Chapter 7
Conclusions
In this thesis an investigation of a DC-DC Converter was completed, in order to find the efficiency at several point of steady-state operation. The efficiency maps were also drawn. One significant problem for this kind of simulation was their speed. The speed of the simulation was influenced by different factors, mainly:
• The high switching frequency • The large number of simulations
For this reason the efficiency was calculated for only one switching period of the steady state. To do that, the simulations were run using the initial values of the storage elements (Capacitors and inductor) of one switching period at the steady state. In order to reach a more precise value of the initial condition, the formulae of the buck and the boost converter were rewritten considering the parasitic component resistance.
Efficiency Maps were drawn in three different level of accuracy:
Theoretical Using the formulae with the parasitic resistance, considering the mean values of the currents and the voltages (switching losses are neglected and the value are considered constant)
Simulink Consideration of the differential equations of the restive circuit model (switching losses are neglected)
Simplorer Consideration of a detailed model of the circuit.
The maps show that the current has the biggest impact on the losses. The diode gives larger losses than the Mosfet. For that reason in the buck operation mode, the lines of the efficiency are inclined so that for the same load current, the larger the duty cycle (and hence
the shorter the diode conducting time) the greater the efficiency. In the boost operation the lines of the efficiency are inclined like the buck, but for another reason: The larger the duty cycle (hence the difference between Vin and Vload), the larger the current in the device
(Iind= Ilaod/(1 − d)), and thus the greater the losses.
A brief investigation of the three phase converter was also completed, in order to investigate the ripple output voltage and on the ripple inductance current. A few simulations were run, but the advantages are recognizable, especially in some points of operation (d = 1/3, d = 2/3). The switching losses are larger because there is more switching, but they can be reduced by choosing smaller components. Moreover, to reach the same ripple, the switching period can be made smaller.
