Buck Converter Performance Improvement
Weekly Progress
September 11, 2015
Final corrections in the dissertation document and delivery of the thesis.
September 4, 2015
Revision of all the chapter and final measurements and tests with the buck converter.
August 27, 2015
Finished the third and fourth chapters, writing of the final chapter. Revision of the 4 previous chapters.
August 21, 2015
Writing of the third and fourth chapters. Final tests of the transient response of the buck converter and sweep load to prove the estimation accuracy.
August 14, 2015
Finish of the two first chapters. Writing of the third chapter. Simulations with injected noise in the output voltage and evaluation of the results.
August 7, 2015
Start of the writing of the dissertation, by the first two chapters. Repeat of the simulations and test on the buck converter of the current and resistor estimation accuracy.
July 31, 2015
Optimization of the control algorithm, to lower the overall execution time. Implementation of a oversampling of the output voltage to help the calculation of the average value of the output voltage.
July 24, 2015
Design and test of the control loop, including the voltage and current loop. Test of the algorithm in the microcontroller and evaluation of the results.
July 17, 2015
Implementation of a Kalman Filter library in the microcontroller, including also a matrix operations library.
July 10, 2015
Evaluation of the results with EKF. Offline dead time optimization was performed instead of introducing it in the model.
July 3, 2015
Design and implementation of an extended Kalman filter algorithm to estimate the load of the buck converter, alongside the current estimation.
June 26, 2015
Modeling of the buck converter during the dead time period and evaluation of results.
June 19, 2015
Simulations of current estimation were performed with the ideal model and their accuracy was evaluated. Modeling of the parasitic parameters of the capacitance and inductor.
June 12, 2015
Modeling of the buck converter, without considering the parasitics of the model. The model was achieved using state-space averaging method.
June 5, 2015
The focus of the dissertation will now be the aplication of the Kalman filter in a buck converter. The main objective will be to improve the response of this device when its output voltage is affected by noise, improving then indirectly the ADC performance.
May 29, 2015
Change of the theme of the dissertation. The results obtained until this point were not conclusive and not reproducible in all microcontroller channels and the supply voltage influence was difficult to evaluate. Therefore, a discussion with the dissertation supervisors and Infineon Engineers was done to evaluate the best course of action.
May 22, 2015
Discussion and analisys of the obtained results. Identification of the influences of clock and supply voltages in the ADC noise.
May 15, 2015
Evaluation of the results for other microcontrollers from both XMC1000 and XMC4000 families, to evaluate the consistency of the results.
May 8, 2015
Continuation of the last week work. Discussion of the results.
May 1, 2015
Analysis of the infuence due to clock activity on the ADC. The objective was triyng to understand why the clock and clock changes have a influence on the ADC result and why this is different in different channels. Discussion of the results.
April 24, 2015
Early implementation of the Kalman filter algorithm, using measured signals as input. Tuning of the filter based on the parameters of the measured signals.
April 17, 2015
The tests were performed in multiple devices. The goal is to prove that the results are consistent when the model is the same, even in different devices, and to prove that the results are device specific and not production dependant. Confirmation and discussion of the results.
April 10, 2015
Introduction of a more stable singal to measure the error introduced only by the ADC in the signal. This is important due to the fact that the measured error might come from the signal itself and not only from the ADC converter. The most stable signal was achieved using a voltage regulator. The tests were repeated and compared to the previously obtained results.
April 3, 2015
Redo the test proedures measuring also the supply and core voltages. Nedded to implement another microcontroller to measure this voltages, in order to have independent results. Analysis and discussion of the influence of this voltages in the final outcome of the ADC conversion.
March 27, 2015
Repeated the temperature and activity tests measuring the results in two channels. Generated random activity by constantly enable/disable clocks, instead of only enabling them only one time. Mixing of the both tests and analysis and discussion of the results.
March 20, 2015
The tests that were done previously were repeated, now for two different microcontroller series. Analysis and discussion of the obtained results with the supervisors.
March 13, 2015
Analysis of the ADC response under controlled conditions: temperature variation and internal activity. The temperature test was made by heating the microcontroller and compare the ADC output with the output obtained at normal operation temperature. The internal activity test required enabling and disabling of multiple clocks that generate PWM waves and compare the results with the results obtained with no activity. The results were discussed and analyzed with the supervisors.
March 06, 2015
Measurement of the supply and core voltages of the microcontroller and analysis of the error magnitude. Update of the logging program in order to record the temperature of the die of the microcontroller. Development of a MATLAB algorithm to display the logged results to provide a better analysis.
February 27, 2015
Study how the microcontroller and its internal signals/activity influence the ADC. Determination of the internal microcontroller registers that will be used in order to measure the desired values. Coding of the base program that will be used to retrieve and log the desired variables in order to analyze and compare multiple experiments.