripple signals to pass through. A differential operational amplifier AD8221 is used as the preamplifier, with a common-mode rejection ratio of 100dB, and its gain can be adjusted via an external resistor. After amplification by the AD8221, the ripple signal is connected to a true RMS converter chip AD637, which outputs a DC voltage proportional to the RMS ripple value. The output voltage of the AD637 is then RC-filtered and connected to the microcontroller's ADC6 channel, with a sampling frequency of 10kHz. This sampling branch has a bandwidth of 10Hz~1MHz, accurately acquiring low-frequency and high-frequency ripple from the power supply output, with a ripple measurement range of 0.1mV~100mV and a measurement accuracy of ±5%. Ripple sampling calibration method: Using a standard signal source, input a 1mV, 1kHz sine wave signal, adjust the gain resistor of the AD8221 until the AD637 output voltage is 1V, completing the calibration. The calibration parameters are stored in the microcontroller's internal Flash memory and are automatically loaded upon power-on.

[0094] The RS485 communication interface module uses the MAX485 chip, with a baud rate of 9600bps, 8 data bits, 1 stop bit, no parity check, and adopts the Modbus-RTU communication protocol. The microcontroller acts as a slave station, with address 0x01, and supports the system master control SOC in reading power supply operating parameters and sending control commands.

[0095] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. The microcontroller main control unit has built-in mains voltage hysteresis judgment and control logic, and sets a voltage judgment constraint formula:

[0096] In series mode:

[0097] The input is determined to be AC110V, and the system switches to parallel mode.

[0098] It is determined to be AC220V input and remains in series mode;

[0099] In parallel mode:

[0100] It is determined to be AC110V input, and the parallel connection mode is maintained.

[0101] The input is determined to be AC220V, and the system switches to series mode.

[0102] in, Real-time sampling voltage of the transformer secondary rectifier; This is the upper limit threshold of the secondary voltage corresponding to the 110V input in series mode; This is the lower limit threshold of the secondary voltage corresponding to a 220V input in series mode; This is the upper limit threshold of the secondary voltage corresponding to a 110V input in parallel mode. This is the lower limit threshold of the secondary voltage corresponding to a 220V input in parallel mode.

[0103] When the sampled voltage is within the hysteresis range between the two thresholds of the corresponding mode The microcontroller main control unit maintains the current transformer winding connection state unchanged;

[0104] In parallel mode, if Immediately shut down the power output and trigger an overvoltage fault. This is the overvoltage threshold.

[0105] In this embodiment, the toroidal transformer has a turns ratio of 220V:24V, therefore:

[0106] In series mode: with AC110V input, the secondary open-circuit voltage is approximately 12V. Considering grid fluctuations of ±10%, the maximum voltage is 13.2V, therefore, the setting is... =14V; When AC220V input is applied, the secondary open-circuit voltage is approximately 24V. Considering grid fluctuations of ±10%, the minimum voltage is 21.6V, therefore, it is set to... =20V;

[0107] In parallel mode: with AC110V input, the secondary no-load voltage is approximately 24V. Considering grid fluctuations of ±10%, the maximum voltage is 26.4V, therefore, the setting is... =28V; When AC220V input is applied, the secondary open-circuit voltage is approximately 48V. Considering grid fluctuations of ±10%, the minimum voltage is 43.2V, therefore, it is set to... =40V; Overvoltage threshold =45V. The secondary rectified voltage sampling circuit uses a 10kΩ + 2.2kΩ voltage divider resistor to convert the maximum 50V DC voltage after secondary rectification to ≤9V. After a second voltage divider of 10kΩ + 10kΩ, it is converted to ≤4.5V and connected to the ADC0 channel of the microcontroller. The sampling frequency is 1kHz, and a 16-times moving average filtering algorithm is used to improve the sampling accuracy.

[0108] After the system is powered on, the microcontroller continuously samples the secondary rectified voltage 10 times and takes the average value as... Select the corresponding judgment threshold according to the current winding connection mode; if a mode switch is required, execute the winding switching procedure; if the sampled voltage is in the hysteresis range, maintain the current state and resample and judge every 1 second; in parallel mode, if If the voltage is ≥45V, immediately block the output and trigger an overvoltage fault.

[0109] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. The microcontroller main control unit has built-in winding switching electrical interlocking timing logic. After the voltage type is determined, it controls the currently operating relay to disconnect and maintain a static delay. ;

[0110] The relay in the target connection mode is activated, and the relay contact level status is synchronously collected for readback verification.

[0111] If the verification result is consistent with the preset winding topology, the switching is determined to be complete. If the verification is inconsistent, a single switching retry is automatically initiated. If the retry fails, a status fault flag is triggered and stored in the fault register.

[0112] In this circuit, a hardware interlock is set up in the relay drive circuit: the normally open auxiliary contact of K1 is connected in series in the series drive circuit of K2, and the normally open auxiliary contact of K2 is connected in series in the series drive circuit of K1. This prevents the winding short circuit caused by one relay being engaged and the other being released, and allows the two relays to be engaged or released at the same time.

[0113] It should be noted that the fault register is a non-volatile memory inside the microcontroller, used to store the fault type, occurrence time and related operating parameters, and the data is not lost after power failure.

