In this section, frequency responses are validated with the experimental results of the versatile buck–boost converter prototype. Efficiency power conversion results of the BBV converter as part of a composite architecture (Composite converter A, see
Figure 6) and a classical noninverting buck–boost converter (BB) (Composite converter B, see
Figure 7) are compared. The experimental and HIL (Hardware in the loop) testing are developed to verify the proposed composite architecture. The HIL test has been split into two subsystems corresponding to the plant and the controller. On the one hand, the plant subsystem corresponds to the composite converter (A based on the BBV converter or B based on the BB converter) and modeled on the PLECS RT Box 1 for HIL evaluating. On the other hand, the controller subsystem is implemented in an inexpensive signal controller (DSC) from Texas Instruments LAUNCHXL-F28069M. The controller consists of a double loop control algorithm to regulate the dc–dc converter’s output voltage (BBV or BB). The BBV converter design was realized following the steps shown in Algorithm 1 for the input parameters:
A,
W, for boost mode (
V and
V),
μs, maximum current
A, and
V. Therefore, the selected components are listed in
Table 3.
4.2. Hardware in the Loop (HIL) Validation
The double loop control strategy implemented using HIL has been validated with experimental tests. In
Figure 9 and
Figure 10, the system’s response can be observed when the output voltage reference changes ±20 V in a step form.
The figures show experimental and HIL resulting waveforms of the input, output current, and output voltage when the input voltage
V and a constant resistive load
200 Ω from
Figure 4. For buck mode results,
Figure 9 illustrates the signal waveforms when the output voltage reference changes between 100 V and 120 V. Note that the output current reference is tracked perfectly. In
Figure 10, the results of the converter working in boost mode can be observed, and where the output voltage reference changes between the values of 294 V and 314 V, the results are shown in the ac component to appreciate the output voltage variation in the boost mode. A good agreement between the experimental and HIL results is observed, which validates the correct design of the HIL system.
The current loop response for the BBV is shown in
Figure 11. This figure shows HIL results with a step change of the current reference of ±4 A for buck and boost mode, respectively. The input voltage is set in 300 V, the output voltage is
V for boost mode and
V for buck mode. As shown, the output current is well regulated. For buck mode, the peak-to-peak output current ripple (
) is near 3 A, and the output current ripple value is near 1.5 A for boost mode. Thus, the proposed current control strategy performance during current step reference change is also validated.
4.3. BBV in Composite Converter
Finally, tests of the BBV converter as part of the composite converter were performed. The rated operating conditions of each module and the composite converter are listed in
Table 5.
Transient responses are realized by a step change in dc bus voltage (
) of the composite converter A (
Figure 6) with a
of 300 V. The dc bus voltage is changed from 900 V to 1000 V and from 1000 V to 900 V for boost mode in
Figure 12a,b, whereas the
is changed from 700 V to 800 V and from 800 V to 700 V for buck mode shown in
Figure 12c,d. As it is shown, the output voltage followed the reference quickly for both operation modes.
Figure 13 shows the dc bus voltage for the composite converter A while the
is regulated to 900 V, and the battery voltage (
) changes from 360 V to 220 V to emulate a discharge. In this case, the buck–boost converter’s output voltage
and the DCX’s output voltage (
) increase when the battery voltage is being discharged to keep the output voltage regulated at 900 V. It can be observed that the dc bus voltage is tightly controlled to its desired value, showing a soft transition from DCX+buck mode to DCX+boost mode.
Figure 14 and
Figure 15 show the transient response front of the 50% step change in the composite converter’s current load. In the case of the buck mode, the dc bus voltage (
) is regulated to 800 V, and the battery voltage (
) has a voltage of 300 V while the power load changes between 1600 W and 3200 W (see
Figure 14). In the boost mode, the dc bus voltage is regulated to 1000 V, and the battery voltage (
) has a voltage of 300 V while the power load changes between 2000 W and 4000 W (see
Figure 15). In these experiments, an outstanding regulation of the inner DSMCC loop is demonstrated, evidenced by a zero steady-state error in the dc bus voltage.
