GaN power device thermal path state evaluation method based on frequency response
By designing a GaN power device test circuit based on frequency response, and evaluating the fatigue state of the thermal path of the GaN power device using AC power loss, the accuracy and efficiency of thermal path state evaluation in the prior art are solved, and efficient and economical are achieved thermal path status monitoring.
Patent Information
- Application Number
- CN202411424609.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-12
- Publication Date
- 2025-02-07
AI Technical Summary
The existing thermal path state evaluation methods of GaN power devices are difficult to accurately reflect the internal thermal path state of the device, and the calculation is complex, time-consuming and costly.
A GaN power device test circuit based on frequency response is designed to generate AC power loss for GaN power devices to be tested. By comparing the phase frequency characteristic curves before and after the experiment, the fatigue position and degree of thermal path are evaluated.
It realizes a simple and effective evaluation of the fatigue position and degree of the thermal path of GaN power devices without damaging the device packaging, reduces operation and maintenance costs, and improves the accuracy and efficiency of evaluation.
Smart Images
Figure CN119395491A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of health status assessment of GaN power devices, and in particular to a method for evaluating the thermal path status of a GaN power device based on frequency response. Background Art
[0002] Power electronic devices have been widely used in the fields of renewable energy power generation, traffic traction, aerospace, industrial automation, transportation, etc. Improving the reliability of power electronic devices can meet more stringent safety and cost requirements. Wide bandgap semiconductor materials are the third generation of new semiconductor materials that have developed rapidly after the first and second generation semiconductor materials represented by silicon (Si) and gallium arsenide (GaAs). Gallium nitride (GaN) is one of the wide bandgap semiconductor materials. It has outstanding performance advantages such as high voltage resistance, low on-resistance, fast switching speed, high switching frequency, low junction-shell thermal resistance and higher junction temperature. It is more suitable for power devices with voltage levels below 1kV. However, complex and harsh working conditions and alternating thermal-mechanical stresses make GaN power devices prone to fatigue damage. Unforeseen failures caused by fatigue of GaN power devices will bring expensive costs. Health status analysis of GaN power devices has also become an economical and effective way to improve the reliability of power electronic systems.
[0003] The existing methods for evaluating the thermal path status of GaN power devices are mainly characterized by Foster thermal networks and Cauer thermal networks, and thermal imaging technology is also used as a monitoring and evaluation method. The Foster thermal network can be equivalent to the external characteristics of the device's heat transfer structure, and its parameter acquisition and numerical calculation are more convenient. However, it cannot represent the physical nature of the heat transfer of the power device itself, and the thermal network structure is more complex than the Cauer thermal network. For large networks, the calculation of the Foster thermal network may become very complicated, requiring more computing resources and time. The Cauer thermal network characterizes the heat conduction process inside the package from a physical essence. The thermal resistance and thermal capacitance represent the thermal resistance and thermal capacitance values of each layer of structural materials between the junction and shell of the semiconductor device. The thermal resistance and thermal capacitance values have practical significance, but their parameters are difficult to obtain and experimentally verify.
[0004] Some articles have proposed methods for converting between Foster thermal network and Cauer thermal network. Foster thermal network and Cauer thermal network are mostly used for junction temperature prediction, but there are few studies on the impact of high and low operating frequency on the health status and fatigue of GaN power devices. Thermal imaging uses infrared imaging technology to observe the temperature distribution on the surface of the device to evaluate the state of the thermal path. However, thermal imaging is limited by the temperature distribution on the surface of the device and cannot accurately reflect the state of the thermal path inside the device. Summary of the invention
[0005] In order to solve the above technical problems, the present invention provides a method for evaluating the thermal path status of a GaN power device based on frequency response. The method designs a GaN power device test circuit to make the GaN power device to be tested generate AC power loss. During long-term AC power operation, the thermal path fatigue of the GaN power device to be tested will cause the equivalent thermal resistance of the thermal path to change. By comparing the phase-frequency characteristic curve of the GaN power device before the experiment with the phase-frequency characteristic curve after long-term AC power operation, the position and degree of thermal path fatigue of the GaN power device are evaluated. The present invention does not destroy the packaging of the GaN power device, nor does it need to measure the junction temperature of the device by the temperature-sensitive parameter method. It can simply and effectively realize the evaluation of the position and degree of thermal path fatigue of the GaN power device, thereby ensuring the safe operation of the GaN power device and reducing the operation and maintenance cost.