[0114] After the voltage type is determined, the microcontroller first controls the currently operating relay to disconnect, maintaining a static delay Td=100ms (the typical relay release time is 10ms, leaving a 10-fold margin to ensure the contacts are completely disconnected). After the delay, the relay in the target connection mode is driven to engage, and the microcontroller's I / O port is used to collect the level status of the relay auxiliary contacts for verification. If the verification result is consistent with the preset winding topology (e.g., in parallel mode, both normally open contacts of K1 and K2 are at a high level), the switching is considered complete, and the soft-start process begins. If the verification is inconsistent, a single switching retry is automatically initiated, and the retry process is exactly the same as the first switching. If the retry still fails, a relay switching failure fault flag is triggered, stored in the microcontroller's internal fault register (address 0x08005000), and the winding is kept in series mode, disabling the output voltage. Hardware interlock circuit: The normally open auxiliary contact of K1 is connected in series in the series mode drive circuit of K2, and the normally open auxiliary contact of K2 is connected in series in the series mode drive circuit of K1. When any one relay is engaged, the series drive circuit of the other relay is physically disconnected, eliminating the risk of short circuit in the winding caused by one relay being engaged and the other being released at the hardware level. At the same time, it allows the two relays to be engaged simultaneously to achieve parallel mode.

[0115] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. The microcontroller main control unit has built-in adaptive stepped soft-start control logic with current hysteresis constraint, and sets a dynamic boost step size calculation formula:

[0116] ,in, This represents the nth step size for real-time voltage boost, ranging from 0.1V to 0.5V. This is the initial reference boost step size, ranging from 0.4V to 0.6V. This is the step size decay coefficient, with a value ranging from 0.7 to 0.9. This is the output voltage for the current gear. This is the rated output voltage of the power supply;

[0117] Simultaneously, the current hysteresis recovery judgment condition is set: ,in, To output instantaneous current, This is the overcurrent threshold, and its value is 1.1 to 1.3 times the rated current. This is the current hysteresis margin, with a value ranging from 0.4A to 0.6A;

[0118] The microcontroller main control unit gradually increases the output voltage according to the dynamic boost step size, wherein the instantaneous output current is sampled in real time during the boost process. ,when At this time, pause the stepped voltage boost and lock the current voltage level until the desired voltage is met. When the hysteresis recovery condition is met, the boost process for the subsequent gears continues according to the dynamic boost step size.

[0119] The adaptive stepped soft-start with current hysteresis constraint is a soft-start method that dynamically adjusts the boost step size based on the output voltage and prevents inrush current through current hysteresis protection. It achieves this by using a microcontroller's built-in DAC to output a gradually increasing reference voltage, thus solving the problem of excessive inrush current in traditional fixed-step soft-start under large capacitor loads. Its specific parameters and process are as follows:

[0120] Initial reference boost step size =0.5V;

[0121] Step size decay coefficient =0.8;

[0122] Power supply rated output voltage =24V;

[0123] Overcurrent threshold =6A (1.2 times the rated current);

[0124] Current hysteresis margin =0.5A;

[0125] boost time interval =10ms;

[0126] Overcurrent recovery detection cycle =10ms.

[0127] The dynamic boost step size calculation formula achieves an adaptive effect of rapid initial boost and slow boost when approaching the rated value, which shortens the soft start time and avoids overshoot.

[0128] Soft start process:

[0129] After the winding switching is completed, the microcontroller initializes the DAC output to 0V, and the output voltage is 0V.

[0130] Calculate the target voltage for the current gear according to the dynamic boost step size, control the DAC to output the corresponding voltage, and maintain it for 10ms;

[0131] Real-time sampling of output current ,like >6A, pause boost and lock the current DAC output;

[0132] The output current is checked every 10ms. When the voltage is ≤5.5A, resume the boost process;

[0133] The soft start is complete when the output voltage reaches 24V.

[0134] Connect 10000 in parallel at the output. In the case of a large capacitor load, the surge current of a traditional fixed-step soft-start is 12A, while the surge current of the adaptive stepped soft-start of this invention is only 5.8A, effectively protecting the precision circuitry at the back end.

[0135] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. The microcontroller main control unit is equipped with a pre-established adaptive calculation formula for the sampling period. ,in, For real-time sampling period, The reference sampling period ranges from 80ms to 120ms. This is the load regulation coefficient, with a value ranging from 0.4 to 0.6. This is the temperature regulation coefficient, with a value ranging from 0.2 to 0.4. This is the ripple adjustment factor, with a value ranging from 0.1 to 0.3. This refers to the real-time output power of the power supply. This refers to the rated output power of the power supply. This refers to the real-time temperature of the transformer. The reference temperature is 25℃. The over-temperature protection threshold has a range of 110℃ to 130℃. To output the RMS value of the ripple; The preset reference ripple value ranges from 0.8mV to 1.2mV.

[0136] The ADC sampling interval for voltage, current, and temperature parameters is dynamically adjusted based on the aforementioned adaptive calculation formula for the sampling period.

[0137] The ADC sampling interval for voltage, current, and temperature parameters is dynamically adjusted based on the aforementioned adaptive calculation formula for the sampling period.

[0138] Adaptive sampling period is a control method that dynamically adjusts the ADC sampling interval based on the real-time output power of the power supply. It reduces the sampling frequency under light load to decrease the computational burden on the microcontroller, and increases the sampling frequency under heavy load to ensure monitoring accuracy. Its specific parameters and implementation are as follows:

[0139] Reference sampling period =100ms;

[0140] Load adjustment coefficient =0.5;

[0141] Temperature regulation coefficient =0.3;

[0142] Ripple adjustment coefficient =0.2;

[0143] Rated output power of power supply =120W;

[0144] Lower limit of sampling period =100ms;

[0145] upper limit of sampling period =500ms.

