Temperature monitoring device and server

Abstract

The embodiment of the utility model discloses a temperature monitoring device and a server. The server comprises a board card and a substrate controller. The board card power supply is connected with the corresponding N slots through the N groups of cables and the N power supply connectors. The temperature monitoring device comprises N temperature monitoring circuits, and each temperature monitoring circuit comprises at least one semiconductor device and a temperature chip; the at least one semiconductor device is located at a target position and is in contact with the to-be-monitored cable; the temperature chip is respectively connected with the at least one semiconductor device and the substrate controller; the at least one semiconductor device is used for generating a first electric signal according to the temperature change of the target position and sending the first electric signal to the temperature chip; the temperature chip is used for processing the first electric signal to obtain a second electric signal and sending the second electric signal to the substrate controller; the substrate controller is used for determining the temperature of the to-be-monitored cable according to the second electric signal. The service life of the cable is prolonged, and the performance of the whole server system is ensured.

Description

Technical Field

[0001] This application relates to the field of temperature monitoring technology, and more particularly to a temperature monitoring device and server. Background Technology

[0002] With the rapid development of servers, the performance requirements for server components are increasing, as are the quantity requirements, thus increasing the overall power consumption of servers. These server components can be various types of expansion cards, such as graphics processing units (GPUs) and data processing units (DPUs) and other accelerator cards.

[0003] Currently, server components are typically powered using standard cables. If these cables are not properly inserted, prolonged operation of the server components can lead to loose power connections. The cables are also susceptible to sudden surges in current, causing overcurrent issues. When cables operate under high current for extended periods, it not only accelerates cable aging and shortens their lifespan, but it can also burn out or deform the connectors between the cables and circuit boards, potentially causing the circuit boards to burn out or be damaged, thus affecting the entire server system's usability. Utility Model Content

[0004] This application provides a temperature monitoring device and a server for real-time monitoring of the temperature of cables connecting the slots and the power supply of the circuit board, thereby improving the service life of the cables and ensuring the performance of the entire server system.

[0005] In a first aspect, one embodiment of this application provides a temperature monitoring device applied to a server. The server includes a board and a baseboard controller disposed on the board. The board is also provided with N slots. Each slot is used to insert M expansion cards. The board power supply is connected to the corresponding N slots through N sets of cables and N power connectors. M and N are integers greater than or equal to 1.

[0006] The temperature monitoring device includes N temperature monitoring circuits, and each temperature monitoring circuit includes at least one semiconductor device and a temperature chip;

[0007] At least one semiconductor device is located at a target position and is in contact with the cable to be monitored; wherein, the target position is a position on the cable to be monitored near the power connector; the cable to be monitored is any one of N groups of cables;

[0008] The temperature chip is mounted on the board and is connected to at least one semiconductor device and a substrate controller.

[0009] At least one semiconductor device is used to generate a first electrical signal based on the temperature change at the target location, and to send the first electrical signal to a temperature chip;

[0010] The temperature chip is used to process the first electrical signal to obtain the second electrical signal, and then send the second electrical signal to the substrate controller.

[0011] The baseboard controller is used to determine the temperature of the cable to be monitored based on the second electrical signal.

[0012] In this embodiment, the temperature monitoring device includes N monitoring circuits. At least one semiconductor device in each temperature monitoring circuit is positioned on the cable to be monitored near the power connector. Utilizing the characteristic that the signal output by the semiconductor device changes with temperature, a temperature chip processes the signal and sends it to the baseboard controller. The baseboard controller then determines the temperature of the cable to be monitored based on the received signal. This design allows for real-time monitoring of the temperature of the cable connecting the slot and the board's power supply, improving cable lifespan and ensuring the performance of the entire server system.

[0013] In one alternative implementation, the temperature monitoring circuit includes two semiconductor devices; the semiconductor devices are transistors.

[0014] The first terminal of the first transistor is connected to the first temperature monitoring pin of the temperature chip, the second terminal of the first transistor is connected to the second temperature monitoring pin of the temperature chip, and the third terminal of the first transistor is connected to the second temperature monitoring pin through the first capacitor.

[0015] The first terminal of the second transistor is connected to the third temperature monitoring pin of the temperature chip, and the second terminal of the second transistor is connected to the fourth temperature monitoring pin of the temperature chip; the third terminal of the second transistor is connected to the fourth temperature monitoring pin through the second capacitor.

[0016] In the above embodiment, since each group of cables has a certain width, in order to monitor the temperature more accurately and avoid local temperatures being too high or too low, two transistors can be set in the temperature monitoring circuit. Each transistor is connected to the temperature chip and outputs a corresponding signal so that the substrate controller can obtain a more accurate temperature.