[0006] The technical solution adopted by the present invention is:
[0007] A GaN power device test circuit, comprising:
[0008] A voltage source for supplying power to the power circuit;
[0009] The driving circuit sends out a sinusoidal driving signal and controls the GaN power device under test;
[0010] A sampling resistor, used to collect the AC current flowing through the GaN power device to be tested;
[0011] The GaN power device to be tested generates AC loss, which provides AC excitation for the equivalent Cauer thermal path model.
[0012] The driving circuit includes a microprocessor and a driving chip. The microprocessor is connected to the driving chip. The microprocessor sends a sinusoidal signal through the driving chip to turn on the GaN power device to be tested, and sets the end time of the experiment.
[0013] The sampling resistor and the GaN power device to be tested are connected in series in the power loop of the voltage source, and the amplitude and frequency of the current flowing through the GaN power device to be tested are measured through the sampling resistor; the power loss of the GaN power device to be tested is obtained by multiplying the voltage across the voltage source and the current flowing through the sampling resistor.
[0014] The power P(t) generated by the GaN power device to be tested is used as an excitation signal of the equivalent Cauer thermal path model to provide necessary conditions for the operation of the equivalent Cauer thermal path model. The thermal path fatigue of the GaN power device after long-term operation will cause the thermal resistance R θ Changes, and the shell temperature phase-frequency response curves of Cauer thermal paths under different thermal resistances are compared.
[0015] The shell temperature phase-frequency response curve is obtained by linearizing the Cauer thermal path model equivalent to the GaN power device to be tested, and the shell temperature response varies with the thermal resistance R θ The law of change.
[0016] The method for evaluating the thermal path status of a GaN power device based on frequency response comprises the following steps:
[0017] S1: Build the above-mentioned GaN power device test circuit;
[0018] S2: The amplitude and frequency of the AC current flowing through the sampling resistor are obtained through the GaN power device test circuit built in S1. The power loss of the GaN power device to be tested is obtained by multiplying the voltage across the voltage source and the current flowing through the sampling resistor, and the product is used as the excitation signal of the Cauer thermal path model.
[0019] S3: Build an equivalent Cauer thermal path model of the GaN power device to be tested.
[0020] S4: Implementing linear analysis of the equivalent Cauer thermal path model of the GaN power device to be tested;
[0021] S5: Through the linear analysis of the equivalent Cauer thermal path model of S4, the phase-frequency offset in different frequency intervals and the phase offset in the same frequency interval of the phase-frequency response curve are compared to analyze and evaluate the location and degree of thermal path fatigue of the GaN power device to be tested.
[0022] In S3, the chip layer, substrate layer, thermal interface material, and heat sink of the GaN power device to be tested are each equivalent to a thermal resistance R θ and a heat capacity C θ , so that the four-layer R θ C θ The structure is connected in parallel and R θ C θ The parallel excitation signal at the front end of the parallel structure is obtained by S2, which can build an equivalent Cauer thermal path model of the GaN power device to be tested.
[0023] In S4, the input module is connected to the controllable current source module as the input signal of the equivalent Cauer thermal path model of the GaN power device to be tested, and the connection line between the input module and the controllable current source module is selected as the closed-loop input. The input signal is connected to the fourth-order R θ C θ The model is connected, and the voltage measurement module is connected to the output module as the output signal of the shell temperature measurement, and the connection line between the voltage measurement module and the output module is selected as the open-loop output.
[0024] In S5, the specific location of the thermal path of the GaN power device responds differently to the frequency range. High-frequency signals are likely to affect the chip layer of the GaN power device, medium-frequency signals are likely to affect the substrate layer of the GaN power device, and low-frequency signals are likely to affect the thermal interface material of the GaN power device.
[0025] For a determined GaN power device thermal path position, in the same frequency range, the greater the phase deviation between its phase-frequency characteristic curve and the original phase-frequency characteristic curve and the smaller the phase value, the higher the thermal resistance value corresponding to this layer of material, that is, the higher the fatigue degree of this material layer.
[0026] The present invention provides a method for evaluating the thermal path status of a GaN power device based on frequency response, and the technical effects are as follows:
[0027] 1) The GaN power device test circuit of the present invention calculates the power consumption generated by the GaN power device to be tested in the test circuit by reasonably constructing and using the microprocessor STM32F334, the voltage source, the sampling resistor, and the GaN power device to be tested.