[0146] Real-time output power calculation formula: ,in, This is the current output voltage sample value. The current output current is sampled. The two parameters are sampled synchronously with a sampling time difference of <1ms.

[0147] The sampling period is limited to between 100ms and 500ms to prevent the sampling frequency from being too high or too low.

[0148] Voltage, current, and temperature parameters use the same adaptive sampling period. The microcontroller triggers ADC sampling via a timer interrupt, updating the sampling period after each sampling, and the next interrupt executes according to the new period. This method achieves an adaptive effect by reducing the sampling frequency under light load to reduce the computational load on the microcontroller, and increasing the sampling frequency under heavy load, high temperature, and high ripple to ensure monitoring accuracy.

[0149] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. The microcontroller main control unit is equipped with a pre-established formula for calculating the maximum output power under transformer temperature constraints. ,in, This represents the maximum allowable output power at the current temperature. This is the power attenuation coefficient, with a value ranging from 0.7 to 0.9;

[0150] The real-time clamping power supply output power is calculated based on the maximum output power calculation formula.

[0151] The maximum output power under transformer temperature constraints is a protection mechanism that linearly adjusts the maximum allowable output power based on the real-time transformer temperature to prevent transformer overheating damage. This addresses the problem of sudden output interruption caused by traditional fixed-threshold over-temperature protection. Its specific parameters and implementation are as follows:

[0152] Power attenuation coefficient =0.8;

[0153] room temperature reference temperature =25℃;

[0154] Over-temperature protection threshold =120℃;

[0155] Temperature warning threshold =80℃.

[0156] Power clamping implementation method: The microcontroller calculates the maximum output power at the current temperature in real time. And convert it to the maximum allowable output current. = When the output current exceeds At this time, the microcontroller reduces the output voltage through the DAC to ensure that the output power does not exceed [the specified value]. .

[0157] Over-temperature protection procedure:

[0158] When the transformer temperature When the temperature reaches ≥80℃, a temperature warning is triggered, and the warning information is reported through the communication interface;

[0159] when When the temperature reaches ≥120℃, immediately block the power output channel (disconnect the MOSFET in the output circuit), trigger the over-temperature protection severe fault, and report the fault code and real-time temperature;

[0160] When the temperature drops below 70℃, the fault is automatically reset and the power-on process is re-executed.

[0161] Existing low-noise power supplies mostly use a single adjustment method with fixed feedback parameters for ripple suppression. This method has several drawbacks, such as the inability to balance deviation convergence speed and adjustment accuracy with fixed gain, difficulty in distinguishing ripple disturbance attributes to adapt to targeted adjustment strategies, and the tendency for over-adjustment oscillation or ripple deterioration after changes in operating conditions. It also cannot maintain stable ultra-low ripple output under complex operating conditions such as sudden load changes and temperature variations. Therefore, this invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring, and is also equipped with a ripple sampling hardware branch.

[0162] The microcontroller main control unit acquires the RMS value of the ripple output from the linear voltage regulator module in real time through the ripple sampling hardware branch. ;

[0163] Will Compared with the preset reference ripple value The real-time ripple deviation is obtained by performing a difference calculation.

[0164] The threshold range to which the real-time ripple deviation belongs is detected, and the corresponding error amplification gain level is matched and called to complete the coarse adjustment of the graded level. The coarse adjustment gain level has cross-range switching and locking logic.

[0165] The microcontroller uses fixed sampling points as the timing window to continuously sample the residual ripple deviation after coarse adjustment for multiple cycles, thus constructing a discrete deviation timing sampling sequence.

[0166] Using a preset baseline ripple value as a zero-deviation baseline, the positive and negative deviation polarities are determined by comparing the sampling sequences.

[0167] Perform differential operations on adjacent sampled values ​​to calculate the real-time rate of change of the deviation amplitude and its positive or negative polarity;

[0168] The sampling sequence is divided into low-amplitude, medium-amplitude, and high-amplitude intervals based on the magnitude of the deviation sampling values. Sampling points in the same interval are grouped together, and a linear distribution is determined by piecewise linear fitting using the least squares method with the minimum value of each group as the benchmark.

[0169] The sampled sequence is divided into secondary discrete clusters according to the positive and negative polarity and magnitude of the deviation change rate to determine the secondary linear distribution;

[0170] By comparing the first-order and second-order linear distributions, the maximum overlapping layer is determined and the independent overlapping range within the layer is extracted to determine the deviation disturbance attribute.

[0171] The BJT feedback bias parameter adjustment direction is uniquely matched based on the deviation polarity, and the adjustment rate level is matched based on the deviation amplitude and deviation attribute.

[0172] Set a constraint on the maximum adjustment step size of the BJT feedback bias fine-tuning in a single operation. After each fine-tuning is completed, resample and update the deviation time series, and dynamically iterate to adjust the adjustment direction and rate.

[0173] When the residual ripple deviation of multiple consecutive samplings falls within the preset allowable range, the current BJT feedback bias parameter is locked until a sudden load change or environmental temperature disturbance is detected that causes the ripple deviation to exceed the range, at which point the gain coarse adjustment and bias fine adjustment are restarted.

