Plate burning prevention method based on temperature detection

Abstract

The invention provides a board burning prevention method based on temperature detection, and belongs to the field of computers, the board burning prevention method is based on temperature detection, three levels of temperature thresholds are divided according to attributes of board cards and devices, reverse order judgment logic is adopted, E-FUSE current protection measures are combined, the response speed to over-current and over-temperature abnormal conditions is increased, and the board burning prevention method is high in reliability. And active temperature detection three-level anti-burning plate protection is realized.

Description

Technical Field

[0001] The invention relates to the field of computers, and in particular to a method for preventing a burning board based on temperature detection. Background Art

[0002] With the development of new Internet technologies such as cloud computing, AI intelligence, and big data, high requirements are placed on server performance. The power of the CPU and memory is greater, the demand for power supply is more stringent, the dynamic response requirements are higher, and the loss under dynamic conditions is also greater. In the actual application of high-performance and high-power server boards, there are often overloads caused by local overcurrent or short circuits caused by board assembly errors and other board failures. Overcurrent and short circuits will have a serious impact on the normal operation and life of the board. In the most serious cases, the board will burn out and the board and device will be completely scrapped. Summary of the invention

[0003] In view of the problem of board burning caused by untimely passive protection of servers or computers in the event of load overcurrent or short circuit, the present invention provides a method for preventing board burning based on temperature detection.

[0004] The technical solution of the present invention is:

[0005] A method for preventing board burnout based on temperature detection implements an active three-level protection method based on temperature detection. This method is based on temperature detection, divides the temperature threshold into three levels according to the board and device properties, adopts reverse judgment logic, and combines E-FUSE current protection measures to speed up the response to overcurrent and overtemperature abnormalities, and realizes active temperature detection three-level board burnout protection.

[0006] The server active temperature detection three-level protection method has the following main features:

[0007] 1) Based on the board temperature detection, actively detect the temperature of the main position of the board and the temperature of the high-power consumption device in real time;

[0008] 2) The system is divided into three different temperature steps according to the PCB material and semiconductor characteristics of the device. The system takes different actions at different temperature thresholds.

[0009] 3) Adopt the inverted step temperature detection method, first detect and determine whether the highest temperature threshold is reached, speed up the judgment of the danger signal of burning board, and realize the anti-burning protection of boards and devices with the fastest logic;

[0010] 4) If the temperature reaches the third-level temperature threshold, the BMC and other management chips will send shutdown commands to high-power devices to stop them from continuing to work, and send turbo operation instructions for the fans. At the same time, they will send control panel error signals to inform the backend user that the server has failed, and the user will make a judgment on the abnormal operation status of the server.

[0011] 5) If the temperature detection does not reach the third-level temperature threshold but reaches the second-level temperature threshold, the BMC and other management chips send an over-temperature warning signal to the high-power consumption device. After receiving this warning signal, the high-power consumption device will actively reduce the frequency to avoid over-temperature caused by over-current, and then send a fan turbo operation command;

[0012] 6) If only the first-level temperature threshold is detected, the BMC and other management chips send a full-speed cooling fan speed signal (not a turbo speed signal) to speed up the cooling of high-power devices and prevent the temperature rise of high-power devices from burning local and peripheral devices;

[0013] 7) In any temperature detection stage, the E-FUSE and BMC control chip detect the temperature simultaneously and in parallel. As long as the detection current exceeds the set overcurrent or short-circuit current, the E-FUSE will perform protection;

[0014] This implements a three-level protection method for active temperature detection of the server.

[0015] The difference between the design of the present invention and the solution before improvement is:

[0016] Before the improvement, the server's board burn-out protection design used only E-FUSE or passive temperature detection protection. However, the system may not react in time to large stresses such as external factors or short-term overcurrent during failure, which may lead to board burnout.

[0017] After the improvement, the active temperature detection three-level protection method is adopted, and the following functions can be realized:

[0018] 1) Through temperature detection and classification into different levels, the over-current and over-temperature abnormalities of the board can be identified in advance to achieve active anti-burn board protection;

[0019] 2) The system takes different actions at different temperature levels and adopts the reverse temperature step detection method. First, it detects whether the highest temperature threshold is reached, speeds up the judgment of the danger signal of burning the board, and realizes the anti-burning protection of the board and the device with the fastest logic;

[0020] 3) Active detection and protection of the system is achieved through the parallel execution of temperature detection and current monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the structure of the technical solution of the present invention;

[0022] Figure 2 It is a schematic diagram of the workflow of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0024] The active temperature detection three-level protection method of the present invention implements the following working steps: Figure 2 As shown:

[0025] 1) If the server board does not have its own temperature detection function, place a temperature sensor nearby according to the device characteristics. If the device has its own temperature detection function, connect it to the management controller chip such as BMC through protocols such as I2C;

[0026] 2) When the system board is operating normally, the BMC management chip can collect temperature at a set time interval, and compare the collected temperature with the set temperature threshold to determine the temperature status of the system operation;

[0027] 3) During system operation, the detected temperature first determines whether it reaches the third-level temperature threshold; if so, the BMC and other management chips send a shutdown command to the high-power devices to stop them from continuing to work, and send a fan turbo operation command.

[0028] At the same time, an error signal is sent to the human-computer interaction interface to inform the backend user that the server has failed, and the user can judge the abnormal operation status of the server; the turbo operation time of the fan is determined according to the characteristics of the fan, and after the turbo operation ends, it continues to run at the sustainable maximum speed;

[0029] The third temperature threshold is set according to the maximum operating temperature characteristics of the device (for example, 150°C). At this temperature, the device may be damaged quickly, resulting in serious burning of the board or device.