4.4. Efficiency Results Comparison
In this section, a comparison has been made regarding the efficiency simulated results between the composite converter using a BBV converter (composite converter A) and the composite converter using a BB converter (composite converter B). The BBV converter for high-voltage application was presented in [
28], and its parameters are listed in
Table 3. The BB converter was designed under the same characteristic as the BBV converter, and the control strategy presented in
Section 3 was implemented for both composite converters. The efficiency results are acquired using the tool PLECS through its thermal model. This simulation is carried out using the heat sink components for the power device SCT2450KEC employed for the dc–dc converters and the DCX module. The conduction, turn-on, and turn-off switching losses are obtained from the datasheet. These were defined as simulation parameters as shown
Figure 16.
The thermal simulation for the power conversion efficiency of the BBV converter under different input and output voltages was validated using an experimental measurement with the converter prototype shown in
Figure 17a. The thermal model of PLECS was validated with efficiency results presented in [
28] for the versatile buck–boost converter. The efficiencies were measured using a Yokogawa WT 3000 precision power analyzer connected at the input and the output of the converter and were taken with the converter working with
A as shown in
Figure 17a.
Figure 17b calculates different relative error points of simulated and experimental power conversion efficiency results for the versatile buck–boost converter. These results indicate a small relative error between the thermal simulation and the experimental results in all the converter’s operation points. Therefore, the thermal model provided by the PLECS simulation is a correct approximation to compare efficiency between the composite converters A and B. The efficiency results for composite A which is based on the BBV converter are presented in
Figure 6, and the efficiency results for composite converter B based on the BB converter are shown in
Figure 7. In these figures, the efficiency of the power conversion is included in each module that forms the composite converter (dc–dc converter and the dc transformer) and the overall efficiency for all the possible operation points. As can be seen in
Figure 6, the BBV converter has higher efficiency with a conversion ratio close to one
), while for the case of the DCX module, this occurs for an output voltage higher than 200 V. The dc–dc converter was designed for a maximum output current (
) of 8 A and a maximum output voltage (
) of 400 V. The composite converter has a conversion ratio
when the output voltage of the dc–dc converter (
) is 0 V (
). The maximum conversion ratio of the composite converter (
) depends on the dc–dc converter’s maximum conversion ratio (
) and the DCX turns ratio selected
. Consequently, the composite converter’s maximum ratio is five (
) as long as it does not exceed the maximum dc voltage
corresponding to 1100 V.
Figure 6 shows the limitation for all the possible operation points.
In the composite converter A shown
Figure 6, the DCX corresponds to a Half-Bridge LLC resonant converter with a resonant inductor (
) and capacitor (
) in series and with a relatively small magnetizing inductance (
) in the transformer. The DCX module is designed following the step proposed in [
40] for an output power of 3.2 kW and a commutation frequency of 100 kHz. The same power conversion efficiency presented for the BBV converter in
Figure 6 (composite converter A) was carried out for the composite converter B based on the BB converter shown in
Figure 7. The composite converter A achieves the highest power conversion efficiency in the vast majority of operation points as shown in
Figure 18, where the efficiency difference between both composite systems (
) is depicted. Outstanding efficiency improvements of up to 6% are achieved if a BBV converter is selected instead of a classical BB one as shown in
Figure 18. Power-loss breakdown results for the noninverting buck–boost converters are shown in
Figure 19 and
Figure 20. These results relate to the overall efficiency results shown in
Figure 6 and
Figure 7 for the buck–boost converters. In
Figure 19 and
Figure 20, the size of the pie charts are proportional to the power loss. The power inductor losses are calculated as in [
41], and the MOSFET conduction and switching losses are provided by the thermal model of the PLECS simulation. For the BBV converter, the total loss is composed of the MOSFET conduction loss, the MOSFET switching loss, the coupled inductor loss, and the damping network loss. The average value for the results in
Figure 19 are 59.6% of power dissipation is associated to conduction loss, 20.6 % to MOSFET switching loss, 15.8% to coupled inductor loss, and 4% to losses in the resistor of the damping network. For the BB converter, the total loss is composed of 73.6% of MOSFET conduction, 9.5% of MOSFET switching, and 16.9% of magnetic loss. From results presented in
Figure 19 and
Figure 20, it can be observed that the BBV converter has an average percentage of 4.9665% of the total input power in loss for all the operating points compared to 6.4058% of the total input power in loss for the BB converter, demonstrating a higher power conversion for the BBV converter.