[0028] 2) The GaN power device test circuit of the present invention realizes the function of generating AC power loss when the GaN power device is in operation. The AC power loss provides a frequency-adjustable excitation signal for the equivalent Cauer thermal path model, thereby ensuring the accurate calculation of the Cauer thermal path model and having practical application value.
[0029] 3) Since GaN power devices will experience thermal path fatigue after long-term operation, this will cause the thermal resistance R θ Changes; By changing the frequency range of the Cauer thermal path model excitation AC signal, the fatigue position of the thermal path of the GaN power device can be evaluated; By fixing the frequency range of the Cauer thermal path model excitation AC signal, the fatigue degree of the thermal path of the GaN power device can be evaluated; This method can be simply and effectively implemented without destroying the GaN power device package. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 It is a flow chart of the method of the present invention.
[0031] Figure 2 It is a GaN power device test circuit.
[0032] Figure 3 is the equivalent Cauer thermal path model of GaN power devices with AC excitation.
[0033] Figure 4 It is a schematic diagram of the method of linearization analysis of Cauer thermal path model.
[0034] FIG5( a ) is a shell temperature phase-frequency characteristic curve corresponding to different thermal resistances of different thermal paths of the first type;
[0035] FIG5( b ) is a shell temperature phase-frequency characteristic curve corresponding to different thermal resistances of the first different thermal paths;
[0036] FIG5( c ) is the shell temperature phase-frequency characteristic curve corresponding to different thermal resistances of the third different thermal paths. DETAILED DESCRIPTION
[0037] The method for evaluating the thermal path status of GaN power devices based on frequency response can evaluate the location of the thermal path of GaN power devices and the degree of fatigue of the thermal path. The basic idea of this method is based on the one-dimensional Cauer thermal network of the device, which is regarded as a multi-order passive linear network. The drift of the network parameters is evaluated according to the change of the input-output frequency characteristics, and then the state of the thermal path of the device is evaluated. In the specific implementation of the method of the present invention, the AC power injected into the device is used as the excitation, and the shell temperature of the device is used as the response. Without destroying the device package and extracting the internal parameters of the device, the change of the phase-frequency characteristic curve constructed by the two is used to evaluate the state of the thermal path of the GaN power device under different fatigue states.
[0038] The method designs a GaN power device test circuit to make the GaN power device to be tested generate AC power loss. The GaN power device test circuit includes:
[0039] A voltage source is used to power a power circuit; the power circuit includes a filter capacitor, a GaN power device to be tested, and a sampling resistor;
[0040] The driving circuit sends out a sinusoidal driving signal and controls the GaN power device under test;
[0041] A sampling resistor, used to collect the AC current flowing through the GaN power device to be tested;
[0042] The GaN power device to be tested generates AC loss, which provides AC excitation for the equivalent Cauer thermal path model.
[0043] The driving circuit includes a microprocessor STM32F334 and a driving chip. The microprocessor STM32F334 is connected to the driving chip. The microprocessor STM32F334 sends a sinusoidal signal through the driving chip to turn on the GaN power device to be tested, and sets the end time of the experiment;
[0044] The driver chip Si8271GB-IS can transfer the voltage input from the VDD pin to the output pin Vo+ with equal amplitude, and transfer the sinusoidal signal input from the VI pin to the output pin Vo+ with equal frequency, thereby controlling the amplitude and frequency of the driving voltage signal of the GaN power device under test.
[0045] The sampling resistor and the GaN power device to be tested are connected in series in the power loop of the voltage source, and the amplitude and frequency of the current flowing through the GaN power device to be tested are measured through the sampling resistor; the power loss of the GaN power device to be tested is obtained by multiplying the voltage across the voltage source and the current flowing through the sampling resistor.
[0046] The power P(t) generated by the GaN power device to be tested is used as an excitation signal of the equivalent Cauer thermal path model to provide necessary conditions for the operation of the equivalent Cauer thermal path model. The thermal path fatigue of the GaN power device after long-term operation will cause the thermal resistance R θ Changes, and the shell temperature phase-frequency response curves of Cauer thermal paths under different thermal resistances are compared.