[0017] In one optional implementation, the power supply pin of the temperature chip is connected to the first power supply terminal of the board, and the power supply pin of the temperature chip is connected to the first terminal of the third capacitor and the first terminal of the fourth capacitor respectively; the second terminal of the third capacitor and the second terminal of the fourth capacitor are both grounded.

[0018] The alarm output pin of the temperature chip is connected to the first end of the first resistor, and the second end of the first resistor is connected to the first power supply terminal.

[0019] The open-drain output and bus address selection multiplexed pin of the temperature chip are connected to the first end of the second resistor, and the second end of the second resistor is connected to the first power supply terminal.

[0020] In the above embodiment, the first power supply terminal is the power supply for the temperature chip. In the connection path between the temperature chip and the first power supply terminal, voltage division can be performed by the first and second resistors, and filtering can be performed by the third and fourth capacitors. Furthermore, the alarm output pin, the open-drain output, and the bus address selection multiplexed pin are all effectively connected to ensure that an alarm is triggered when the temperature is abnormal.

[0021] In one alternative implementation, the target location includes at least two regions, with a first transistor located in the first region of the at least two regions and a second transistor located in the second region of the at least two regions.

[0022] In the above embodiment, placing the two transistors in different areas of the target location can avoid errors caused by excessively high or low local temperatures, and further improve the accuracy of temperature monitoring.

[0023] In one alternative implementation, the baseboard controller is further configured to:

[0024] The first temperature is obtained by converting the first sub-signal in the second electrical signal, and the second temperature is obtained by converting the second sub-signal in the second electrical signal; wherein the first sub-signal is obtained based on the sub-signal output by the first transistor, and the second sub-signal is obtained based on the sub-signal output by the second transistor.

[0025] The smaller of the first and second temperatures is determined as the current temperature of the cable to be monitored.

[0026] In the above embodiment, when two transistors are placed in different positions and two temperatures are obtained, the smaller temperature can be selected as the current temperature of the cable to be monitored, thus ensuring the safety of the cable to be monitored.

[0027] In one alternative implementation, the baseboard controller is further configured to:

[0028] Determine P temperatures of the cable to be monitored within a set time range; where P is an integer greater than or equal to 3.

[0029] Remove Q out of P abnormal temperatures; where Q is an integer less than P.

[0030] Based on the remaining PQ temperatures, determine the temperature of the cable to be monitored within the set time range.

[0031] The above embodiment determines the temperature within a set time range to be monitored based on the temperature remaining after removing abnormal temperatures, which can achieve de-shaking processing and avoid the influence of abnormal temperatures on the final monitored temperature.

[0032] In one alternative implementation, the temperature monitoring device further includes a switching chip;

[0033] The temperature chip in each temperature monitoring circuit is connected to the bus pin of the baseboard controller via the bus pin of the switching chip.

[0034] In the above embodiments, when there are multiple cables to be monitored, the temperature monitoring device may include multiple temperature monitoring circuits, and the temperature monitoring chip in each temperature monitoring circuit is connected to the baseboard controller through a switching chip, so that the baseboard controller can monitor the temperature of each group of cables in real time.

[0035] In one alternative implementation,

[0036] The system management bus clock pin and system management bus data pin of the first temperature chip are respectively connected to the first bus pin of the switch chip;

[0037] The system management bus clock pin and system management bus data pin of the second temperature chip are respectively connected to the second bus pin of the switching chip;

[0038] The first temperature chip is the temperature chip in the first temperature monitoring circuit, the second temperature chip is the temperature chip in the second temperature monitoring circuit, and the first temperature monitoring circuit and the second temperature monitoring circuit are any two monitoring circuits among N monitoring circuits.

[0039] In the above embodiments, when the temperature monitoring device includes multiple temperature monitoring circuits, there are multiple temperature chips. In this case, some temperature chips can be connected to the first bus pin of the switching chip, and other temperature chips can be connected to the second bus pin of the switching chip. This avoids the data transmission delay problem caused by connecting all chips to the same bus pin of the switching chip, and also avoids excessive use of the switching chip's bus pins affecting other functions of the switching chip, thus making reasonable use of the switching chip's bus pins.

[0040] In one alternative implementation,

[0041] The system management bus clock pin of the first temperature chip is connected to the first end of the third resistor, and the second end of the third resistor is connected to the first bus pin of the switch chip.

[0042] The system management bus data pin of the first temperature chip is connected to the first end of the fourth resistor, and the second end of the fourth resistor is connected to the first bus pin of the switch chip.

[0043] In the above embodiment, the system management bus clock pin and system management bus data pin of the first temperature chip are connected to the switching chip through resistors, thereby ensuring the stability of the temperature monitoring device.