[0174] Ripple sampling hardware branch: adopts AC coupling, through Ceramic capacitors isolate the DC component, allowing only AC ripple signals to pass through. A differential operational amplifier AD8221 is used as the preamplifier, with a common-mode rejection ratio of 100dB and gain adjustable via an external resistor. After amplification by AD8221, the ripple signal is connected to a true RMS converter chip AD637, outputting a DC voltage proportional to the RMS ripple value. The AD637's output voltage is then RC-filtered and connected to the microcontroller's ADC6 channel, with a sampling frequency of 10kHz. This sampling branch has a bandwidth of 10Hz~1MHz, accurately acquiring low-frequency and high-frequency ripple from the power supply output, with a ripple measurement range of 0.1mV~100mV and a measurement accuracy of ±5%.

[0175] In this embodiment, the ripple adaptive adjustment algorithm flow is as follows: Gain coarse adjustment stage:

[0176] Preset reference ripple value =1mV;

[0177] The error amplification gain is divided into three levels: G1=10 times, G2=100 times, and G3=1000 times.

[0178] The microcontroller acquires the RMS value of the ripple in real time. Calculate the ripple deviation. ;

[0179] like 10mV, switch to G1 mode, which is suitable for rapid adjustment of large deviations; if Switch to G2 mode, which is suitable for medium deviation transition adjustment; if ≤1mV, switch to G3, which is suitable for fine adjustment of small deviations;

[0180] Cross-range switching interlocking logic: A gain state machine is set up at the software level, and state transitions are only allowed between adjacent levels: G1↔G2↔G3, to prevent system oscillations caused by sudden gain changes; after each gain switch, a 20ms delay is set to wait for the circuit to stabilize before the next ripple sampling is performed; if a non-adjacent level switching request is detected, the system automatically rejects it and maintains the current gain level, while recording a gain switching anomaly log. Compared with traditional fixed gain adjustment, the graded coarse adjustment can shorten the ripple convergence time from 500ms to less than 100ms.

[0181] Biased temporal sampling and cluster analysis stage:

[0182] Using a sampling time window of 10ms, residual ripple deviations were continuously collected for 20 cycles to construct a discrete deviation time-series sampling sequence. ;

[0183] First discrete clustering partition (by amplitude):

[0184] Low amplitude range: ;

[0185] Medium amplitude range: 2mV;

[0186] High amplitude range: 2mV;

[0187] Sampling points in the same interval are grouped together. Based on the minimum value of each group, piecewise linear fitting is performed using the least squares method to obtain a linear distribution y1=k1x+b1.

[0188] Secondary discrete clustering partitioning (by rate of change):

[0189] Calculate the difference between adjacent sample values ;

[0190] Positive growth range: 0.1mV;

[0191] Negative growth range: -0.1mV;

[0192] Stable range: ≤0.1mV;

[0193] Sampling points in the same interval are grouped together, and piecewise linear fitting is performed using the least squares method to obtain a quadratic linear distribution y2=k2x+b2;

[0194] Deviation disturbance attribute determination: Compare the first-order linear distribution with the second-order linear distribution to determine the maximum overlap layer, specifically:

[0195] For example, traversing the three groups (A, B, C) of the first cluster and the three groups (D, E, F) of the second cluster, the number of overlapping sampling points for all 9 combinations is calculated. Overlapping sampling points are defined as sampling points that belong to two groups at the same time, and the overlap degree = number of overlapping sampling points / total number of sampling points × 100%. The combination with the highest overlap degree ≥ 70% is the maximum overlap layer. If the overlap layer is low amplitude and in a stable range, it is judged as steady-state deviation; if the overlap layer is medium or high amplitude and in a positive or negative growth range, it is judged as periodic disturbance; if there is no obvious overlap layer, it is judged as random noise.

[0196] BJT feedback bias fine-tuning stage:

[0197] The BJT feedback bias parameter is the base current of the 2N3904, which is achieved by controlling the 10kΩ base resistor through the output voltage of the microcontroller's DAC. The formula for calculating the base current is as follows: Adjustment range 0~100 .

[0198] Adjustment direction: If the ripple deviation is positive ( ), increase the base current; if it is negative ( ), reduce base current;

[0199] Adjustment rate:

[0200] High amplitude deviation (≥2mV): Fast adjustment, single step size 10 Adjustment interval 1ms;

[0201] Mid-amplitude deviation (0.5~2mV): Mid-range tuning, single step size 1 Adjustment interval 10ms;

[0202] Low amplitude deviation (<0.5mV): Fine-tuning, single step size 0.1mV. The adjustment interval is 100ms.

[0203] The maximum adjustment step size in a single operation is constrained to 10. To prevent ripple oscillation caused by excessive adjustment;

[0204] After each fine-tuning, the ripple deviation is re-acquired, the timing sampling sequence is updated, and the adjustment direction and rate are dynamically iteratively adjusted.

[0205] Parameter locking and re-triggering:

[0206] When the residual ripple deviation of three consecutive sampling rounds all fall within the ±0.1mV tolerance range, the current BJT feedback bias parameter is locked.

[0207] Restart the gain coarse-tuning and bias fine-tuning process when any of the following conditions are detected:

[0208] Output current change rate > 1A / 10ms (load sudden change);

[0209] Transformer temperature change rate > 5℃ / 10s (ambient temperature disturbance);

[0210] The ripple deviation exceeds ±0.5mV.

[0211] Under full load output of 120W, the output ripple of the traditional LM317 voltage regulator circuit is 5.2mV. The adaptive ripple suppression circuit of this invention can stabilize the output ripple between 0.3mV and 0.5mV, improving the ripple suppression ratio by more than 10 times.

