Christian Oeder

A comparison of different design methods for the multiresonant LLC converter with capacitive output filter

Faced with ever shorter product cycles, SMPS designers require fast and reliable modeling tools. In case of resonant converters the derivation of the exact solution is complex and time-consuming. Thus, simplified calculation methods are attractive. However, the reduced accuracy of these approximation methods has to be taken into account carefully. In this paper, the concepts of the First Harmonic Approximation (FHA) and the extended First Harmonic Approximation (eFHA) of the multiresonant LLC converter with capacitive output filter are discussed. Furthermore, the accuracy of both approximation methods is evaluated in detail with respect to the results of the exact solution in the time domain.

Analysis and design of a low-profile LLC converter

Miniaturization is currently one of the most important trends within the field of power electronics. In order to achieve this goal, increased switching frequencies are desirable to reduce the size of passive components. Resonant converters seem to be very suitable for those applications, due to the fact that virtually no switching losses occur, making high switching frequencies together with high efficiency possible. In this paper, based on its mathematically description in the time domain, a low-profile resonant LLC converter with capacitive output filter is designed with respect to a typical notebook adaptor application. Furthermore, special attention is paid towards a low-profile design of the magnetic component.

Analysis and design of a partial-multi-resonant LLC converter

In this paper a resonant converter is shown, which is based on the series resonant parallel loaded LLC converter. Limiting the voltage across the resonant capacitance with clamping diodes to the input voltage range as shown here, has several advantages like excellent zero voltage switching performance.

ZVS investigation of llc converters based on FHA assumptions

While resonant converters have been gaining much attention within the recent years, their exact mathematical description is rather complicated. Thus, simplified calculation methods are very attractive, e.g. the First Harmonic Approximation (FHA). Although the steady state solution of these simplified methods is usually quite acceptable and close to the exact solution, the prediction of the ZVS characteristic is either neglected or treated to simple. The ZVS characteristic is typically directly related to the voltage transfer function resulting in a too optimistic estimation of the operation range obtaining ZVS. Therefore, this paper focuses on a more accurate investigation of the resonant LLC converters ZVS behavior and proposes two approximation methods, which are also based on simplified FHA assumptions allowing a fast estimation of different converter designs.

Automatic measurement of the reverse recovery behavior of ultra-fast diodes

Often data sheets provide only poor information about the recovery behavior of ultra-fast diodes. On the other hand, existing diode models do not predict the real characteristic for all diodes. Nevertheless, due to its importance, this behavior needs to be known and therefore measured. For this purpose, a fully automated measurement set-up for determining the reverse recovery characteristic of ultra-fast diodes in an accurate manner was designed and is described here. All the data obtained is immediately transferred into MATLAB and therefore available for further calculation, model building and model validation. Since the whole set-up is automated, a complete field of variations in the reverse voltage, the forward current, the temperature, and the di/dt can be easily applied to the tested diode. Hence, a lot of information can be obtained effortlessly. This uncomplicated methodology makes it readily available for circuit designers, allowing them to predict the contribution of the reverse recovery of rectifiers to the total losses more accurately. In addition, a real comparison of different diodes at many operation points is made possible.

The resonant LLC vs. LCC converter - comparing two optimized prototypes

Due to the increasingly demand for highly efficient, cost-effective and low-profile applications, resonant converter topologies are massively gaining in importance. With the LCC and LLC converter, the two most promising topologies of this converter class are presented and compared within this paper. Based on their individual configuration, both converters are analyzed in the time domain revealing the first differences and similarities. Moreover, both converter topologies are optimized with respect to a low-profile notebook adapter specification for 100 W. The main focus is placed on the magnetic design of both transformers. Finally, two prototypes using completely identical layouts are developed and measurement results are discussed concerning best operation point and overall efficiency.

Sampled-data modeling applied to output impedance reveals missing term in state-space averaging model

The investigation of the stability of larger power systems requires knowledge about output impedances of switch mode power supplies (SMPSs). Methods for the prediction of this quantity of a SMPS comprise state-space averaging and sampleddata modeling. However, results obtained from both methods exhibit large differences. Furthermore, sampled-data modeling itself yields completely different results for different sampling points. Negative as well as positive incremental impedances may occur. Nevertheless, considerations of the steady-state in time domain prove the existence of these strongly different results, as discussed in this paper. In addition to that, it will be shown that only small additional effort is necessary to simulate the converter behavior with exact averaged quantities. This leads directly to an improvement of the state-space average model by means of adding an operating point dependent resistor.

Influence of component tolerances onto design and losses of resonant LLC converters

There exist several publications addressing and presenting an optimal design strategy for resonant LLC converters to end up with minimal ohmic losses. However, none of these papers takes unavoidable component tolerances into account. As a result, the specification cannot be met anymore for critical operation points. This paper proposes a new design strategy for resonant LLC converters, which incorporates these component tolerances. The resulting design provides minimal rms losses and guarantees the fulfillment of the converter specification under all circumstances at the same time.

Square wave based small signal modeling of frequency controlled converters

This paper discusses major aspects to be addressed for reliable determination of the transfer function of a frequency controlled converter through square wave perturbation. The square wave as a multi frequency perturbation allows to determine the transfer function with fewer simulations than classical methods. However, special care has to be taken in order to obtain valid frequency responses. Especially, matching pairs of distortion frequency and amplitude need to be derived. Furthermore accuracy of steady state determination and influence of aliasing effects must be considered. The simulation results are verified through measurements using a setup with the same perturbation principle.

Application specific efficiency improvement for an industrial point of load converter

Today, many electrical applications use the DC-DC buck converter topology as point of load (POL) converter with synchronous rectifiers to provide the desired output voltage for different loads and components [1-3]. While its efficiency is usually optimized for high output powers, in case of low and mid power it is often off the optimum. Nevertheless, since many applications force these converters to operate at part load condition most of the time, the systems overall power consumption can be reduced noticeably by optimizing the efficiency at its main operating point. This paper deals with the specific efficiency improvement of an existing industrial POL converter (Po,max = 60 W) towards its main operating point at part load with Po = 15 W. In a first step, the converter is analyzed in the time-domain by means of a sampled data modeling (SDM) approach in order to simulate and predict its behavior as accurate as possible. In a second step, based on the derived converter model the POL is optimized for Po = 15 W by simulations using the SDM model. Finally, practical measurements are performed in order to verify the correctness of the identified options for improvements.