CN-122017515-A - Gallium nitride power device packaging parasitic parameter degradation monitoring method

CN122017515ACN 122017515 ACN122017515 ACN 122017515ACN-122017515-A

Abstract

The invention provides a gallium nitride power device package parasitic parameter degradation monitoring method, which comprises the steps of establishing a high-frequency RLC oscillation model of a gallium nitride power device containing package parasitic parameters under a high-frequency switching condition, judging whether an oscillation amplitude is reduced compared with a switching transient current change rate reference waveform by a switching transient current change rate measurement waveform, determining the correlation between the package parasitic parameters and the switching transient current change rate oscillation amplitude, predicting whether the package of the gallium nitride power device is degraded, reflecting the switching transient current change rate change by the package parasitic parameters, monitoring whether the package of the gallium nitride power device is degraded by the package parasitic parameters, effectively identifying early signals of early signal package degradation of device package degradation, capturing early tiny package degradation signs, and having high package degradation monitoring accuracy.

Inventors

CHEN YIQIANG
LU XING
HOU BO

Assignees

中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室））

Dates

Publication Date: 20260512
Application Date: 20260409

Claims (9)

1. A method for monitoring package parasitic parameter degradation of a gallium nitride power device is characterized by comprising the following steps: S1, establishing a high-frequency RLC oscillation model of the gallium nitride power device containing packaging parasitic parameters under a high-frequency switching condition; S2, collecting the change rate of the transient current of the switch, and extracting a measurement waveform of the change rate of the transient current of the switch; S3, comparing the switching transient current change rate measurement waveform with a switching transient current change rate reference waveform, judging whether the switching transient current change rate oscillation amplitude is reduced or not, if so, carrying out S4; S4, determining the correlation between the packaging parasitic parameter and the oscillation amplitude of the transient current change rate of the switch; S5, predicting whether the package of the gallium nitride power device is degraded or not.
2. The method for monitoring package parasitic parameter degradation of GaN power device of claim 1, wherein the package parasitic parameter comprises equivalent parasitic resistance Equivalent parasitic inductance Equivalent parasitic capacitance 。
3. The method for monitoring the package parasitic parameter degradation of the gallium nitride power device according to claim 2, wherein S4 comprises the following steps: s4.1 by Determining a frequency coefficient delta; S4.2, through Determining oscillation frequency ; S4.3 by Determining frequency coefficient delta and switching transient current change rate waveform Is of negative correlation with oscillation frequency And switching transient current rate of change waveform Is a positive correlation.
4. The method for monitoring package parasitic parameter degradation of GaN power device of claim 3, wherein S5 is specifically that when oscillation amplitude of transient current change rate of switch decreases, frequency coefficient delta increases, and equivalent parasitic resistance increases And (5) increasing, and predicting the package degradation of the gallium nitride power device.
5. The method for monitoring the degradation of the parasitic parameters of the gallium nitride power device package of claim 4, wherein S5 further comprises, according to the equivalent parasitic resistance Predicting the package degradation of the gallium nitride power device.
6. The method for monitoring package parasitic parameter degradation of a gallium nitride power device according to claim 3, wherein in S3, the switching transient current change rate measurement waveform comprises a measurement peak value, the switching transient current change rate reference waveform comprises a reference peak value, and comparing the measurement peak value with the reference peak value to judge whether the switching transient current change rate oscillation amplitude is reduced.
7. The method for monitoring package parasitic parameter degradation of a gallium nitride power device according to claim 6, wherein the measurement peak value is a maximum absolute value of first two peaks of a switching transient current change rate measurement waveform, and the reference peak value is a maximum absolute value of first two peaks of a switching transient current change rate reference waveform.
8. The method for monitoring package parasitic parameter degradation of a gallium nitride power device according to claim 3, wherein the method further comprises presetting stress action time, presetting gallium nitride power device grid voltage, continuously applying dynamic grid bias stress to the gallium nitride power device in the stress action time, and obtaining package parasitic parameters after the step S5.
9. The method for monitoring the package parasitic parameter degradation of the gallium nitride power device according to claim 8, wherein the method is characterized in that the preset cycle times are adopted, and the dynamic grid bias stress is repeatedly applied to the gallium nitride power device in each stress acting time according to the cycle times; Recording the switching transient current change rate and packaging parasitic parameters of the gallium nitride power device under different cycle times; and verifying the changes of the package parasitic parameters and the change rate of the switching transient current of the gallium nitride power device along with the accumulation of stress acting time.