[0212] This adaptive ripple control scheme achieves rapid convergence and precise suppression of ripple deviation through the coordinated control of a dedicated ripple sampling hardware branch, graded coarse gain adjustment and cross-interval blocking logic, two-level discrete clustering fitting deviation disturbance attributes, and BJT feedback bias fine-tuning iteration. It solves the problem of ripple deterioration and regulation instability that traditional fixed feedback ripple suppression is prone to under conditions such as sudden load changes and temperature variations. It can stably control the output ripple at an extremely low level, improve the low noise performance and regulation robustness of the power supply under complex operating conditions, and is suitable for the power supply needs of high-precision instruments and audio equipment that are sensitive to power supply ripple.

[0213] This invention provides a hybrid low-noise power supply system with adaptive input voltage switching and status monitoring. The single-chip microcomputer main control unit is also used to classify faults into three levels: early warning type, self-healing minor faults, and non-self-healing severe faults.

[0214] When it is a warning type fault, the warning information is only reported through the communication interface, and no reset or shutdown operation is triggered.

[0215] When it is a minor fault, a delay is triggered. It will then automatically reset and restart.

[0216] When a severe fault occurs, the power output channel is immediately blocked. At this time, when a fault event is triggered, the composite information of the fault event is split and the indivisible fault information atomic units are extracted. The fault information atomic units include a fault level identifier unit, a fault type identifier unit, and an associated operating condition parameter unit.

[0217] Each fault information atomic unit is mapped to a preset standardized fault coding meta-set to generate a basic coding value corresponding to each atomic unit. The standardized fault coding meta-set contains standardized coding rules corresponding to each fault information atomic unit.

[0218] The real-time operating parameters corresponding to the associated operating condition parameter unit are quantized. Based on the preset parameter threshold and encoding length constraint, the quantized parameter values ​​are range-shrinked to obtain the target quantized value that adapts to the encoding field length.

[0219] For the basic coding value corresponding to each fault information atomic unit, corresponding correction factors are introduced and combined to generate the intermediate coding value corresponding to each atomic unit. The correction factors include fault level correction factor, fault type correction factor and operating condition parameter correction factor, which are used to correct the deviation of the basic coding value.

[0220] Perform iterative intersection verification on each intermediate encoded value. If there is a field conflict, adjust the encoded value based on the preset priority rules to ensure that there are no conflicts in all fields.

[0221] The verified intermediate code values ​​are coupled and fused together, and a verification field generated based on all the intermediate code values ​​is added to form a fixed-length hierarchical standardized fault code.

[0222] The standardized fault code and fault occurrence timestamp are synchronously packaged and uploaded to the system main control SOC through the communication interface module. The system main control SOC restores the fault level, fault type, associated operating condition parameters and verification information from the standardized fault code based on the preset encoding parsing rules.

[0223] In this embodiment, fault event triggering refers to the moment when the parameters collected by the microcontroller through the power status monitoring module exceed the preset threshold or a logical abnormality is detected. For example, when the output current is detected to exceed 6A for 1 second, the transformer temperature exceeds 120°C, or the relay switching retry fails, it is determined to be a fault event triggering.

[0224] In this embodiment, composite information refers to all the original data related to the fault stored inside the microcontroller at the time the fault occurred, including the time of the fault occurrence, the severity of the fault, the specific type of fault, the output voltage, output current, transformer temperature, voltage difference of the voltage regulator circuit, and all other operating parameters.

[0225] In this embodiment, the following are considered early warning faults: transformer temperature ≥80℃ and <120℃, output voltage deviation ≥5% and <10%, and ripple RMS value ≥1mV and <2mV. Early warning faults are only reported via the RS485 interface; the power supply continues to operate normally, reminding the user to perform timely inspection and maintenance to prevent further deterioration of the fault.

[0226] Self-healing minor faults: single relay switching failure, instantaneous overcurrent, automatically reset and restart after trigger delay (T_r=5s);

[0227] Non-self-healing severe faults: relay switching retry failure, continuous overcurrent ≥1s, transformer overheating ≥120℃, output short circuit, overvoltage, immediately block output.

[0228] In this embodiment, the fault information atomic unit refers to the smallest information unit that can independently describe a certain dimension of the fault and cannot be further subdivided. Specifically, it includes a fault level identifier unit, a fault type identifier unit, and an associated operating condition parameter unit. The fault level identifier unit is used to uniquely identify the severity of the fault, the fault type identifier unit is used to uniquely identify the specific type of fault, and the associated operating condition parameter unit is used to store a single key operating condition parameter directly related to the fault type, rather than all operating parameters. For example, the associated operating condition parameter for a relay switching failure fault is the input voltage at the switching time, the associated operating condition parameter for a continuous overcurrent fault is the output current at the time the fault occurs, and the associated operating condition parameter for an overtemperature fault is the transformer temperature at the time the fault occurs. This ensures the reasonableness of the encoding length and the relevance of the information.

[0229] In this embodiment, the standardized fault coding metadata set is a set of coding rules pre-stored in the non-volatile memory inside the microcontroller. For each fault information atomic unit, a unique coding length and coding mapping relationship are defined. For example, the coding length of the fault level identifier unit is 2 bits, and the mapping relationship is 00 for warning type faults, 01 for self-healing minor faults, 10 for non-self-healing severe faults, and 11 is a reserved bit; the coding length of the fault type identifier unit is 6 bits, and the mapping relationship is 000001 for relay switching failure faults, 000010 for continuous overcurrent faults, 000011 for transformer over-temperature faults, 000100 for output short-circuit faults, and 000101 for voltage regulator circuit abnormal faults. The remaining codes are reserved for adding fault types later; the coding length of the associated operating condition parameter unit is 6 bits, and its basic coding value is the initial code corresponding to the original parameters before subsequent quantization processing.

