How Buck Converters Work: Feedback Control Explained
Understanding how the LM2596S-ADJ controls output voltage through feedback - it's like an op-amp, but with switching!
The Question
When looking at the LM2596S-ADJ circuit, a natural question arises:
"This is a DC-DC converter, right? We input 15V, but how does it output 13.5V? Does it continuously monitor the FB (feedback) pin and control the voltage? Is this similar to an op-amp?"
Short answer: Yes! It's exactly like an op-amp comparator, but instead of adjusting output voltage continuously, it adjusts the duty cycle of an internal switch.
What is a Buck Converter?
A buck converter (step-down converter) is a DC-DC switching regulator that converts a higher input voltage to a lower output voltage with high efficiency.
Buck Converter vs Linear Regulator
Linear Regulator (like LM7812):
15V ──┬──[Transistor]──┬──→ 12V
│ (Variable │
│ Resistance) │
Heat! Load
Power loss = (15V - 12V) × Current
= 3V × 1A = 3W of heat! 🔥
Efficiency ≈ 80%
Buck Converter (like LM2596S-ADJ):
15V ──┬──[Switch ON/OFF]──[Inductor]──┬──→ 13.5V
│ 150kHz PWM │
│ │
GND Load
Power loss ≈ 10-15% = 0.2W of heat ❄️
Efficiency ≈ 85-90%
Key difference: Linear regulators waste excess voltage as heat. Buck converters use high-frequency switching to efficiently "chop" the voltage down.
The Feedback Control Loop (Like an Op-Amp!)
Internal Block Diagram
LM2596S-ADJ Internal Circuit
┌──────────────────────────────────────┐
│ │
15V ────┤ VIN │
│ │ │
│ ├──→ [Power Switch] ──→ VOUT ──────┤──→ To Inductor
│ (MOSFET) 4 │
│ ↑ │
│ │ │
│ ┌────────┴─────────┐ │
│ │ PWM Control │ │
│ │ Logic │ │
│ └────────┬─────────┘ │
│ │ │
│ ┌────────┴─────────┐ │
│ │ Error │ │
│ │ Amplifier │ │
│ │ (like op-amp) │ │
│ └───┬──────────┬───┘ │
│ │ │ │
│ [1.23V] FB ───────────────────┤──→ To voltage divider
│ (Internal 2 │
│ Reference) │
└──────────────────────────────────────┘
The Comparison: Op-Amp Analogy
Your intuition is 100% correct! The LM2596S-ADJ uses an error amplifier that works just like an op-amp:
| Op-Amp Circuit | LM2596S-ADJ Buck Converter |
|---|---|
| Non-inverting input (+) | Internal 1.23V reference |
| Inverting input (-) | FB pin (pin 2) |
| Error amplifier | Compares FB to 1.23V |
| Output adjustment | Changes PWM duty cycle |
| Goal | Make inputs equal |
| V(+) = V(-) | FB = 1.23V |
Op-Amp Voltage Regulator:
┌──────┐
┌────┤+ │
│ │ Op- ├──── Vout (continuous voltage)
│ ┌──┤- Amp│
│ │ └──────┘
│ │ ↑
[Vref] [Feedback from voltage divider]
LM2596S-ADJ:
┌──────┐
[1.23V]─┤+ │
│ Error├──→ PWM ──→ Switch (150kHz)
┌────┤- Amp│
│ └──────┘
│ ↑
[Feedback from voltage divider via FB pin]
How Feedback Sets the Output Voltage
The Magic Formula
The LM2596S-ADJ has an internal 1.23V reference. The chip tries to keep the FB pin at exactly 1.23V.