[0030] 4) If the system detects that the temperature has not reached the third temperature threshold but has reached the second temperature threshold, the BMC management chip sends a frequency reduction instruction to the device. After receiving the frequency reduction instruction, the device will reduce the operating frequency, and the power consumption and loss will decrease accordingly. Then the temperature begins to drop, and a warning signal is sent to the user through the human-computer interaction interface, prompting the machine to enter the high power consumption stage; the BMC and other management chips send turbo operation instructions to the fan.

[0031] After turbo operation ends, if the temperature drops to the temperature range after the device MTBF value is derated, the BMC management chip will send the maximum sustainable speed of the fan and continue for a period of time (for example, 10 minutes, determined according to the loss, thermal resistance and other parameters of the high-power device), and then the system will enter normal operation.

[0032] Among them: The second-level temperature threshold is set according to the upper limit of the sustainable operating temperature characteristics of the device, such as 115°C. At this temperature, the system can run for a relatively long time, but it cannot be guaranteed that the MTBF value meets the life requirements of the entire machine at this temperature.

[0033] 5) If the system detects that the temperature threshold has only reached the first level, the BMC and other management chips will send the fan to the maximum sustainable speed. If the device temperature drops back to the temperature after the device MTBF value is reduced after a period of full speed operation (for example, 5 minutes, the full speed operation time needs to be determined based on the loss, thermal resistance and other parameters of the high-power device to ensure that the temperature inside the device does not exceed an unbearable level), the fan will return to normal speed;

[0034] The first-level temperature threshold is set according to the typical characteristic value of the high-power devices and the sustainable operating temperature of the devices on the board (for example, 85°C). This temperature threshold can ensure that the MTBF value of the system and each device is met, and the value is reduced to a certain extent;

[0035] 6) Regardless of whether the device temperature reaches the third temperature threshold for the first time or rises to the third temperature threshold again after the device frequency is reduced, the system enters the third level protection state;

[0036] 7) Regardless of the device temperature being at the second temperature threshold, the system enters the second level protection state;

[0037] 8) In any of the above system working states, if the E-FUSE detects the current overcurrent signal of the device, it will immediately take protective action;

[0038] Through the above logic, the three-level anti-burning board protection method based on temperature detection proposed by the present invention is realized.

[0039] The above description is only a preferred embodiment of the present invention, which is only used to illustrate the technical solution of the present invention, and is not used to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A method for preventing burning of boards based on temperature detection, characterized in that: Based on temperature detection, an active three-level protection method is implemented; based on temperature detection, three-level temperature thresholds are divided according to the board and device properties, and reverse judgment logic is adopted. Combined with E-FUSE current protection measures, the reaction speed to over-current and over-temperature abnormal conditions is accelerated, and active temperature detection three-level anti-burn board protection is realized.

2. The method according to claim 1, characterized in that: Based on the board temperature detection, it actively detects the temperature of each position of the board and the temperature of the device body in real time; The system is divided into three different temperature levels according to the PCB material of the board and the semiconductor characteristics of the device. The system takes different actions at different temperature thresholds.

3. The method according to claim 2, characterized in that The inverted step temperature detection method is adopted to first detect and determine whether the maximum temperature threshold is reached, speed up the judgment of the danger signal of burning board, and realize the anti-burning protection of boards and devices with the fastest logic.

4. The method according to claim 3, characterized in that If the temperature reaches the third-level temperature threshold, the management chip sends a shutdown command to high-power devices to stop them from continuing to work, and sends a turbo fan operation command. At the same time, it sends a control panel error signal to inform the backend user that the server has failed, and the user determines the abnormal operation status of the server. If the temperature detection does not reach the third-level temperature threshold but reaches the second-level temperature threshold, the management chip sends an over-temperature warning signal to the high-power device. After receiving the warning signal, the high-power device will actively reduce the frequency to avoid over-temperature caused by over-current, and then send a fan turbo operation command; If only the first-level temperature threshold is detected, the management chip sends a full-speed cooling fan speed signal instead of a turbo speed signal to speed up the cooling of high-power consumption components.

5. The method according to claim 4, characterized in that At any temperature detection stage, the E-FUSE and the control chip detect the temperature simultaneously and in parallel. As long as the detection current exceeds the set overcurrent or short-circuit current, the E-FUSE will provide protection.

6. The method according to claim 5, characterized in that The server board is based on the device characteristics. If the device does not have its own temperature detection function, a temperature sensor is placed nearby. If the device has its own temperature detection function, it is connected to the management controller chip through the I2C protocol. When the system board is operating normally, the management chip collects temperature at a set time interval, compares the collected temperature with the set temperature threshold, and determines the temperature condition of the system operation.

7. The method according to claim 6, characterized in that The fan turbo operation time is determined by the characteristics of the fan. After the turbo operation ends, the fan will continue to run at the maximum sustainable speed. After turbo operation ends, if the temperature drops to the temperature range after the device MTBF value is derated, the management chip will send the maximum sustainable speed of the fan and continue for a set period of time, and then the system will enter normal operation.

8. The method according to claim 7, characterized in that Regardless of whether the device temperature reaches the third temperature threshold for the first time or rises to the third temperature threshold again after the device frequency is reduced, the system enters the third level protection state; Regardless of the device temperature being at the second temperature threshold under any circumstances, the system enters the second level protection state.

9. The method according to claim 8, characterized in that In any of the above system working states, if the E-FUSE detects an overcurrent signal from the device, it will immediately take protective action.

Images