[0230] In this embodiment, the basic encoding value refers to the initial encoding obtained by directly converting according to the mapping relationship of the standardized fault encoding metadata set. For example, when a continuous overcurrent fault occurs, the basic encoding value of the fault level identifier unit is 10, and the basic encoding value of the fault type identifier unit is 000010.

[0231] In this embodiment, quantization refers to the process of converting real-time operating condition parameters in analog form into digital codes. The parameter threshold value refers to the boundary value of the operating condition parameter under normal operation and fault conditions. The code length constraint means that the associated operating condition parameter unit is fixed at 6 bits, which can represent an integer range from 0 to 63. For example, for a continuous overcurrent fault, the associated operating condition parameter is the output current, with a preset parameter threshold value of 0A to 10A and a code length constraint of 6 bits. Therefore, during quantization, the current range of 0A to 10A is linearly mapped to the integer range of 0 to 63, with each integer unit corresponding to a current value of approximately 0.1587A. Range contraction refers to uniformly mapping parameter values ​​exceeding the parameter threshold value range to the maximum or minimum value of the code range. For example, when the output current exceeds 10A, it is uniformly mapped to 63; when the output current is below 0A, it is uniformly mapped to 0. This ensures that all operating condition parameters can adapt to the length of the 6-bit code field, avoiding code overflow. For example, for a transformer over-temperature fault, the associated operating condition parameter is the transformer temperature. The preset critical value is 25℃ to 120℃. During quantization, the temperature range of 25℃ to 120℃ is linearly mapped to an integer range of 0 to 63. Each integer unit corresponds to a temperature value of approximately 1.508℃. When the temperature exceeds 120℃, it is uniformly mapped to 63, and when the temperature is below 25℃, it is uniformly mapped to 0.

[0232] In this embodiment, the correction factor is a correction coefficient pre-determined based on the measurement error and coding deviation characteristics of each atomic unit. It is used to eliminate systematic deviations in the basic coding value and improve coding accuracy. The fault level correction factor and the fault type correction factor are both 1, because the coding of fault level and fault type is a discrete fixed mapping relationship, and there is no systematic deviation. The operating condition parameter correction factor is determined based on the measurement accuracy of the operating condition parameter. For example, the measurement error of the output current is ±2%, so the operating condition parameter correction factor is 0.98. The combined processing refers to multiplying the basic coding value by the corresponding correction factor and then rounding it to obtain the intermediate coding value. For example, when the output current is 5.0A, the quantized basic coding value is 32. Multiplying it by the correction factor 0.98 gives 31.36, and rounding it gives the intermediate coding value 31.

[0233] In this embodiment, iterative intersection verification refers to sequentially checking whether the actual meanings corresponding to every two intermediate encoded values ​​conform to a preset set of logical rules. The set of logical rules consists of fault attribute constraints pre-stored in the microcontroller, such as: continuous overcurrent faults can only correspond to non-healing severe faults, transformer temperature warnings can only correspond to warning-type faults, and output short-circuit faults can only correspond to non-healing severe faults. The verification process is as follows: first, all intermediate encoded values ​​are converted to their corresponding actual meanings; then, the set of logical rules is traversed to check for any rule violations. If a violation exists, it is determined to be a field conflict. The preset priority rule is that fault type priority is higher than fault level priority. That is, when fault type and fault level conflict, the intermediate encoded value of the fault level is adjusted based on the level corresponding to the fault type. For example, when a continuous overcurrent fault is detected, if the intermediate encoded value of the fault level is 01 (mild fault), and the logical rule stipulates that continuous overcurrent can only correspond to severe faults, then the intermediate encoded value of the fault level is adjusted to 10 (severe fault). Iteration refers to re-executing the above verification process after adjusting the encoded values ​​until all fields conform to the logical rules and there are no conflicts.

[0234] In this embodiment, coupling fusion refers to concatenating the verified intermediate encoded values ​​into a complete binary code in a preset fixed order. This scheme uses a 16-bit fixed-length encoding structure, and the concatenation order is as follows:

[0235] Bits 15-14: Fault Level (00=Warning, 01=Minor, 10=Severe, 11=Reserved);

[0236] Bits 13-8: Fault type (000001=Relay switching failure, 000010=Continuous overcurrent, 000011=Transformer overheating, 000100=Output short circuit, 000101=Voltage regulator circuit malfunction, 000110=Overvoltage).

[0237] Bit 7-2: Associated operating condition parameters (quantization value 0-63);

[0238] Bits 1-0: CRC-2 checksum (generator polynomial) ).

[0239] For example, when a continuous overcurrent fault occurs and the output current is 5.0A, the intermediate code value of the fault level after verification is 10, the intermediate code value of the fault type is 000010, the intermediate code value of the associated operating condition parameter is 31 (011111 in binary), the first 14 bits of the code are 10000010011111, and the check code is 01 calculated by CRC-2. Therefore, the final 16-bit standardized fault code is 1000001001111101 (0x827D in hexadecimal).

[0240] In this embodiment, synchronous packaging refers to encapsulating the code and the fault occurrence timestamp into a communication frame conforming to the Modbus-RTU protocol within 10ms after the standardized fault code is generated, and uploading it. The fault occurrence timestamp is generated by the real-time clock module inside the microcontroller, accurate to the second, and includes year, month, day, hour, minute, and second information. After receiving the communication frame, the system main control SOC can directly extract the fault level, fault type, and associated operating condition parameters from the 16-bit code based on the preset parsing rules that are completely consistent with the standardized fault code metadata set, and verify the correctness of the code through the verification field, without the need for additional custom protocol parsing.