Circuit:
R1 (10kΩ)
+13.5V output ──────┬────────────┬──── FB pin (pin 2)
│ │
│ R2 (1kΩ)
│ │
│ GND
│
[Voltage divider makes FB = 1.23V]
Voltage divider equation:
V_FB = V_OUT × (R2 / (R1 + R2))
The chip forces: V_FB = 1.23V
Therefore:
1.23V = V_OUT × (R2 / (R1 + R2))
Solving for V_OUT:
V_OUT = 1.23V × (R1 + R2) / R2
= 1.23V × (1 + R1/R2)
Example (our +13.5V circuit):
V_OUT = 1.23V × (1 + 10kΩ/1kΩ)
= 1.23V × (1 + 10)
= 1.23V × 11
= 13.53V ✓
Want Different Output Voltage?
Just change the resistor ratio!
| Target Voltage | R1 | R2 | Calculation |
|---|---|---|---|
| 3.3V | 1.7kΩ | 1kΩ | 1.23V × (1 + 1.7) = 3.32V |
| 5V | 3.1kΩ | 1kΩ | 1.23V × (1 + 3.1) = 5.04V |
| 7.5V | 5.1kΩ | 1kΩ | 1.23V × (1 + 5.1) = 7.50V |
| 12V | 8.7kΩ | 1kΩ | 1.23V × (1 + 8.7) = 11.94V |
| 13.5V | 10kΩ | 1kΩ | 1.23V × 11 = 13.53V |
The Control Process: How It Maintains Voltage
Let's see what happens when the output voltage changes:
Scenario 1: Output Voltage Drops (Load increases)
Step 1: Heavy load pulls voltage down
+13.5V → drops to 13.2V
Step 2: Voltage divider responds
FB pin = 13.2V × (1kΩ/11kΩ) = 1.20V
FB is now LESS than 1.23V reference!
Step 3: Error amplifier detects difference
Error = 1.23V - 1.20V = +0.03V (positive error)
"Need more output voltage!"
Step 4: PWM controller increases duty cycle
Was: 90% duty cycle
Now: 92% duty cycle
(Switch stays ON longer)
Step 5: More energy delivered to output
Inductor current increases
Output voltage rises back to 13.5V
Step 6: System stabilizes
FB pin returns to 1.23V
Error = 0
Duty cycle adjusts to maintain balance ✓
Scenario 2: Output Voltage Rises (Load decreases)
Step 1: Light load allows voltage to rise
+13.5V → rises to 13.8V
Step 2: Voltage divider responds
FB pin = 13.8V × (1kΩ/11kΩ) = 1.25V
FB is now MORE than 1.23V reference!
Step 3: Error amplifier detects difference
Error = 1.23V - 1.25V = -0.02V (negative error)
"Need less output voltage!"
Step 4: PWM controller decreases duty cycle
Was: 90% duty cycle
Now: 88% duty cycle
(Switch stays OFF longer)
Step 5: Less energy delivered to output
Inductor current decreases
Output voltage drops back to 13.5V
Step 6: System stabilizes
FB pin returns to 1.23V
Error = 0
Duty cycle adjusts to maintain balance ✓
How Switching Creates Lower Voltage: Duty Cycle
What is Duty Cycle?
Duty cycle is the percentage of time the internal switch is ON during each cycle.
One switching period (1/150kHz = 6.67µs):
Switch ON for 6µs, OFF for 0.67µs
Duty Cycle (D) = ON time / Total time
= 6µs / 6.67µs
= 90%
Output Voltage = Input Voltage × Duty Cycle
= 15V × 0.9
= 13.5V ✓
Visualizing the Switching
Time axis (microseconds) →
0 1 2 3 4 5 6 6.67
Switch state:
ON ████████████████████████████ OFF ██
(6µs = 90% duty cycle) (0.67µs)
Voltage at VOUT pin (before inductor smoothing):
15V ┐ ┌─────────────────────┐ ┌─
│ │ │ │
0V └──┘ └─────┘
After inductor smoothing:
15V ┐
│ ───────────────────────────────
13.5V (Average = 15V × 0.9 = 13.5V)
│
0V ┘
The inductor acts like a flywheel, smoothing the chopped voltage into steady DC!
