Electro-thermal characterization of Si-Ge HBTs with pulse measurement and transient simulation

Amit Kumar Sahoo,Sébastien Fregonese,Mario Weiß,Nathalie Malbert,Thomas Zimmer
DOI: https://doi.org/10.1109/ESSDERC.2011.6044190
2013-03-19
Abstract:This paper describes a new and simple approach to accurately characterize the transient self-heating effect in Si-Ge Heterojunction Bipolar Transistors (HBTs), based on pulse measurements and verified through transient electro-thermal simulations. The measurements have been carried out over pulses applied at Base and Collector terminals simultaneously and the time response of Collector current increase due to self-heating effect are obtained. Compared to previous approach, a complete calibration has been performed including all the passive elements such as coaxial cables, connectors and bias network. Furthermore, time domain junction temperature variations, current of heat flux and lattice temperature distribution have been obtained numerically by means of 3D electro-thermal device simulations. The thermal parameters extracted from measurements using HiCuM HBT compact model have been verified with the parameters extracted from electro-thermal transient simulation. It has been shown that, the standard R-C thermal network is not sufficient to accurately model the thermal spreading behavior and therefore a recursive network has been employed which is more physical and suitable for transient electro-thermal modeling.
Materials Science,Instrumentation and Detectors
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: how to accurately characterize the transient self - heating effect in Si - Ge heterojunction bipolar transistors (HBTs). Specifically, the author proposes a new method based on pulse measurement and transient electro - thermal simulation to solve this problem. ### Research Background As the size of electronic devices continues to shrink, the impact of self - heating effects on device performance and reliability has become increasingly significant. Especially for Si - Ge HBTs used in high - frequency applications, since they operate at high current densities, power consumption increases, leading to severe self - heating effects. Traditional steady - state methods can only extract thermal resistance and heat capacity, while transient responses require more complex measurement and modeling techniques. ### Main Problems 1. **Limitations of Traditional Methods**: - Traditional self - heating effect characterization methods are mainly based on steady - state conditions and are difficult to capture transient behavior. - There are calibration challenges in pulse measurement, especially calibration related to passive components such as coaxial cables and connectors. 2. **Selection of Electro - Thermal Network Model**: - The standard R - C thermal network model is not sufficient to accurately describe thermal diffusion behavior. - A recursive network model that is more physical and suitable for transient electro - thermal modeling is required. ### Solutions 1. **Combination of Pulse Measurement and Transient Simulation**: - By applying pulse voltages to the base and collector simultaneously and measuring the change in collector current, the response of the transient self - heating effect is obtained. - Use 3D electro - thermal simulation tools (such as Sentaurus Device Simulator) for numerical simulation to verify the measurement results. 2. **Complete Calibration Process**: - Include accurate calibration of all passive components (such as coaxial cables, connectors, bias networks, etc.) to eliminate noise effects. 3. **Recursive Electro - Thermal Network Model**: - Introduce a recursive network model that can better describe the thermal diffusion behavior from the heat source to the thermal ground. - Model parameters are extracted and verified through experimental measurement and numerical simulation. ### Conclusion Through the above methods, the author has successfully achieved accurate characterization of the transient self - heating effect in Si - Ge HBTs and verified the effectiveness and accuracy of the recursive electro - thermal network model. This provides important theoretical and technical support for the future design of Si - Ge HBTs with higher performance and reliability. ### Formula Representation In the recursive electro - thermal network model mentioned in the paper, thermal diffusion can be simplified to a spherical heat conduction path, and each element is equivalent to a heating sphere. For a basic spherical slice at a distance z from the heat source with a thickness of Δz, its capacitance C(z) and resistance R(z) are defined as follows: \[ R(z)\Delta z=\frac{\Delta z}{kA(z)} \] \[ C(z)\Delta z = C_vA(z)\Delta z \] where: - \(k\) is the thermal conductivity of silicon, - \(C_v\) is the specific heat at constant volume of silicon, - \(A(z)\) is the spherical isothermal surface area at a distance z. These formulas are used to describe the time - varying characteristics of thermal diffusion. The thermal resistance and heat capacity in the recursive network change with distance and are adjusted by multiplication factors \(K_r(< 1)\) and \(K_c(> 1)\) respectively.