[P1]
The main contribution of the paper is the accurate small-signal modeling of the PCMC converters. However, in the scope of this thesis the most important contribution is the derivation of the open-loop dynamical profile from practical measurements. Due to the current-source nature of the PCMC at open-loop, a resistive load had to be used in the measurements. It was noticed that the resistive load effectively hid the nominal behavior, which could lead to wrong deductions of the model accuracy. To recover the internal dynamics in the control-to-output transfer function a mixed-data method was developed. As a result, it was discovered that the analytical model and the profile derived by using the mixed-data method coincided.
[P2]
The characterization of regulated converters was addressed to enable the assessment of the stability, performance, supply and load interactions as well as the transient responses. A canonical model of a converter was proposed and applied to create a set of parameters that were able to fully describe the dynamics associated with a converter as such as well as under external interactions. System theory was used to develop methods to study the internal stability of cascaded subsystems such as an EMI filter, load and other converters. A framework that allowed the evaluation of different converter topologies under different operation and control modes was established.
[P3]
The stability and performance of a regulated converter was analyzed based on its
functions that defined the internal stability of an interconnected system consisting of the source and load converters. The internal stability was described in terms of the ratio of the output impedance of the source converter and the input impedance of the load converter known as the minor-loop gain. Thus, the closed-loop output impedance of a source converter could be used to define the safe operating areas that avoided instabilities imposed by the load impedance. It was shown that the margins associated with the minor-loop gain (i.e. the gain and phase margins) did not generally match with the margins of the output-voltage loop gain. The relationship was especially weak at the frequencies close to and beyond the crossover frequency of the loop gain. This means that the margins specified to the minor-loop gain should be gradually increased as the voltage-loop-gain crossover frequency is approached in order to avoid performance degradation (i.e. changes in margins and crossover frequency) in the supply converter. Experimental evidence was provided based on a buck converter under VMC and PCMC.
[P4]
The effect of the load impedance on the dynamics and performance of a regulated converter was investigated. Theoretical formulation was derived utilizing two-port representation. It was definitively shown that the load interactions were reflected into the converter dynamics via the internal open-loop output impedance. At the frequencies, where the loop gain was much higher than unity, the internal closed-loop output impedances acted as a boundary for the control-bandwidth reduction. The loop gain was always affected, whenever the internal open-loop output impedance was equal or greater than the load impedance. It was found out that the converters were sensitive especially to the capacitive and resonant-type loads. The sensitivity was dependent on the control mode, and could not be much reduced by means of the basic controller design.
[P5]
The characterization of regulated converters was investigated in order to establish a set of dynamical parameters defining the interactions arising in the interconnected systems such as DPS systems. The commercially available converters are usually vaguely specified in respect to those interactions. Provided information do not suffice for predicting the stability and performance. It was noticed that there were certain double reflections, which were not previously recognized but could increase the load sensitivity if not properly considered. The defined parameter set could also be used to design the converters to be more insensitive to different interactions.
[P6]
The paper investigated the dynamical differences of the load and supply interactions in the direct-duty-ratio or VMC buck converters in the continuous and discontinuous conduction modes. The dynamical parameters i.e. the phase margin and control bandwidth of the CCM and DCM converters were designed to be the same. It was shown that the sensitivity for the interactions could be concluded from the measured frequency responses. The investigations showed that a buck converter operating in the CCM was more sensitive to capacitive loads than a converter in the DCM, while the DCM converter was more prone to instability caused by the load. The supply interactions caused e.g. by an EMI filter were shown to be smaller in the CCM converter except near the converter-output-filter resonant frequency, where the forward-voltage transfer function had an amplifying effect on the supply interactions.
[P7]
The paper studied the stability and performance analysis of a regulated converter based on the impedance ratio known commonly as the minor-loop gain. System theory was used to prove that the minor-loop gain could be addressed to the internal stability of an interconnected system in a scientifically sound manner. As a consequence, the output impedance of the source system could be used to specify the stability boundary of the source converter in respect to the load impedance. The margins assigned to the minor-loop gain did not, however, coincide with the margins associated to the loop gain of the converter. Therefore, the converter performance can be deteriorated unless the minor-loop gain margins are sufficient for avoiding that.
Experimental evidence was provided to support the ideas presented in the paper.
[P8]
The paper investigated the effects of an EMI filter on the dynamics of a buck converter. It was shown theoretically that the EMI filter could increase significantly the load sensitivity of the VMC converter, but the PCMC converter was quite insensitive to the EMI filter interactions. Experimental validations were carried out using a buck converter with three different control modes - VMC, PCMC and PCMC-OCF. The investigations showed that the phenomenon causing the instability under PCMC was the negative-resistor–oscillation (NRO) phenomenon, and confirmed also the excess EMI-filter sensitivity of the VMC converter.
[P9]
The paper investigated the load interactions in a constant-current-controlled buck
derived from the voltage-output-converter model by applying duality. It was observed that the constant-current control made the converter very sensitive to load interactions implying that a proper controller design is a necessity. The high crossover frequency in the nominal control-to-output transfer function may lead to performance degradation if the initial control design is done with a resistive load, because the typical load for the constant-current-controlled converters is a low impedance storage-battery recovering the nominal feature. It was shown that electronic loads, typically used in the prototype testing and verification, might have unexpected characteristics. This could lead to wrong deductions of the converter stability and performance.
[P10]
The paper investigated the dynamical properties of the current-output converter, and the set of transfer functions representing the dynamics was developed based both on the state-space averaging method and on the two-port model of the corresponding voltage-output converter. According to the investigations, the observed tendency to increasing crossover frequency could be addressed to the internal dynamics of the converter, which the storage battery invoked into effect. This meant that the origin of the problem was an erroneous control design.
[P11]
The mechanism and characterizing parameters causing the source reflected interactions were investigated in this paper both theoretically and experimentally using a buck converter under a VMC, PCMC and input voltage-feedforward (IVFF) control. The reflected interactions would be eliminated if the forward-voltage transfer function could be made zero, but in practice such a condition could not be fully achieved. The investigations showed that the VMC converter was very sensitive to the source-reflected interactions.