Key Takeaways
- Converter theory holds when students run a model of each topology in the same week they derive it.
- A term built on six two-week blocks, each adding one new idea to a familiar circuit, gives students judgment as well as vocabulary.
- Instructors get further by converting one block per term and grading the comparison between prediction and simulated result.
A power electronics course teaches more when every converter derivation arrives with a model your students can run that same week.
Most syllabi still put the lab at the end of term. Students meet converter behaviour weeks after the maths that explains it, and the two never quite connect. A meta-analysis of 225 studies of undergraduate science and engineering classes found that students in traditional lecture sections were 1.5 times more likely to fail than students in active learning sections. Lecture first and lab last is the same pattern on a 12-week delay.
The fix is structural. Put an editable converter model beside each topic and let simulation work carry the term, with lectures setting up what each model will show. Your students will predict a waveform, run it, and explain the gap before the next class.
“A student who has derived four converters and simulated all four understands more than one who has derived seven and simulated none.”
Why converter waveforms belong in the same week as the derivation
Students keep a converter derivation when they see its switching waveform within days of the lecture. Volt-second balance on an inductor stays an abstract equation until a ripple current trace confirms it. Pairing each derivation with a runnable model gives the maths something to predict.
Take the buck converter. You derive the duty ratio relationship on Tuesday, and on Thursday students open the model, move the duty ratio from 0.4 to 0.6, and watch the output voltage and inductor ripple respond. They’ll also see what the averaged equation hides, such as the startup overshoot and the ripple on the capacitor. When the lab sits in week 10, that prediction and its check are separated by two months and a midterm.
The cost is lecture time. You’ll drop a derivation or two to make room, and that’s a fair trade. A student who has derived four converters and simulated all four understands more than one who has derived seven and simulated none.
What topics belong in a power electronics syllabus
A first syllabus needs six blocks. Those are switch behaviour with losses, direct current conversion, magnetics with filtering, rectification, inversion with pulse width modulation, and closed loop control. Each block maps to a converter family your students can model, which keeps the theory tied to a circuit they have run themselves.
Losses are the block most often cut, and they’re the one that makes the rest feel like engineering. Once students have a working buck model, ask them to add conduction and switching losses to the switch and estimate the heat it must shed at 20 kilohertz and again at 100 kilohertz. That single exercise connects device physics, efficiency, and thermal limits.
Leave resonant converters, multilevel topologies, and grid interaction for a second course. A first syllabus that tries to touch everything gives students vocabulary without judgment. Six blocks covered with a model each will serve them better than ten covered on slides.
How to sequence converter topics across a 12-week term
Sequence converter topics so that each week adds one new idea to a model students already trust. Start with a single switch and a resistive load, add energy storage, then a second operating mode, then alternating current, then feedback. Every model should reuse most of the previous week’s circuit.
The boost converter shows how this works. It arrives after two weeks on the buck, so the inductor, the switch, and the ripple calculation are familiar. The only new idea is that energy reaches the load while the switch is off, and students can spend the whole week on that one point. Many courses open with line frequency rectifiers because the textbooks do. Commutation overlap and discontinuous source current are harder to reason about than a buck converter, so rectifiers work better once students can already read a switching waveform.
A week-by-week course map with a model per topic
A 12-week course map works best as six two-week blocks, each with one editable model that students open in lecture and modify for the assignment. The first week of each block covers the derivation and a guided run. The second week hands students a design question that the model can answer.
| Weeks | Topic covered in lecture | Editable model beside the topic |
| Weeks 1 to 2 | Switch behaviour, conduction loss, and switching loss | A single switch model feeding a resistive load, with device parameters students can edit |
| Weeks 3 to 4 | Buck conversion, ripple, and conduction modes | A buck converter model where duty ratio and inductance set the ripple students predicted |
| Weeks 5 to 6 | Boost and buck-boost conversion with inductor and filter sizing | A boost converter model that students resize to meet a ripple target |
| Weeks 7 to 8 | Rectification, capacitor filtering, and input current quality | A single-phase diode bridge rectifier model with an adjustable filter capacitor |
| Weeks 9 to 10 | Inversion and pulse width modulation | A three-phase inverter model that shows how modulation index shapes the output spectrum |
| Weeks 11 to 12 | Closed loop voltage control and the final project | A closed loop buck converter model with a compensator students tune themselves |
The models in this map are standard converter topologies, so you can build each one from a component library and keep it for future terms. SPS SOFTWARE models run inside MATLAB/Simulink with open equations and editable parameters, which means the file you demonstrate in lecture is the same file students modify afterward. That matters most in weeks 1 and 2, when students need to see exactly which parameter produces the loss they calculated.
What students should build in a power electronics course

Students should build a sequence of working models that each answer a design question, ending with one converter they’ve specified themselves. A model counts as built when the student has predicted its behaviour on paper, run it, and explained any difference. Copying a finished file and pressing run doesn’t meet that bar.
Five deliverables cover the term.
- A switch loss estimate checked against simulated conduction and switching losses.
- A buck converter sized to meet a stated ripple and load specification.
- A boost converter with an inductor chosen to stay in continuous conduction.
- A rectifier or inverter study that reports harmonic content at the terminals.
- A closed loop converter with a tuned compensator and a documented load step response.
Grade the comparison between prediction and result, since that’s where the learning shows. Simulation work won’t teach probing, soldering, or layout. Keep a bench session if your department has the hardware, and place it after students have modelled the same circuit so their time at the bench goes toward measurement.
What separates an advanced power electronics course from an introductory one
An advanced power electronics course moves from single converters in isolation to converters that interact with a wider system and with their own limits. Students stop asking what a topology does and start asking how it behaves when the grid is weak, the load steps, or the devices heat up.
Grid connection deserves the largest share of that second course, because a United States Department of Energy presentation projected that 80% of electricity could flow through power electronics by 2030. Graduate students can take the three-phase inverter from weeks 9 and 10, add an output filter, a phase-locked loop, and a current controller, then study how the control loop responds as grid impedance rises. The introductory model becomes the starting point for a system study.
The same reuse applies to modelling choices. Advanced students should compare a switched model against an averaged one and decide which suits the question, because picking the right level of detail is a professional skill. Expect fewer topics and longer assignments than the first course carries.
How instructors bring simulation models into next term’s syllabus
The practical way to rebuild a course around simulation is one block at a time. Pick the topic your students struggle with most, place a model beside it next term, and move that week’s assignment from a problem set to a predicted and simulated result. You’ll know within one cohort if the timing helps.
Restructuring a full term in one pass is tempting and usually stalls. A single well-run block teaches you how much lecture time a model needs, how students handle the software, and which parameters they reach for first. Courses that improve year over year tend to belong to instructors who change a little each term and keep what worked.
Access is the other practical question. SPS SOFTWARE has an educational offer that begins with a request from the professor teaching the course, and an onboarding guide walks you through the first steps once that request is made. Start there if the models in this map fit your syllabus. What you’re really building is a habit in your students, the habit of checking every equation against a waveform, and that habit outlasts any single topic you teach.
“Courses that improve year over year tend to belong to instructors who change a little each term and keep what worked.”


