Key Takeaways
- Open-loop = 'blind' control, no sensing, no correction
- Closed-loop = sense Vout + compute D + auto-adjust
- Formula: D = Vout_target / VIN → constant output under any VIN
Lesson 3
Conclusion: D Must Be Auto-Adjusted
Open-loop is 'blind' — doesn't know how far output has drifted. Solution: use a feedback loop to detect Vout in real time and adjust D.
Combining Lessons 1 and 2: load changes cause DCM deviation, VIN changes cause linear drift — with both present simultaneously, open-loop is completely helpless.
Root cause: open-loop system doesn't 'see' the output, doesn't 'know' how far it drifted, and never auto-corrects. The solution is intuitively simple: add a 'feedback path' from output back to PWM control — measure Vout, compute needed D = Vout_target/VIN, auto-set PWM width.
The next chapter will show: as long as D adjusts correctly with VIN, output can stay constant under all conditions.
Monte Carlo: Output Spread Under Random Load
R_load~N(50,σ≈13)Ω, 100 samples. Output spreads from 6.0V to 7.8V (σ≈0.44V). Open-loop unable to suppress load variation.
After reading this section, run the simulation and observe the waveforms. To explore further, open the example in a standalone page.
Watch Items
- Monte Carlo spread plot visualizes open-loop output uncertainty
- Think: if D could auto-adjust from 60% to 37.5%, output would stay at 6V