[0047] A method for evaluating the thermal path status of GaN power devices based on frequency response, the implementation method of which is as follows Figure 1 As shown, the specific steps include:
[0048] Step 1: Figure 2 As shown, a GaN power device test circuit is constructed, including a voltage source for powering a power circuit; a microprocessor (STM32F334) for issuing a sinusoidal drive signal and controlling the GaN power device to be tested through a drive chip; a sampling resistor for collecting the AC current flowing through the GaN power device to be tested; and the GaN power device to be tested for generating AC loss, which generates AC loss to provide AC excitation for the equivalent Cauer thermal path model.
[0049] Step 2: Pass Figure 2 The GaN power device test circuit can obtain the amplitude and frequency of the AC current flowing through the sampling resistor. The sampling resistor has a low resistance value, so its power loss in the GaN power device test circuit is ignored. The power loss of the GaN power device to be tested is obtained by multiplying the voltage across the voltage source and the current flowing through the sampling resistor, and it is used as the excitation signal of the Cauer thermal path model.
[0050] Step 3: If Figure 3 As shown in the figure, the equivalent Cauer thermal path model of the GaN power device to be tested is constructed. The chip layer, substrate layer, thermal interface material, and heat sink of the GaN power device to be tested are each equivalent to a thermal resistance R θ and a heat capacity C θ . Make four layers of R θ C θ The structure is connected in parallel and R θ C θThe network front end is connected in parallel with the excitation signal obtained in step 2, and the equivalent Cauer thermal path model of the GaN power device to be tested can be built. The initial thermal resistance and thermal capacitance values are provided by the data sheet of the GaN power device model GS66508P.
[0051] Step 4: Figure 4 As shown in the figure, the linearization analysis of the equivalent Cauer thermal path model of GaN power devices is realized in Matlab / Simulink. It is required to use the input module connected with the controllable current source module as the input signal of the equivalent Cauer thermal path model of GaN power devices, and select the connection line between the input module and the controllable current source module as the closed-loop input. θ C θ The model is connected, and the voltage measurement module is connected to the output module as the output signal of the shell temperature measurement, and the connection line between the voltage measurement module and the output module is selected as the open-loop output.
[0052] Input module: Inport1, connected to a controllable current source, provides AC excitation input for the equivalent Cauer thermal path model;
[0053] Controlled Current Source module: Controlled Current Source, the input terminal s is connected to the input Inport1, and the positive and negative ends are connected to C θ1 , as the input module signal transmission medium;
[0054] Voltage measurement module: Voltage Measurement, measure the output terminal v connected to the output Outport2, and connect the positive and negative ends to R θ4 , as the output module signal transmission medium;
[0055] Output module: Outport2, connected to the voltage measurement module, measures the shell temperature Tc output signal.
[0056] Step 5: Through the linear analysis of the Cauer thermal path model in step 4, compare the phase-frequency offset in different frequency intervals of the phase-frequency response curve and the phase offset in the same frequency interval, and analyze and evaluate the degree of fatigue position of the thermal path of the GaN power device to be tested.
[0057] Example:
[0058] In this embodiment, Figure 4 As shown in the figure, the equivalent Cauer thermal path model of GaN power device is built in Matlab / simulink. Figure 4 The parameters of the relevant simulation components are shown in Table 1 below:
[0059] Table 1 Simulation component parameters
[0060] Controllable current source current 10A Controllable current source frequency 1KHz-1MHz <![CDATA[Thermal resistance R θ1 > 0.015℃ / W <![CDATA[Heat capacity C θ1 > 8e-5W·s / ℃ <![CDATA[Thermal resistance R θ2 > 0.23℃ / W <![CDATA[Heat capacity C θ2 > 7.4e-4W·s / ℃ <![CDATA[Thermal resistance R θ3 > 0.24℃ / W <![CDATA[Heat capacity C θ3 > 6.5e-3W·s / ℃ <![CDATA[Thermal resistance R θ4 > 0.015℃ / W <![CDATA[Heat capacity C θ4 > 2e-3W·s / ℃
[0061] In the simulation circuit of Matlab / Simulink, the parameters of the controllable current source, thermal resistance, and thermal capacitance are set as shown in the table, and the equivalent Cauer thermal path model is linearized and analyzed to obtain the phase-frequency characteristic curve corresponding to the junction temperature.
