I can now give a more useful answer than one winning converter name. The XL4015 and MP2307DN are commercial chips with different off-state paths and nominal switching frequencies. The XL4015 example uses an external diode; the MP2307DN controls a low-side MOSFET. I can follow that difference through the current path and the loss equations.
The conditions make the comparison more demanding. A 12-to-5 V heading does not make a 3.3 V substitution equivalent, and two different ripple allowances do not isolate frequency by themselves. That gives this study a concrete conclusion: explain the job, follow the model and keep the operating point attached to the result.
The current path is the clearest place to start. In the non-synchronous arrangement, a diode provides the path during the off interval. In the synchronous arrangement, a controlled low-side transistor takes that role. Once I follow those paths, the topology names stop feeling like words I have to remember and start describing a difference I can locate in a circuit.
The ideal duty-ratio calculation gives another piece of the picture. For 12 V in and 5 V out, the simplified relationship gives about 0.417. The useful part isn't the division. It’s understanding how the inductor current rises and falls over the intervals and why those changes balance in the steady ideal case. Then I know what conditions came with the equation.
The inductor calculation goes a little further. Keeping the same ideal voltages and a 0.9 A peak-to-peak ripple target, changing the frequency from 180 kHz to 340 kHz changes the calculated inductance from about 18.0 to 9.53 microhenries. They make one effect visible because the other conditions stay fixed.
That last detail has become important across the whole comparison. If I change the output voltage or allowed ripple in one column, I need to say so. Otherwise a difference in the answers can look like a property of the converter when it actually comes from changing the task. Looking back at the mismatched examples is making me pay more attention to the conditions before checking the arithmetic.
Losses give me another way to ask the question. The diode path has a forward-drop contribution. The MOSFET path has an on-resistance contribution. Their conducting times matter. Then there are other conduction, switching and inductor losses. Reducing one term can be useful without proving a universal efficiency claim about the entire module.
The maximum efficiency figure is therefore only a beginning. I want the input voltage, output voltage and load beside it. A curve can show more than a peak, but it still needs a source and stated conditions. If I can't say whether it is manufacturer data, a model or a measurement from the actual module, I shouldn't turn it into my own result.
The heat question has also moved earlier in my thinking. I used to put it after the circuit explanation, as if it were mainly a packaging concern. Now it follows directly from the power balance. Loss needs a path out of the components, and that physical path affects the temperature. One neat bar chart doesn't supply the temperatures or the rule for choosing a heatsink.
For the write-up, I'm organising the material around those questions rather than trying to make every page sound equally conclusive. A circuit explanation can be clear while a graph's source remains unresolved. A calculation can be correct under its assumptions without being a measured result. Keeping those distinctions visible makes it easier to see what the study actually supports.
My mentor's help belongs in that account. There are parts of the reasoning I can now work through and parts where I still need the assumptions checked. I don't want either to disappear when the report is tidied. The same goes for examples in the documentation. They can help me understand a step, but the actual circuit and source information decide whether that explanation fits this case.
A useful final check is whether I could talk somebody through one operating point without jumping to the conclusion. Start with the input and desired output. Explain the current paths. Work through the simple model. Identify the losses that matter. Then say which information is still missing before making a practical choice. That sequence is more useful than a broad claim that one technology is newer or more efficient.
There are still measurements that would strengthen the comparison, especially a matched efficiency sweep with a clear setup and records behind it. A matched sweep would add a different kind of result to the analytical comparison. The graphs and illustrations also need their origins kept clear.
The matched ideal example is useful alongside the report's differing cases. At 12 V input, 5 V output and 0.9 A peak-to-peak ripple, I calculate about 18.0 microhenries at 180 kHz and 9.53 at 340 kHz. That isolates the frequency term on paper. It does not select a component or measure a converter, but it shows the kind of question I can now explain. The answer has a condition, a mechanism and a limit, rather than only a smaller number.
The study has become more interesting as the easy answer has become harder to give. I can look at a small converter now and find several connected questions instead of just a voltage setting and a current rating. I don't think that makes choosing one impossible. It gives me a much better idea of what I need to know before I choose.