[0044] Secondly, one embodiment of this application provides a server, including a board and a baseboard controller disposed on the board; the board is also provided with N slots; each slot is used to insert M expansion cards; the board power supply is connected to the corresponding N slots through N sets of cables and N power connectors; M and N are integers greater than or equal to 1;

[0045] The temperature monitoring device of the first aspect includes N temperature monitoring circuits, each temperature monitoring circuit including at least one semiconductor device and a temperature chip; each temperature monitoring circuit is used to measure the temperature at a target location of a cable to be monitored, the target location being a position on the cable to be monitored near the power connector. Attached Figure Description

[0046] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 An application scenario diagram of a server system provided in one embodiment of this application;

[0048] Figure 2 This is a schematic diagram illustrating monitoring at different locations according to an embodiment of this application;

[0049] Figure 3 This is a schematic diagram of the structure of a temperature monitoring device provided in one embodiment of this application;

[0050] Figure 4 A circuit diagram of a temperature monitoring circuit provided in one embodiment of this application;

[0051] Figure 5 This is a schematic diagram of different monitoring areas provided in one embodiment of this application;

[0052] Figure 6 A schematic diagram illustrating the connection relationship between a temperature monitoring device and a baseboard controller, provided in an embodiment of this application;

[0053] Figure 7 This is a schematic diagram illustrating the connection relationship between a temperature chip and a transistor, provided as an embodiment of this application. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0055] For ease of understanding, the terms used in the embodiments of this application are explained below:

[0056] (1) A PCIe slot (Peripheral Component Interconnect Express Slot, PCIeSlot) is an interface on a computer motherboard used to connect expansion cards (such as graphics cards, network cards, sound cards, memory cards, etc.). It is currently the mainstream hardware expansion interface standard, used to transfer data between the motherboard and expansion cards, and has advantages such as high speed, low latency, and support for hot-swapping. In the embodiments of this application, the slot refers to a PCIe slot.

[0057] (2) The meaning of standard capacitor parameters, taking 2200pF 50V X7R 0402 as an example:

[0058] Capacitance value: 2200pF (i.e., 2200 picofarads, which can also be expressed as 2.2nF or 0.0022μF);

[0059] Withstand voltage: 50V (the maximum voltage the capacitor can withstand);

[0060] Temperature characteristics: X7R (indicates the temperature coefficient of the capacitor, with a value change of ±15% in the range of -55℃ to +125℃. It belongs to Class II ceramic capacitors, with stable performance and is suitable for filtering, coupling and other scenarios).

[0061] Package size: 0402 (imperial code, indicating a length of 0.04 inches and a width of 0.02 inches, approximately 1.0mm × 0.5mm, belonging to small surface mount capacitors).

[0062] (3) The meaning of standard resistor parameters, taking 4.7KR1%0402 as an example:

[0063] Resistance value: 4.7K indicates that the resistance is 4.7 kiloohms (kΩ), or 4700 ohms;

[0064] Accuracy: R1%, indicates that the accuracy of this resistor is ±1%, that is, the deviation between the actual resistance value and the nominal resistance value of 4.7KΩ is within ±1%.

[0065] Package size: 0402 indicates the package size of the resistor. 0402 is an imperial size code, representing that the resistor is 0.04 inches long and 0.02 inches wide, which is approximately 1.0 mm long and 0.5 mm wide in metric units.

[0066] The number of any elements in the accompanying drawings is for illustrative purposes only and not as a limitation, and any naming is for distinction only and has no limiting meaning.

[0067] In practice, reliability design has always been a key consideration in server design. Currently, most server components are still powered by connecting to the power supply of the circuit board via ordinary cables. If these cables are not properly inserted, prolonged operation can lead to a loose power connection, making the cable susceptible to sudden high current surges and overcurrent issues. Furthermore, when the ambient temperature is high, prolonged operation under high current conditions accelerates cable aging and shortens its lifespan.

[0068] On GPU-based systems, there have been instances where, after prolonged testing and power-on, the connector area near the power cable contact point has turned black and slightly charred. In severe cases, the entire cable or connector may burn out or deform, potentially leading to the burning and damage of the circuit board, affecting the use and operation of the entire server system.

[0069] Therefore, to detect the risk of cable burnout due to overheating or exceeding temperature limits, a temperature monitoring device can be designed. This application provides a temperature monitoring device that can monitor the temperature of cables connecting the power supply and slots of the circuit board at a target location (near the power connector, where temperatures are prone to overheating). The device then outputs this temperature in real time through the baseboard controller, allowing users to be informed promptly and take appropriate measures when the temperature exceeds a threshold. This design not only improves cable lifespan but also ensures the overall safety of the server system to a certain extent.

[0070] After introducing the design concept of the embodiments of this application, the following is a brief introduction to the application scenarios to which the technical solutions of the embodiments of this application can be applied. It should be noted that the application scenarios described below are only for illustrating the embodiments of this application and are not intended to limit the scope. In specific implementation, the technical solutions provided by the embodiments of this application can be flexibly applied according to actual needs.