[0062] The thermal resistance R θ1 The value of is increased by 1.067 times, 1.134 times, and 1.2 times respectively, and the other parameters remain unchanged. The phase-frequency characteristic curve corresponding to the junction temperature is compared with the phase-frequency characteristic curve corresponding to the original junction temperature. As shown in Figure 5(a), it is observed that the phase-frequency characteristic curve changes with the thermal resistance R in the high frequency range. θ1 As the phase increases, it gradually shifts to the left, that is, the phase becomes lower and lower in the high-frequency range, and the four phase-frequency characteristic curves approximately coincide with the medium-frequency range in the low-frequency range.
[0063] The thermal resistance R θ2 The value of is increased by 1.067 times, 1.134 times, and 1.2 times respectively, and the other parameters remain unchanged. The phase-frequency characteristic curve corresponding to the junction temperature is compared with the phase-frequency characteristic curve corresponding to the original junction temperature. As shown in Figure 5(b), it is observed that the phase-frequency characteristic curve changes with the thermal resistance R in the intermediate frequency range. θ2 As the frequency increases, it gradually shifts to the left, that is, the phase becomes lower and lower in the intermediate frequency range, and the four phase-frequency characteristic curves approximately coincide with the high frequency range in the low frequency range.
[0064] The thermal resistance R θ3 The value of is increased by 1.067 times, 1.134 times, and 1.2 times respectively, and the other parameters remain unchanged. The phase-frequency characteristic curve corresponding to the junction temperature is compared with the phase-frequency characteristic curve corresponding to the original junction temperature. As shown in Figure 5(c), it is observed that the phase-frequency characteristic curve changes with the thermal resistance R in the low frequency range. θ3 As the frequency increases, it gradually shifts to the left, that is, the phase becomes lower and lower in the low-frequency range, and the four phase-frequency characteristic curves approximately coincide with the high-frequency range in the intermediate-frequency range.
[0065] As shown in Figure 5(a), Figure 5(b) and Figure 5(c), after the Cauer thermal path model linear analysis, the specific position of the thermal path of the GaN power device responds differently to the frequency range. High-frequency signals are likely to affect the GaN power device chip layer, medium-frequency signals are likely to affect the GaN power device substrate layer, and low-frequency signals are likely to affect the GaN power device thermal interface material. This method is feasible for evaluating the health status of the thermal path position of the GaN power device.
[0066] For the determined thermal path position of GaN power devices, in the same frequency range, the greater the phase deviation between its phase-frequency characteristic curve and the original phase-frequency characteristic curve, and the smaller the phase value, the higher the thermal resistance value corresponding to this layer of material, that is, the higher the fatigue degree of this material layer. This method is feasible for evaluating the fatigue degree of the thermal path of GaN power devices.
[0067] The present invention provides a method for evaluating the thermal path state of a GaN power device based on frequency response, which has the following characteristics:
[0068] (1): The AC power injected into the GaN power device is used as the excitation, and the device shell temperature is used as the response. The frequency response characteristics of the two are used to characterize the state of the device heat dissipation path. This method does not need to destroy the GaN power device package and can monitor the fatigue level of the GaN power device thermal path.
[0069] (2): By extracting the frequency response characteristics of GaN power devices under different health states, the frequency range where the phase-frequency characteristic curve is stratified is used as the characteristic parameter to evaluate the location of thermal path fatigue of GaN power devices; the phase offset where the phase-frequency characteristic curve is stratified is used as the characteristic parameter to evaluate the degree of thermal path fatigue of GaN power devices.
[0070] (3): A GaN power device test circuit was constructed. By controlling the driving signal of the GaN power device to be tested, the driving chip can provide a driving voltage with a superposition of a sinusoidal component and a DC component, so that an AC current flows through the GaN power device and the current amplitude and frequency are collected through a sampling resistor; the power consumption of the GaN power device to be tested during operation is calculated.
[0071] (4): The present invention discloses a method for evaluating the state of the thermal path of a GaN power device, which can realize the evaluation of the location of the thermal path of a GaN power device and the degree of fatigue of the thermal path. The basic idea of this method is based on the one-dimensional Cauer thermal network of the device, which is regarded as a multi-order passive linear network. The drift of the network parameters is evaluated according to the change of the input-output frequency characteristics, and then the state of the thermal path of the device is evaluated. In the specific implementation of this method, the AC power injected into the device is used as the excitation, and the shell temperature of the device is used as the response. Without destroying the device package and extracting the internal parameters of the device, the change of the phase-frequency characteristic curve constructed by the two is used to evaluate the state of the thermal path of the GaN power device under different fatigue states.