The Complete Buck Converter Circuit
LM2596S-ADJ (U2)
┌─────────────────┐
+15V ───────────────┤5 VIN │
│ │
┌──────┤3 ON │
│ │ │
│ │ VOUT ├4───┬─→ L1 (100µH) ──┬─→ +13.5V
│ │ │ │ │
│ │ FB ├2───┼────────────────┤
│ │ │ │ │
│ │ GND ├1─┐ │ │
│ └─────────────────┘ │ │ │
│ GND│ │
┌───┴───┐ │ │
│ C5 │ 100µF │ │
│ Input │ │ │
└───┬───┘ │ │
GND │ │
│ │
D1 (SS34) │ │
Schottky ────┘ │
(Flyback) │
│ │
GND │
│
Feedback Divider: │
R1 (10kΩ) │
┌────────────┬───────────────────────┘
│ │
│ R2 (1kΩ)
│ │
│ GND
│
Sets: V_FB = V_OUT × (R2/(R1+R2))
= 13.5V × (1k/11k)
= 1.23V ✓
Component Roles
| Component | Function |
|---|---|
| L1 (100µH) | Stores energy when switch is ON, releases when OFF (smooths output) |
| D1 (Schottky) | Provides current path when switch is OFF (flywheel effect) |
| C5 (input cap) | Stabilizes input voltage, handles transient currents |
| C3 (output cap) | Filters ripple, provides stable output voltage |
| R1, R2 | Voltage divider - sets output voltage by creating 1.23V at FB pin |
Why Buck Converters Are Better Than Linear Regulators
Power Loss Comparison
Our circuit: 15V → 13.5V at 1.3A
Linear Regulator (LM7812 equivalent):
Power in = 15V × 1.3A = 19.5W
Power out = 13.5V × 1.3A = 17.6W
Power loss = 19.5W - 17.6W = 1.9W 🔥
Efficiency = 17.6W / 19.5W = 90%
(Still wastes 1.9W as heat)
Heat sink required: Large (to dissipate 1.9W)
Buck Converter (LM2596S-ADJ):
Power in = 15V × 1.3A = 19.5W
Power out = 13.5V × 1.3A = 17.6W
Internal loss ≈ 10% = 1.95W (efficiency loss)
Power loss ≈ 0.2W ❄️
Efficiency = 90% (typical for buck converters)
Heat sink required: Minimal or none
Result: Buck converter runs much cooler and wastes less energy!
Why We Use Both in This Project
Power Supply Architecture:
Stage 1: USB-PD (15V) → DC-DC Buck → 13.5V
(High efficiency, large voltage drop)
Stage 2: DC-DC (13.5V) → Linear Regulator → 12V
(Ultra-low noise, small voltage drop)
Best of both worlds:
✓ Efficient voltage reduction (buck converter)
✓ Ultra-clean output (linear regulator)
✓ Low overall heat generation
✓ Less than 1mVp-p ripple (perfect for audio circuits)
Key Takeaways
- Buck converters are switching regulators that use PWM to efficiently reduce voltage
- Feedback control works like an op-amp comparing FB pin to internal 1.23V reference
- Voltage divider sets output voltage by creating 1.23V at FB pin
- Duty cycle determines output: V_OUT = V_IN × Duty Cycle
- Much more efficient than linear regulators for large voltage drops
- The chip constantly adjusts duty cycle to maintain FB = 1.23V
Common Mistakes to Avoid
❌ Wrong: "The LM2596S-ADJ outputs 15V and we use resistors to drop it to 13.5V"
- This would waste power like a linear regulator!
✅ Correct: "The LM2596S-ADJ uses PWM switching to create 13.5V directly. The resistors just tell it what voltage to target (by creating 1.23V at FB)."
❌ Wrong: "R1 and R2 pass current from the output"
- The feedback resistors carry almost no current (less than 1mA)!
✅ Correct: "R1 and R2 form a voltage divider that samples the output voltage and reports it to the FB pin."
See Also
- LM2596S-ADJ Documentation - Full component specifications
- Circuit Diagrams - See buck converters in context
- Open-Drain Outputs - Another control mechanism explained