[0071] refer to Figure 1 This is an application scenario diagram of a server system provided in an embodiment of this application. Figure 1 In this example, the server is configured to support eight 4U (server chassis height) dual-width GPU cards. Since each dual-width GPU card requires a maximum power consumption of 300W to 350W, and each PCIe slot provides a maximum of 75W, each dual-width GPU card requires an external cable (hereinafter referred to as the cable) for power supply. This example includes eight sets of cables, each set connected to the slot on one side via a power connector and to the board's power supply on the other side. Each set of cables... Figure 1 It appears as a single cable, but is actually composed of multiple cables. For example, Figure 1Cable group 11 is a cable group consisting of multiple cables.

[0072] The following section explains the key monitoring locations within each cable group:

[0073] In practical applications, the temperature at different locations on each group of cables can be determined by measuring with a heat gun. Figure 2 This is a schematic diagram illustrating monitoring at different locations, provided as an embodiment of this application. In this example, the area near the power connector on the cable is subdivided into three positions (position 0, position 1, and position 2). The temperature of each cable at position 0, position 1, position 2, and the housing position is measured at two different times. Additionally, it should be noted that... Figure 2 Only one set of cables is shown, which includes 6 cables.

[0074] Table 1 shows the temperature at different locations. As can be seen from Table 1, the highest temperature in each cable group is at the contact point with the power connector. Furthermore, numerous cases of similar blackening and burning issues indicate that the blackening of power cables consistently occurs at this location. Therefore, this location can be identified as a key area for monitoring (e.g., ...). Figure 1 (Position 111 in the middle).

[0075] Table 1 shows the temperatures of the two sets of cables at different times.

[0076]

[0077] As shown in Table 1, for cable group 1, the temperature at position 0 is the highest in data 1 at the first moment; the temperature at position 0 is also the highest in data 2 at the second moment. For cable group 2, the temperature at position 1 is the highest in data 1 at the first moment; the temperature at position 1 is also the highest in data 2 at the second moment. In other words, the temperatures at positions 0, 1, 2, and the housing are all relatively high.

[0078] Therefore, the data in Table 1 further confirms that the location on the cable near the power supply and the machine is the key area that needs to be monitored.

[0079] Of course, the methods provided in the embodiments of this application are not limited to... Figure 1 The application scenarios shown can also be used in other possible scenarios, and this application does not impose any limitations. Figure 1 The functions that each device in the application scenario shown can achieve will be described in subsequent method embodiments, and will not be elaborated on here.

[0080] To further illustrate the technical solutions provided in the embodiments of this application, a detailed description is provided below in conjunction with the accompanying drawings and specific implementation methods. Although the embodiments of this application provide method operation steps as shown in the following embodiments or drawings, the method may include more or fewer operation steps based on conventional or non-inventive methods. In steps where there is no logically necessary causal relationship, the execution order of these steps is not limited to the execution order provided in the embodiments of this application.

[0081] The following is combined Figure 1 and Figure 2 The application scenarios shown illustrate the technical solutions provided in the embodiments of this application.

[0082] refer to Figure 3 This application provides a schematic diagram of a temperature monitoring device applied to a server. The server includes a circuit board and a baseboard controller mounted on the circuit board. The circuit board also has N slots, each for inserting M expansion cards (this embodiment illustrates that each slot can insert one expansion card). The circuit board's power supply is connected to the corresponding N slots via N sets of cables and N power connectors, where M and N are integers greater than or equal to 1.

[0083] The temperature monitoring device includes N temperature monitoring circuits, each including at least one semiconductor device and a temperature chip. At least one semiconductor device is located at a target position and in contact with the cable to be monitored; this target position is a location on the cable near the power connector. The temperature chip is mounted on a board and is connected to both the at least one semiconductor device and a baseboard controller. The cable to be monitored can be any one of the N groups of cables; the temperature monitoring method is the same for the other groups.

[0084] At least one semiconductor device can generate a first electrical signal based on the temperature change at the target location and send the first electrical signal to a temperature chip. The temperature chip processes the first electrical signal to obtain a second electrical signal and sends the second electrical signal to a substrate controller. The substrate controller determines the temperature of the cable to be monitored based on the second electrical signal.

[0085] In this embodiment, the temperature monitoring device includes N monitoring circuits. At least one semiconductor device in each temperature monitoring circuit is positioned on the cable to be monitored near the power connector. Utilizing the characteristic that the signal output by the semiconductor device changes with temperature, a temperature chip processes the signal and sends it to the baseboard controller. The baseboard controller then determines the temperature of the cable to be monitored based on the received signal. This design allows for real-time monitoring of the temperature of the cable connecting the slot and the board's power supply, improving cable lifespan and ensuring the performance of the entire server system.