Description

Gallium nitride power device packaging parasitic parameter degradation monitoring method Technical Field The invention relates to the field of degradation monitoring of gallium nitride power devices, in particular to a method for monitoring package parasitic parameter degradation of a gallium nitride power device. Background With the development of high frequency and high power density of power electronic technology, gallium nitride (GaN) power devices are widely used in the fields of energy systems, data center power supplies and the like by virtue of the advantages of wide forbidden band, high electron mobility and the like. Gallium nitride (GaN) power devices are affected by high current change rate (di/dt) and voltage change rate (dv/dt) during high frequency switching, which not only can lead to degradation of electrical parameters (such as threshold voltage Vth and on-resistance r_on) of the gallium nitride (GaN) power devices, but also can accelerate failure of package contact, such as bonding wire aging, metal interconnection interface chemical reaction and the like, resulting in changes of inductance and parasitic resistance of the package. At present, degradation sensing and technology of gallium nitride (GaN) power devices mainly depend on single electrical parameters (such as Vth and R_on), however, in a high-frequency switching process, the depth of packaging parameters (parasitic resistance, parasitic inductance and parasitic capacitance) and the energy exchange process of switching transient have great influence on the performance of the gallium nitride (GaN) power devices. Monitoring the current change alone often fails to fully reflect the package degradation state under high frequency switching conditions, and is prone to "non-alert failure". In China application number 202411335513.8 and publication day 2025.1.10, a circuit and a method for monitoring aging of a SiC power MOSFET device are disclosed, wherein a collecting device collects current and voltage data in a turn-off process, and the bonding wire failure process of the SiC power MOSFET is monitored in real time by combining the relation between a Kelvin source electrode and a power source electrode inductance value. However, the integration and division calculation processes mentioned in this patent may increase the computational complexity, especially in real-time monitoring scenarios, which may lead to monitoring delays that cannot immediately reflect the state changes of the device. At the same time, the patent uses invasive measurement, which not only increases hardware cost and design complexity, but also introduces additional parasitic parameters that may affect the high frequency performance of the main circuit. In addition, the existing current probe measurement is non-contact, but only focuses on current amplitude, and lacks in-depth modeling analysis on quantitative relation between transient oscillation characteristics of a switch and package parasitic parameters, so that precise decoupling and monitoring on a package degradation mechanism (such as parasitic resistance increase) are difficult to realize. Disclosure of Invention Therefore, the invention aims to provide the gallium nitride power device package parasitic parameter degradation monitoring method, which is used for monitoring whether the package of the gallium nitride power device is degraded or not by using the package parasitic reflection switch transient current change rate, so that the early signal of the early signal package degradation of the device package degradation can be effectively identified, the early tiny package degradation sign can be captured, and the package degradation monitoring accuracy is high. In order to solve the technical problems, the invention adopts the following technical scheme: The invention relates to a gallium nitride power device packaging parasitic parameter degradation monitoring method, which comprises the following steps: s1, a high-frequency RLC oscillation model of the gallium nitride power device containing packaging parasitic parameters under the high-frequency switching condition is established. S2, collecting the change rate of the switching transient current, and extracting a measuring waveform of the change rate of the switching transient current. S3, comparing the switching transient current change rate measurement waveform with the switching transient current change rate reference waveform, judging whether the switching transient current change rate oscillation amplitude is reduced, and if so, performing S4. S4, determining the correlation between the packaging parasitic parameter and the oscillation amplitude of the transient current change rate of the switch. S5, predicting whether the package of the gallium nitride power device is degraded or not. Preferably, the package parasitic parameter includes an equivalent parasitic resistanceEquivalent parasitic inductanceEquivalent parasitic capacitance。 Preferably