Accurate junction temperature characterization plays a critical role in the performance as well as reliability enhancement design of Insulated gate bipolar transistor (IGBT) modules. However, most current methods are only focused on predicting the temperature at a specific time rather than tracking its temporal variation. To address this challenging issue, we propose a long short term memory (LSTM) neural network-based method for predicting the junction temperature time series of IGBTs, and it can comprehensively characterize the time-dependent junction temperature according to the measured and validated experimental results. As the convolutional neural network is limited in its ability to remember temporal information, we integrate the LSTM networks for training, which can fast capture long-time dependencies in the data. Therefore, we exploit the superior performance of LSTM network for predicting the junction temperature of IGBTs here. It is shown that there is a strong correlation between the maximum junction temperature of IGBT and its parameters, such as bus voltage, load current, carrier, and output frequencies, as well as its heat sink parameters. By leveraging the LSTM-based method, an accurate time series information of the junction temperature is successfully captured under different operating conditions. Notably, in comparison with the current machine learning method, the proposed method can outperform in the same training dataset.

A miniaturized ultrawideband active electric probe with high sensitivity for near-field measurement is proposed in this article. The probe is designed with a four-layer printed circuit board (PCB) with mainly two chips, an amplification chip (HMC797APM5E) which can work from dc to 22 GHz and a power management chip (HMC980LP4E). The probe can work from 9 kHz to 18 GHz with 15-dB average gain compared to the passive probe, and the flatness is about ±4 dB in terms of $\vert S_{21}\vert $ (transfer factor) from 2.5 to 18 GHz. The maximum size of the probe is 64 mm $\times41.5$ mm with 1-mm spatial resolution, and the width of the tip is only 4.1 mm, so the probe can handle most test environments. On the whole, the probe is very efficient for near-field measurement to detect weak signals in a wide bandwidth.

Coupling analysis of shielded and unshielded star quad cables excited by an intentional electromagnetic pulse (IEMP) is investigated using the transmission line theory-based perturbation technique. A detailed modal analysis is presented to obtain the modal per-unit-length parameters of the modal signals in such shielded/unshielded star quad cables. Due to their geometrical balanced structures, with opposite wires used for +leg and −leg to transfer the differential-mode (DM) signal, the external coupling is dramatically suppressed. Moreover, stochastic analyses based on the Monte Carlo simulation are also performed to study the sensitivity of the induced common-mode (CM) and DM currents, due to the deformation effects of the twisting structure of the cable. It is found that the induced CM and DM currents are slightly affected by the twisting deformation of the cable because the interference cancelations do not mainly depend on the twisting loops of the cable. In particular, time domain analysis is performed to study the field-to-wire coupling of the star quad cables exposed to an IEMP, where the worst-case waveform signals in the presence of an IEMP attack are captured and analyzed.

Bonding wire arrays (BWA) play an important role in multi-chip systems. Especially the inductance, Q value, and loss of BWA are closely related to the performance of the chip systems. However, in industrial applications, full-wave numerical design approaches brings data burden and high computational costs, particularly for large-scale BWA. To alleviate the problem, we propose the multi-fidelity data enabled neural networks for inversely designing BWA, which consist of an inversion network and a residual network. Specifically, we firstly introduce an approximate analytical model of BWA, and perform low-fidelity data collection with the analytical model to train the inversion network. Subsequently, a small set of high-fidelity data is used to train the residual network, which is targeted to learn the residual between the analytical solutions and high-fidelity results. During the testing process, we cascade the two trained networks. Intensive numerical results are used to test the proposed method, where the average testing accuracy reaches to 94.30% with 5000 tests. To demonstrate the effectiveness of the proposed method, the inversely-designed BWA is fabricated and measured experimentally, where the measured results satisfy the performance demands. Furthermore, a power amplifier (PA) using bonding wires as matching inductors is designed, fabricated, and measured by using the proposed method.

An efficient transmission line model of high-speed cables is established, to predict its response at a high-frequency region (up to 20 GHz). Especially, an appropriate conformal mapping technique is applied to formulate the per-unit-length parameters of high-speed cables. Moreover, the skin and proximity effects of the conducting wires and the outer shield are derived in the closed form. As a result, the AC resistance/ inductance of all the conducting materials due to the skin and proximity effects are accurately incorporated for both layered and unlayered conductors. Furthermore, the mixed-mode S-parameters are precisely predicted for both balanced and unbalanced transmission line cables in the high-frequency region. In addition, the effects of the nonuniformities and cable geometry deformation on the mode conversions are also investigated. The proposed model is firstly validated with commercial software tools, COMSOL, FEKO, and HFSS; then, we further performed physical measurements to verify our new algorithm.

In this article, the skin and proximity effects formulas of solid and hollow conductors with different radii are accurately formulated using exact expressions of electric and magnetic fields around them. Moreover, both approximation and simplification of the developed formulas are also provided for higher frequency applications, where the skin depth is significantly smaller than the conductor radius. Consequently, it is shown that these approximate expressions can help us to understand the coupling mechanisms of noncoaxial and twinax cables with hollow shields, where the transfer impedances of the shields are determined with an inner conductor located at a noncoaxial location. The coupling through a hollow shield of a twinax cable is also investigated, and the induced common- and differential-mode currents are predicted in very fast and accurate way. The proposed method is validated by making comparisons with both COMSOL Multiphysics and HFSS software with different numerical methods employed.