[0086] Figure 4 A circuit diagram of a temperature monitoring circuit provided in an embodiment of this application is shown below. Figure 4 The structure and connection relationships of each temperature monitoring circuit are explained:

[0087] To prevent localized overheating or underheating, each temperature monitoring circuit can include two semiconductor devices, each positioned at the target location. Each semiconductor device is a transistor, referred to as Transistor 1 and Transistor 2. These transistors can be, for example, NPN transistors, such as the IMPT3904LT1G. This allows for temperature monitoring of the cable by utilizing the PN junction characteristics of the NPN transistor (where forward voltage is linearly related to temperature).

[0088] Optionally, the temperature chip can be an EMC1413 chip. The EMC1413 chip has two ports (DP and DN) connected to an NPN transistor for receiving and processing temperature data. Specifically, the first terminal (collector) of the first transistor Q117 is connected to the first temperature monitoring pin DP1 of the temperature chip, and the second terminal (emitter) of the first transistor Q117 is connected to the second temperature monitoring pin DN1 of the temperature chip. The third terminal (base) of the first transistor Q117 is connected to the second temperature monitoring pin DN1 through a first capacitor C1116. The first terminal (collector) of the second transistor Q118 is connected to the third temperature monitoring pin DP2 of the temperature chip, and the second terminal (emitter) of the second transistor Q118 is connected to the fourth temperature monitoring pin DN2 of the temperature chip. The third terminal (base) of the second transistor Q118 is connected to the fourth temperature monitoring pin DN2 through a second capacitor C1117.

[0089] In a specific example, the parameters of the first capacitor C1116 and the second capacitor C1117 can be 2200pF 50VX7R 0402.

[0090] Because each cable group has a certain width in this design, to achieve more accurate temperature monitoring and avoid localized overheating or underheating, two transistors can be incorporated into the temperature monitoring circuit. Each transistor is connected to a temperature chip and outputs a corresponding signal, allowing the baseboard controller to obtain a more accurate temperature reading.

[0091] See also Figure 4The power supply pin VCC of the temperature chip is connected to the first power supply terminal P3V3_STBY of the board. The power supply pin VCC of the temperature chip is connected to the first terminal of the third capacitor C1120 and the first terminal of the fourth capacitor C1122 respectively. The second terminals of the third capacitor C1120 and the fourth capacitor C1122 are both grounded. The alarm output pin ALERT_N of the temperature chip is connected to the first terminal of the first resistor R278. The second terminal of the first resistor R278 is connected to the first power supply terminal P3V3_STBY. The open-drain output and bus address selection multiplexing pin THERM_N / ADDR of the temperature chip is connected to the first terminal of the second resistor R279. The second terminal of the second resistor R279 is connected to the first power supply terminal P3V3_STBY.

[0092] In a specific example, the parameters of the third capacitor C1120 and the fourth capacitor C1122 are 1uF 16V X5R0402. The parameters of the first resistor R278 and the second resistor R279 are 4.7K R1%0402.

[0093] In this design, since the first power supply terminal powers the temperature chip, the connection path between the temperature chip and the first power supply terminal can be divided by the first and second resistors, and filtered by the third and fourth capacitors. Furthermore, the alarm output pin, the open-drain output, and the bus address selection multiplexed pin are all effectively connected to ensure that an alarm is triggered in case of abnormal temperature.

[0094] In addition, since the target location includes at least two regions, in order to ensure that the temperature at each location of the cable is monitored, the first transistor Q117 can be located in the first region of the at least two regions, and the second transistor Q118 can be located in the second region of the at least two regions.

[0095] Figure 5 This is a schematic diagram of different monitoring areas provided in an embodiment of this application, wherein A is the first area and V is the second area.

[0096] This design, which places the two transistors in different areas of the target location, can avoid errors caused by excessively high or low local temperatures, and further improve the accuracy of temperature monitoring.

[0097] In practical applications, a single board may contain multiple sets of cables, which requires multiple temperature monitoring circuits. If the temperature monitoring device includes multiple temperature monitoring circuits, each temperature chip is connected to two transistors and then to the baseboard controller.

[0098] In this case, the temperature monitoring device also includes a switching chip, such as a PCA945 chip, and the temperature chip in each temperature monitoring circuit is connected to the bus pin of the baseboard controller through the bus pin of the switching chip.

[0099] With this design, when there are multiple cables to be monitored, the temperature monitoring device can include multiple temperature monitoring circuits, and the temperature monitoring chip in each temperature monitoring circuit can be connected to the baseboard controller through a switching chip, so that the baseboard controller can monitor the temperature of each group of cables in real time.