Claims
1. A GaN power device test circuit, characterized in that include: A voltage source for supplying power to the power circuit; The driving circuit sends out a sinusoidal driving signal and controls the GaN power device under test; A sampling resistor, used to collect the AC current flowing through the GaN power device to be tested; The GaN power device to be tested generates AC loss, which provides AC excitation for the equivalent Cauer thermal path model.
2. A GaN power device test circuit according to claim 1, characterized in that: The driving circuit includes a microprocessor and a driving chip. The microprocessor is connected to the driving chip. The microprocessor sends a sinusoidal signal through the driving chip to turn on the GaN power device to be tested, and sets the end time of the experiment.
3. A GaN power device test circuit according to claim 1, characterized in that: The sampling resistor and the GaN power device to be tested are connected in series in the power loop of the voltage source, and the amplitude and frequency of the current flowing through the GaN power device to be tested are measured through the sampling resistor; the power loss of the GaN power device to be tested is obtained by multiplying the voltage across the voltage source and the current flowing through the sampling resistor.
4. A GaN power device test circuit according to claim 1, characterized in that: The power generated by the GaN power device to be tested is used as an excitation signal of the equivalent Cauer thermal path model to provide necessary conditions for the operation of the equivalent Cauer thermal path model. The thermal path fatigue of the GaN power device after long-term operation will cause the thermal resistance R θ Changes, and the shell temperature phase-frequency response curves of Cauer thermal paths under different thermal resistances are compared.
5. A GaN power device test circuit according to claim 4, characterized in that: The shell temperature phase-frequency response curve is obtained by linearizing the equivalent Cauer thermal path model of the GaN power device to be tested, and the shell temperature response changes with the thermal resistance R θ The law of change.
6. A method for evaluating the thermal path status of a GaN power device based on frequency response, characterized in that The following steps are involved: S1: constructing a GaN power device test circuit as claimed in any one of claims 1 to 4; S2: The amplitude and frequency of the AC current flowing through the sampling resistor are obtained through the GaN power device test circuit built in S1. The power loss of the GaN power device to be tested is obtained by multiplying the voltage across the voltage source and the current flowing through the sampling resistor, and the product is used as the excitation signal of the Cauer thermal path model. S3: Build an equivalent Cauer thermal path model of the GaN power device to be tested; S4: Implementing linear analysis of the equivalent Cauer thermal path model of the GaN power device to be tested; S5: Through the linear analysis of the equivalent Cauer thermal path model of S4, the phase-frequency offset in different frequency intervals and the phase offset in the same frequency interval of the phase-frequency response curve are compared to analyze and evaluate the location and degree of thermal path fatigue of the GaN power device to be tested.
7. The method for evaluating the thermal path status of a GaN power device based on frequency response according to claim 6, characterized in that: In S3, the chip layer, substrate layer, thermal interface material, and heat sink of the GaN power device to be tested are each equivalent to a thermal resistance R θ and a heat capacity C θ , so that the four-layer R θ C θ The structure is connected in parallel and R θ C θ The parallel excitation signal at the front end of the parallel structure is obtained by S2, which can build an equivalent Cauer thermal path model of the GaN power device to be tested.
8. The method for evaluating the thermal path status of a GaN power device based on frequency response according to claim 6, characterized in that: In S4, the input module is connected to the controllable current source module as the input signal of the equivalent Cauer thermal path model of the GaN power device to be tested, and the connection line between the input module and the controllable current source module is selected as the closed-loop input. The input signal is connected to the fourth-order R θ C θ The model is connected, and the voltage measurement module is connected to the output module as the output signal of the shell temperature measurement, and the connection line between the voltage measurement module and the output module is selected as the open-loop output.
9. The method for evaluating the thermal path status of a GaN power device based on frequency response according to claim 6, characterized in that: In S5, the specific location of the thermal path of the GaN power device responds differently to the frequency range. High-frequency signals are likely to affect the chip layer of the GaN power device, medium-frequency signals are likely to affect the substrate layer of the GaN power device, and low-frequency signals are likely to affect the thermal interface material of the GaN power device. For a determined GaN power device thermal path position, in the same frequency range, the greater the phase deviation between its phase-frequency characteristic curve and the original phase-frequency characteristic curve and the smaller the phase value, the higher the thermal resistance value corresponding to this layer of material, that is, the higher the fatigue degree of this material layer.