[0241] The beneficial effects of the above technical solution are as follows: by atomically splitting and standardizing the mapping of fault information, a unified structure of fault coding is achieved. Combined with quantitative contraction of operating parameters and multi-field iterative verification, the accuracy and compatibility of the coding are ensured. The additional verification field effectively avoids coding errors caused by communication interference in the industrial environment. The host computer can directly parse the complete fault information, which greatly improves the efficiency of fault diagnosis and tracing and reduces the complexity of system integration.

[0242] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A hybrid low-noise power supply system with adaptive input voltage switching and status monitoring, characterized in that, include: The power input module, transformer power supply module, rectifier and linear regulator module, microcontroller main control unit, power soft start module, power status monitoring module, ripple sampling hardware branch and communication interface module; The power input module is compatible with AC110V and AC220V mains power input and has a built-in 12V switching power supply circuit, which supplies power to the microcontroller main control unit and system control circuit. The transformer power supply module adopts a toroidal transformer with two sets of windings of the same specification in the primary stage. The two sets of windings are connected to controlled relays K1 and K2 respectively. The relays are used to switch the two sets of windings to be connected in series or in parallel. When the system is powered on, the two sets of windings are initially connected in series by default. The microcontroller main control unit samples the secondary rectified output voltage of the toroidal transformer and controls the relay to switch the winding connection mode according to the sampled voltage and the current winding connection mode, adapting to AC110V or AC220V mains power input. The rectification and linear regulation module is provided with a bridge rectifier circuit, a filter capacitor and a composite low-noise linear regulation circuit in sequence. The composite low-noise linear regulation circuit is based on LM317 and integrates an operational amplifier error amplification structure, a BJT feedback control structure and a MOS power transistor output adjustment structure. The secondary output of the toroidal transformer is connected to a composite low-noise linear voltage regulator circuit after passing through a bridge rectifier circuit and a filter capacitor. The power soft-start module is electrically connected to the microcontroller main control unit. By controlling the reference input voltage of the operational amplifier, the output voltage is raised in a controlled and gradual manner after the transformer winding connection is configured, so as to achieve 0V start-up adjustment. The power status monitoring module is deployed at key working nodes of the power system to collect output voltage, output current, transformer temperature and voltage regulator circuit operating status parameters. The monitoring signal is connected to the ADC interface of the microcontroller main control unit via an analog sampling circuit. The ripple sampling hardware branch is connected to the output of the composite low-noise linear voltage regulator circuit to collect the effective value of the output ripple in real time. The microcontroller main control unit has built-in dual interlocking logic of winding switching software timing interlock and hardware physical interlock, adaptive step soft start control logic of current hysteresis constraint, and adaptive ripple suppression logic based on two-level discrete clustering fitting. The communication interface module is a serial communication interface. The microcontroller main control unit establishes communication with the system main control SOC through the serial communication interface and transmits output voltage, output current, transformer temperature, power supply operating status, fault alarm information and parameter self-calibration records to the system main control SOC.

2. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 1, characterized in that, The microcontroller main control unit has built-in mains voltage hysteresis detection and control logic, and sets the voltage determination constraint formula: In series mode: The input is determined to be AC110V, and the system switches to parallel mode. It is determined to be AC220V input and remains in series mode; In parallel mode: It is determined to be AC110V input, and the parallel connection mode is maintained. The input is determined to be AC220V, and the system switches to series mode. in, Real-time sampling voltage of the transformer secondary rectifier; This is the upper limit threshold of the secondary voltage corresponding to the 110V input in series mode; This is the lower limit threshold of the secondary voltage corresponding to a 220V input in series mode. This is the upper limit threshold of the secondary voltage corresponding to a 110V input in parallel mode. This is the lower limit threshold of the secondary voltage corresponding to a 220V input in parallel mode. When the sampled voltage is within the hysteresis range between the two thresholds of the corresponding mode, the microcontroller main control unit keeps the current transformer winding connection state unchanged. In parallel mode, if Immediately shut down the power output and trigger an overvoltage fault. This is the overvoltage threshold.

3. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 2, characterized in that, The microcontroller main control unit has built-in winding switching electrical interlock timing logic. After the voltage type is determined, it controls the currently working relay to disconnect and maintain a static delay. ; The relay in the target connection mode is activated, and the relay contact level status is synchronously collected for readback verification. If the verification result is consistent with the preset winding topology, the switching is determined to be complete. If the verification is inconsistent, a single switching retry is automatically initiated. If the retry fails, a status fault flag is triggered and stored in the fault register. In this circuit, a hardware interlock is set up in the relay drive circuit: the normally open auxiliary contact of K1 is connected in series in the series drive circuit of K2, and the normally open auxiliary contact of K2 is connected in series in the series drive circuit of K1. This prevents the winding short circuit caused by one relay being engaged and the other being released, and allows the two relays to be engaged or released at the same time.

4. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 1, characterized in that, The microcontroller main control unit has built-in adaptive stepped soft-start control logic with current hysteresis constraint, and sets a dynamic boost step size calculation formula: ,in, This represents the real-time boost step size for the nth step. As the initial reference boost step size, This is the step size decay coefficient. This is the output voltage for the current gear. This is the rated output voltage of the power supply; Simultaneously, the current hysteresis recovery judgment condition is set: ,in, To output instantaneous current, The overcurrent threshold, This is the current hysteresis margin; The microcontroller main control unit gradually increases the output voltage according to the dynamic boost step size, wherein the instantaneous output current is sampled in real time during the boost process. ,when At this time, pause the stepped voltage boost and lock the current voltage level until the desired voltage is met. When the hysteresis recovery condition is met, the boost process for the subsequent gears continues according to the dynamic boost step size.

5. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 1, characterized in that, The microcontroller main control unit is equipped with a pre-established adaptive calculation formula for the sampling period: ,in, For real-time sampling period, As the reference sampling period, This is the load adjustment coefficient. This is the temperature regulation coefficient. This is the ripple adjustment coefficient. This refers to the real-time output power of the power supply. This refers to the rated output power of the power supply. This refers to the real-time temperature of the transformer. The reference temperature is room temperature. The over-temperature protection threshold; To output the RMS value of the ripple; The preset reference ripple value; The ADC sampling interval for voltage, current, and temperature parameters is dynamically adjusted based on the aforementioned adaptive calculation formula for the sampling period.

6. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 5, characterized in that, The microcontroller main control unit is equipped with a pre-established formula for calculating the maximum output power under transformer temperature constraints: ,in, This represents the maximum allowable output power at the current temperature. The power attenuation coefficient; The real-time clamping power supply output power is calculated based on the maximum output power calculation formula.

7. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 1, characterized in that, The microcontroller main control unit acquires the RMS value of the ripple output from the linear voltage regulator module in real time through the ripple sampling hardware branch. ; Will Compared with the preset reference ripple value The real-time ripple deviation is obtained by performing a difference calculation. The threshold range to which the real-time ripple deviation belongs is detected, and the corresponding error amplification gain level is matched and called to complete the coarse adjustment of the graded level. The coarse adjustment gain level has cross-range switching and locking logic. The microcontroller uses fixed sampling points as the timing window to continuously sample the residual ripple deviation after coarse adjustment for multiple cycles, thus constructing a discrete deviation timing sampling sequence. Using a preset baseline ripple value as a zero-deviation baseline, the positive and negative deviation polarities are determined by comparing the sampling sequences. Perform differential operations on adjacent sampled values ​​to calculate the real-time rate of change of the deviation amplitude and its positive or negative polarity; The sampling sequence is divided into low-amplitude, medium-amplitude, and high-amplitude intervals based on the magnitude of the deviation sampling values. Sampling points in the same interval are grouped together, and a linear distribution is determined by piecewise linear fitting using the least squares method with the minimum value of each group as the benchmark. The sampled sequence is divided into secondary discrete clusters according to the positive and negative polarity and magnitude of the deviation change rate to determine the secondary linear distribution; By comparing the first-order and second-order linear distributions, the maximum overlapping layer is determined and the independent overlapping range within the layer is extracted to determine the deviation disturbance attribute. The BJT feedback bias parameter adjustment direction is uniquely matched based on the deviation polarity, and the adjustment rate level is matched based on the deviation amplitude and deviation attribute. Set a constraint on the maximum adjustment step size of the BJT feedback bias fine-tuning in a single operation. After each fine-tuning is completed, resample and update the deviation time series, and dynamically iterate to adjust the adjustment direction and rate. When the residual ripple deviation of multiple consecutive samplings falls within the preset allowable range, the current BJT feedback bias parameter is locked until a sudden load change or environmental temperature disturbance is detected that causes the ripple deviation to exceed the range, at which point the gain coarse adjustment and bias fine adjustment are restarted.

8. The hybrid low-noise power supply system with adaptive input voltage switching and status monitoring according to claim 1, characterized in that, The microcontroller main control unit is also used to classify faults into three levels: early warning type, self-healing minor faults, and non-self-healing severe faults. When it is a warning type fault, the warning information is only reported through the communication interface, and no reset or shutdown operation is triggered. When it is a minor fault, a delay is triggered. It will then automatically reset and restart. When a severe fault occurs, the power output channel is immediately blocked. At this time, when a fault event is triggered, the composite information of the fault event is split and the indivisible fault information atomic units are extracted. The fault information atomic units include a fault level identifier unit, a fault type identifier unit, and an associated operating condition parameter unit. Each fault information atomic unit is mapped to a preset standardized fault coding meta-set to generate a basic coding value corresponding to each atomic unit. The standardized fault coding meta-set contains standardized coding rules corresponding to each fault information atomic unit. The real-time operating parameters corresponding to the associated operating condition parameter unit are quantized. Based on the preset parameter threshold and encoding length constraint, the quantized parameter values ​​are range-shrinked to obtain the target quantized value that adapts to the encoding field length. For the basic coding value corresponding to each fault information atomic unit, corresponding correction factors are introduced and combined to generate the intermediate coding value corresponding to each atomic unit. The correction factors include fault level correction factor, fault type correction factor and operating condition parameter correction factor, which are used to correct the deviation of the basic coding value. The intermediate encoded values ​​are iteratively intersected and checked. If there are field conflicts, the encoded values ​​are adjusted based on the preset priority rules to ensure that there are no conflicts in all fields. The verified intermediate code values ​​are coupled and fused together, and a verification field generated based on all the intermediate code values ​​is added to form a fixed-length hierarchical standardized fault code. The standardized fault code and fault occurrence timestamp are synchronously packaged and uploaded to the system main control SOC through the communication interface module. The system main control SOC restores the fault level, fault type, associated operating condition parameters and verification information from the standardized fault code based on the preset encoding parsing rules.

Images