[0100] Figure 6 This diagram illustrates the connection relationship between a temperature monitoring device and a baseboard controller, as provided in an embodiment of this application. Some temperature chips are connected to the baseboard controller via bus pins of a switching chip, while other temperature chips are connected via other bus pins of the switching chip. Taking a first temperature chip and a second temperature chip as examples, the first temperature chip is the temperature chip in a first temperature monitoring circuit, and the second temperature chip is the temperature chip in a second temperature monitoring circuit. The first temperature monitoring circuit and the second temperature monitoring circuit are any two monitoring circuits from N monitoring circuits.

[0101] Among them, the system management bus clock pin SMCLK and system management bus data pin SMDATA of the first temperature chip are respectively connected to the first bus pin I of the switch chip. 2 C_0 connection; the system management bus clock pin SMCLK and system management bus data pin SMDATA of the second temperature chip are respectively connected to the second bus pin I of the switching chip. 2 C_1 connection.

[0102] The basic controller includes multiple bus pins. In this embodiment, the I / O pins of the PCA945 switching chip can be used. 2 The C_IN pin is related to the I-pin of the baseboard controller. 2 The connection of pin C_13 is only an example and does not constitute a specific limitation.

[0103] For example, when a temperature monitoring device includes eight temperature monitoring circuits, that is, eight temperature chips, the PCA945 switching chip includes four bus pins, each connected via the first bus pin I. 2 C_0, Second Bus Pin I 2 C_1, Third bus pin I 2 C_2 and the fourth bus pin I 2C_3 indicates that, in practice, the eight temperature chips can be appropriately distributed among these four bus pins to avoid the latency issues caused by connecting them all to the same bus pin simultaneously, and also to avoid the situation where all four bus pins are occupied, leaving no bus pins available for other circuit modules.

[0104] This design allows for the connection of some temperature monitoring chips to the first bus pin of the switching chip, and other temperature chips to the second bus pin, when the temperature monitoring device includes multiple temperature monitoring circuits (i.e., multiple temperature chips). This avoids the data transmission delay issues caused by connecting all chips to the same bus pin of the switching chip, and also prevents excessive use of the switching chip's bus pins from interfering with other functions, thus making efficient use of the switching chip's bus pins.

[0105] See also Figure 4 The system management bus clock pin SMB_PCA9546_0_SCL of the first temperature chip is connected to the first terminal of the third resistor R305, and the second terminal of the third resistor R305 is connected to the first bus pin I of the switch chip. 2 C_0 connection; the system management bus data pin SMB_PCA9546_0_SDA of the first temperature chip is connected to the first terminal of the fourth resistor R299, and the second terminal of the fourth resistor R299 is connected to the first bus pin I of the switch chip. 2 C_0 connection.

[0106] In a specific example, the parameters of the third resistor R305 are 0R1%0402, and the parameters of the fourth resistor R299 are 33R1%0402. The third resistor R305 is a 0Ω resistor, which may melt to protect the circuit when the current is too large.

[0107] This design connects the system management bus clock pin and system management bus data pin of the first temperature chip to the switching chip via resistors, ensuring the stability of the temperature monitoring device.

[0108] Figure 7 This is a schematic diagram illustrating the connection relationship between a temperature chip and a transistor, provided in an embodiment of this application. Figure 7 The diagram uses only three EMC1413 chips, each connected to two transistors. Thus, the number of monitoring circuits can be configured to match the number of cable groups to be monitored, meaning a corresponding number of temperature chips and twice the number of transistors are required.

[0109] Based on the above technical solution, the substrate controller can generate a first electrical signal according to the temperature change at the target location and send the first electrical signal to the temperature chip. The temperature chip can process the first electrical signal to obtain a second electrical signal and send the second electrical signal to the substrate controller. The substrate controller is used to determine the temperature of the cable to be monitored based on the second electrical signal. If there are two transistors, both the first and second electrical signals include two sub-signals. For example, taking the second signal as an example, the first transistor corresponds to the first sub-signal of the second signal, and the second transistor corresponds to the second sub-signal of the second signal.

[0110] In related technologies, the baseboard controller issues an over-temperature alarm when the monitored temperature exceeds a set temperature threshold. However, during actual testing, it was found that due to external and other factors, directly comparing the real-time monitored temperature with the threshold for alarm may result in false alarms. Therefore, for each cable group, the temperatures at the two measuring points are de-jittered to prevent false alarms from causing errors in the baseboard controller. The specific de-jittering method is as follows:

[0111] First, select the one with the lower temperature from the two temperature monitoring points.

