Andi Buzo

On the influence of angle sensor nonidealities on the torque ripple in PMSM systems — An analytical approach

Accurate measurement of rotor angle in Permanent Magnet Synchronous Motor (PMSM) drive applications is important when high performances as low level torque ripples are required. If the relationship between torque ripple and angle error is known, then, further steps can be done in order to reduce the level of the ripple. Based on an analytical description of the sensor model, a relationship between the angle error and speed error (seen as a noise source for the system) as well as torque ripples are deduced. Simulation results for a specific range of mechanical speed using SystemC-AMS models confirm the theoretical analysis.

On limit cycles suppression in DC-DC Buck converters

This paper focuses on the analysis of limit cycles that appear on the output voltage of a DC-DC Buck converter. Typically, the output filter of the converter is designed in open-loop, considering the maximum allowable values for the output voltage and inductor current ripples. When closing the loop, the converter may exhibit unwanted periodic oscillations in steady state caused by the nonlinearities in the system. The amplitudes of these oscillations are hard to predict. This paper summarizes the conditions that should be taken into account when designing the closed loop converter. These conditions are used furthermore for tuning the PID parameters for a DC-DC Buck converter under input and load step scenarios. The tuning of the PID takes into consideration transient performances such as overshoot and settling time for specified test cases but also the conditions needed for removing the limit cycles from the output voltage.

Simulation-based approach to application fitness for an E-Bike

Apart from fulfilling component electrical specifications, electronics needs to fit into the target application, composed of electronics components as well as mechanical parts. In complex applications such investigations can be made by simulations in which relevant application performances are checked under any allowed variation of operating conditions, design parameters and noise factors. In this paper, we investigate verification methods for application fitness, on top of specification compliance. These are supported by a flow, which supports to assess a components impact on the application performances. This flow is applied on an E-Bike application modelled in SystemC-AMS. The objective is to answer whether the assessed component reasonably serves the application in an environment with uncertainties. To accomplish this, concepts of experiment planning, metamodeling and sensitivity analysis are applied.

On the stability domain of a DC-DC buck converter with software control loop

In recent years, there has been an increasing interest in developing software controlled DC-DC converters because of their fast configurability, e.g. for several supply and load values. This paper aims at exploring the stability of a DC-DC buck converter with a software control loop in the multidimensional space of control parameters. Once the stability domain is determined, it can be further used to search for the optimal parameter values for which the output voltage of the converter satisfies specific performance requirements. In this paper, we determine the stability domain by two approaches. First, we apply an analytical method that uses an analog-to-digital transform followed by the D-decomposition technique. Second, we apply a searching algorithm developed to efficiently characterize pass/fail behavior of electronic devices. The results obtained by the two methods are directly compared and discussed. The analytical approach is more conservative in determining the stability border, identifying a sub-domain of the real one. The search-based method proved to be more accurate in determining the stability border, at the cost of additional simulation runs.

Robust response design for heterogeneous systems based on metamodels

The robustness of the system response to production factors and operating conditions is a big challenge as the systems grow in complexity and heterogeneity. In this paper, we propose a method that solves the robust parameter design problem. First, we define some cost functions based on the distribution of the systems response and, second, we estimate the dependence of the cost function on the controllable factors. With such dependency we are then able to determine the setting of controllable factors that yields the minimum of the cost function. By this method we reduce the number of required simulations by at least one order of magnitude. The method is applied on a beam leveling application and proved to be successful even with high number of heterogeneous factors (> 30). The method is validated by comparing the estimated cost functions with direct simulations. The estimated values differ by only 5% from the simulated ones.

Data improvement in lab verification of smart power products using DoE

The paper focuses on treating and characterizing variations that occur in measurements or that are intrinsic in electronic systems. Methods illustrating basic principles of Design of Experiments such as replication and blocking are implemented and used so that valid and objective conclusions are drawn. Replication is used for three purposes: verifying the metamodel adequacy, defining the confidence interval of the measured mean value of the systems response and determining the minimum number of replications that are needed in order to measure the systems response with a given accuracy. Blocking helps with eliminating the systematic error that appears due to the measuring conditions and was used in the screening process. The methods were applied on a lighting control system used in automotive applications.

Automatic tuning for a DC-DC Buck Converter with adaptive controller

The paper presents an approach based on gain scheduling technique for improving the transient performances of a digitally controlled DC-DC Buck Converter working over a large area of operating conditions. In order to determine optimal settings for the control parameters under different test scenarios, an adaptive mechanism based on machine learning algorithms is used. The experimental results, obtained after using this approach are presented as well. An improvement of 20% is observed in the case of using this gain scheduling controller instead of a controller with constant values for the parameters.

An accurate yield estimation approach for multivariate non-normal data in semiconductor quality analysis

The standard multivariate metrics for semiconductor product yield estimation and prediction in production processes usually assume that the parameters contributing to the yield are all normally distributed. However, the data met in production processes is not always multivariate normal. A variety of methods has been developed for multivariate non-normal data, but these usually rely on no statistical information, address only a specific type of multivariate distributions, or become very time consuming from the point of view of the computational cost. Moreover, the sample size of the multivariate data is often insufficient, as only a limited number of measurements are affordable. This results in inaccurate product yield estimation and high variance of the estimates. In this paper, a multivariate distribution fitting methodology is introduced, which, combined with multivariate random data sampling provides a global yield estimation approach. Compared with the simple failure counts method the estimation variance of the proposed method is two times smaller.

Application-aware lifetime estimation of power devices

The paper proposes a methodology for lifetime estimation of power devices at given applications operating conditions. The active cycling of power devices requires huge testing-time, because the process cannot be accelerated. For this reason, most often, the manufacturers provide information about the lifetime of power devices only for a few specific operating conditions. Most of the current methods are based on the junction temperature swing, which is very difficult to be measured or estimated. Instead, we propose an approach that, based on a few measurements of lifetime at given ambient temperatures, load currents and repetitive energies, is able to make lifetime prediction at any other set of operating conditions. The validation of the method was done by performing lifetime predictions in other operating conditions than those used for fitting the prediction function (metamodel) and it has shown a maximum relative error of 20%. With the proposed methodology, lifetime estimations of power devices can be made in the space of applications operating conditions, using optimal testing resources.

Distribution fitting approach to application fitness assessment

On top of assessing the specification compliance, it is also important to verify the behavior and performance of the electronic components in the targeted application. This is usually achieved by jointly simulating the component and the application. There is a particular interest in finding the application yield caused by the process variation of the electronic component. The cost of the experiments is usually high, and, if the estimation of the Failure Probability (FP) is done from counts of failure from few simulations, then, the FP is not very accurate. We propose a method of FP estimation based on distribution fitness which represents a more robust statistic. The validation of the method is made on an E-Bike use case and the results show that the proposed approach is more accurate than the one based on counts. The investigation is made by taking into account the role of the operating conditions in the experiments. We achieve this by grouping the test scenarios under a number of mission profiles, on which the developed methods are carried out.