[0112] When each temperature monitoring circuit includes two diodes, the first signal and the second signal each include two sub-signals. The sub-signal output from the first transistor is processed by the temperature chip to obtain the first sub-signal in the second signal, and the sub-signal output from the second transistor is processed by the temperature chip to obtain the second sub-signal in the second signal. The baseboard controller converts the first sub-signal in the second electrical signal to obtain the first temperature, and converts the second sub-signal in the second electrical signal to obtain the second temperature; the smaller of the first temperature and the second temperature is determined as the current temperature of the cable to be monitored.

[0113] As shown above, the signal output by each diode of the baseboard controller can be converted into temperature. When determining the current temperature of the cable to be monitored, the smaller of the two temperatures can be selected as the current temperature, which can prevent the baseboard controller from falsely reporting the temperature.

[0114] This design allows for the selection of the smaller temperature as the current temperature of the cable under monitoring when the two transistors are positioned in different locations, thus ensuring the safety of the cable.

[0115] Second, monitor the temperature over a period of time and then process it.

[0116] A1: Determine the P temperatures of the cable to be monitored within the set time range.

[0117] Where P is an integer greater than or equal to 3. For example, the GPU cable temperature is read once every 10 seconds.

[0118] A2: Remove Q abnormal temperatures from P temperatures.

[0119] Where Q is an integer less than P. For example, an abnormal temperature could be the highest and lowest of P temperatures.

[0120] A3: Based on the remaining PQ temperatures, determine the temperature of the cable to be monitored within the set time range.

[0121] Optionally, the average of the remaining PQ temperatures can be used as the temperature of the cable under monitoring within a set time range. This minimizes the risk of false alarms caused by sudden temperature changes due to external factors or changes in the position of the NPN transistor, thus avoiding disruptions to server maintenance.

[0122] This design determines the temperature within a set time range to be monitored based on the temperature remaining after removing abnormal temperatures, thus achieving de-shaking and avoiding the impact of abnormal temperatures on the final monitored temperature.

[0123] In summary, in this embodiment of the application, the temperature chip EMC1413 uses I... 2 The C signal connects to the baseboard controller, which collects temperature data and processes alarms, thus monitoring the cable temperature. This is particularly useful for monitoring the temperature of cables used in servers, which handle high power consumption, high current, and sudden current surges, thereby improving server reliability. Specifically, a combination of an NPN transistor and an EMC1413 chip is used. The NPN transistor is connected to the board and then to the chip using a low-speed cable, and the NPN transistor is fixed at the target location on the cable. Compared to related technologies, this approach not only provides real-time temperature monitoring but also offers stability and robustness. Furthermore, the signal is less susceptible to external interference, reducing the occurrence of false alarms.

[0124] To make the above technical solution easier to understand, the effectiveness of the temperature monitoring device in this application embodiment will be illustrated by comparing the temperature monitored by the above-mentioned temperature monitoring device with the actual temperature. Specifically, in the temperature monitoring device, each cable contact point is monitored by two NPN transistors. The substrate controller acquires data through the temperature chip and records the temperature and determines the threshold alarm. After actual testing in the heat dissipation laboratory, i.e., comparing the temperature readings from the laboratory temperature detection instrument with those from the substrate controller, the monitored temperature and the actual temperature are shown in Table 2. As can be seen from Table 2, the temperature difference is very small, which can achieve the function of real-time and accurate temperature monitoring.

[0125] Table 2. Comparison data between monitored temperature and actual temperature.

[0126]

[0127]

[0128] In summary, the embodiments of this application can perform real-time monitoring of high-current, high-power cables. Specifically, it can effectively monitor abnormal current flow caused by cable aging, cable overheating due to high ambient temperatures, and localized overheating or even burnout caused by poor cable connections. Furthermore, it can effectively monitor abnormal temperature fluctuations and false alarms caused by external influences during long-term temperature readings. Additionally, the debouncing method used during baseboard controller processing effectively avoids false triggers and prevents disruption to server operational efficiency in temperature monitoring.

[0129] It should be noted that the model numbers and parameter values ​​of the components in the embodiments of this application are just examples and do not constitute specific limitations, but any component that can achieve the corresponding function of the embodiments of this application is acceptable.

[0130] The above embodiments are only used to provide a detailed description of the technical solutions of this application. However, the description of the above embodiments is only for the purpose of helping to understand the methods of the embodiments of this application and should not be construed as a limitation on the embodiments of this application. Any changes or substitutions that can be easily conceived by those skilled in the art should be covered within the protection scope of the embodiments of this application.

Claims

1. A temperature monitoring device, characterized in that, The application is in a server, the server including a board and a baseboard controller disposed on the board; the board is also provided with N slots; each slot is used to insert M expansion cards; the board power supply is connected to the corresponding N slots through N sets of cables and N power connectors; M and N are integers greater than or equal to 1; The temperature monitoring device includes N temperature monitoring circuits, and each temperature monitoring circuit includes at least one semiconductor device and a temperature chip. The at least one semiconductor device is located at a target position and is in contact with the cable to be monitored; wherein, the target position is a position on the cable to be monitored near the power connector; the cable to be monitored is any one of the N groups of cables; The temperature chip is disposed on the board and is connected to the at least one semiconductor device and the substrate controller, respectively. The at least one semiconductor device is used to generate a first electrical signal based on the temperature change at the target location, and send the first electrical signal to the temperature chip; The temperature chip is used to process the first electrical signal to obtain a second electrical signal, and then send the second electrical signal to the substrate controller. The baseboard controller is used to determine the temperature of the cable to be monitored based on the second electrical signal.

2. The temperature monitoring device according to claim 1, characterized in that, The temperature monitoring circuit includes two semiconductor devices; the semiconductor devices are transistors. The first terminal of the first transistor is connected to the first temperature monitoring pin of the temperature chip, and the second terminal of the first transistor is connected to the second temperature monitoring pin of the temperature chip; the third terminal of the first transistor is connected to the second temperature monitoring pin through the first capacitor. The first end of the second transistor is connected to the third temperature monitoring pin of the temperature chip, and the second end of the second transistor is connected to the fourth temperature monitoring pin of the temperature chip; the third end of the second transistor is connected to the fourth temperature monitoring pin through the second capacitor.

3. The temperature monitoring device according to claim 1 or 2, characterized in that, The power supply pin of the temperature chip is connected to the first power supply terminal of the board, and the power supply pin of the temperature chip is connected to the first terminal of the third capacitor and the first terminal of the fourth capacitor respectively; the second terminal of the third capacitor and the second terminal of the fourth capacitor are both grounded. The alarm output pin of the temperature chip is connected to the first end of the first resistor, and the second end of the first resistor is connected to the first power supply terminal. The open-drain output and bus address selection multiplexing pin of the temperature chip are connected to the first end of the second resistor, and the second end of the second resistor is connected to the first power supply terminal.

4. The temperature monitoring device according to claim 2, characterized in that, The target location includes at least two regions, with the first transistor located in the first region of the at least two regions and the second transistor located in the second region of the at least two regions.

5. The temperature monitoring device according to claim 2, characterized in that, The baseboard controller is also used for: The first temperature is obtained by converting the first sub-signal in the second electrical signal, and the second temperature is obtained by converting the second sub-signal in the second electrical signal; wherein the first sub-signal is obtained based on the sub-signal output by the first transistor, and the second sub-signal is obtained based on the sub-signal output by the second transistor. The smaller of the first temperature and the second temperature is determined as the current temperature of the cable to be monitored.

6. The temperature monitoring device according to claim 1, characterized in that, The baseboard controller is also used for: Determine P temperatures of the cable to be monitored within a set time range; wherein P is an integer greater than or equal to 3; Remove Q abnormal temperatures from the P temperatures; where Q is an integer less than P. Based on the remaining PQ temperatures, the temperature of the cable to be monitored within the set time range is determined.

7. The temperature monitoring device according to any one of claims 1 to 2 and 4 to 6, characterized in that, The temperature monitoring device also includes a switching chip; The temperature chip in each temperature monitoring circuit is connected to the bus pin of the baseboard controller via the bus pin of the switch chip.

8. The temperature monitoring device according to claim 7, characterized in that, The system management bus clock pin and system management bus data pin of the first temperature chip are respectively connected to the bus pins of the switch chip; The system management bus clock pin and system management bus data pin of the second temperature chip are respectively connected to the second bus pin of the switch chip; Wherein, the first temperature chip is a temperature chip in the first temperature monitoring circuit, the second temperature chip is a temperature chip in the second temperature monitoring circuit, and the first temperature monitoring circuit and the second temperature monitoring circuit are any two monitoring circuits among the N temperature monitoring circuits.

9. The temperature monitoring device according to claim 8, characterized in that, The system management bus clock pin of the first temperature chip is connected to the first end of the third resistor, and the second end of the third resistor is connected to the first bus pin of the switch chip. The system management bus data pin of the first temperature chip is connected to the first end of the fourth resistor, and the second end of the fourth resistor is connected to the first bus pin of the switch chip.

10. A server, characterized in that, The system includes a circuit board and a baseboard controller mounted on the circuit board; the circuit board also has N slots; each slot is used to insert M expansion cards; the circuit board's power supply is connected to the corresponding N slots via N sets of cables and N power connectors; M and N are integers greater than or equal to 1. The temperature monitoring device according to any one of claims 1 to 9 includes N temperature monitoring circuits, each temperature monitoring circuit including at least one semiconductor device and a temperature chip; each temperature monitoring circuit is used to measure the temperature at a target location of a cable to be monitored, the target location being a position on the cable to be monitored near the power connector.

Images