To answer how do AC DC converters work with exact conversion math: when converting a standard 120V AC RMS mains supply to unfiltered DC using a full-wave bridge rectifier, the peak DC output is 168.3V. The exact formula used with values substituted is: V_DC(peak) = (V_AC(RMS) × √2) - (2 × V_f). Substituting standard silicon diode values: (120V × 1.414) - (2 × 0.7V) = 169.68V - 1.4V = 168.28V. This baseline assumes an ideal no-load condition with standard silicon junctions; under a real resistive load with a smoothing capacitor, the voltage will sag based on ripple current, capacitor ESR, and the specific discharge curve.
Rectifier Topologies and Output Formulas
The fundamental mechanism of any passive AC to DC converter is the rectifier, which forces alternating current to flow in only one direction. However, the exact DC output voltage and ripple characteristics depend entirely on the topology you choose. Below is the data-dense reference table for the four most common rectifier configurations used in bench power supplies and industrial motor drives.
| Rectifier Topology | Diode Count | V_DC Formula (No Load) | Ripple Freq (60Hz Mains) | Typical Application |
|---|---|---|---|---|
| Half-Wave | 1 | V_peak / π (0.318 × V_peak) | 60 Hz | Low-cost signal detection, simple heaters |
| Full-Wave Center-Tap | 2 | V_peak - V_f | 120 Hz | Legacy tube amplifiers, dual-polarity rails |
| Full-Wave Bridge | 4 | V_peak - (2 × V_f) | 120 Hz | Standard bench supplies, KBPC5010 modules |
| 3-Phase Bridge | 6 | 1.35 × V_LL (Average DC) | 360 Hz | Industrial VFDs, EV DC fast chargers |
Notice the diode forward voltage drop (V_f). In a full-wave bridge, current always passes through two diodes in series, meaning you lose 1.4V with standard silicon (e.g., 1N4007 or 1N5408). If you are designing a low-voltage converter (like a 5V AC to DC supply for an Arduino), that 1.4V loss is massive. In those cases, swap to Schottky diodes (like the 1N5822) where V_f drops to roughly 0.3V per diode, reducing your bridge loss to 0.6V. For a deeper dive into semiconductor behavior, the All About Circuits rectifier chapter provides excellent schematic breakdowns.
Line Variations: 120V vs 230V vs 3-Phase Shifts
A common mistake in power supply design is assuming the AC mains voltage is a fixed, universal constant. It is not. According to ANSI C84.1 standards, utility voltage can legally fluctuate by ±5% to ±10% depending on your region and grid load. The assumption that fixes your baseline answer is the nominal RMS voltage, but your smoothing capacitor must be rated for the absolute maximum peak voltage during a +10% or +20% surge.
Here is how the peak DC conversion shifts across a ±20% range for a nominal 120V AC system, assuming a full-wave bridge with silicon diodes:
| AC Mains Input (RMS) | Variance | Peak DC Output (No Load) | Average DC (Heavy Load) |
|---|---|---|---|
| 96V | -20% (Brownout) | 134.3V | ~115V |
| 108V | -10% | 151.3V | ~135V |
| 120V | Nominal | 168.3V | ~155V |
| 132V | +10% (Surge) | 185.2V | ~170V |
| 144V | +20% (Fault) | 202.2V | ~185V |
How the math shifts for global and industrial voltages:
- 230V AC (EU/UK/AU Nominal): The formula remains identical, but the numbers scale.
(230V × 1.414) - 1.4V = 323.8V DC peak. Your filter capacitors must be rated for at least 400V to safely handle +10% line surges without venting electrolyte. - 3-Phase 400V Line-to-Line: Three-phase rectification completely changes the ripple profile. The ripple frequency jumps to 360Hz (six times the 60Hz fundamental), which drastically reduces the required filter capacitance. The average DC output is calculated as
1.35 × V_LL, yielding a highly stable 540V DC bus, which is the exact intermediate bus voltage you will measure inside a modern Variable Frequency Drive (VFD) or solar inverter.
When Simple Conversion Math Becomes Meaningless
The linear rectifier math above applies strictly to passive, unregulated AC to DC converters (transformer + diode bridge + capacitor). If you are working with modern Switching Mode Power Supplies (SMPS) or Active Power Factor Correction (PFC) circuits, simple RMS-to-Peak conversion math becomes entirely meaningless for two reasons:
- Switching Regulators Decouple Input from Output: In a flyback converter or a buck-stage SMPS (like those built around the LM2596 or LNK306), the DC output voltage is not dictated by the AC peak. It is dictated by the high-frequency transformer turns ratio and the PWM duty cycle controlled by the feedback optocoupler. A 120V AC input can be converted to 5V DC or 400V DC depending entirely on the controller's duty cycle.
- Unknown Power Factor (PF): If you are trying to calculate the input current draw of an AC/DC converter based on its DC output wattage, the math fails if the Power Factor is unknown. A cheap, capacitor-dropper or uncorrected bridge supply might have a PF of 0.55, meaning it draws nearly twice the apparent current (VA) compared to its real power (W). Modern supplies use Active PFC controllers (like the TI UCC28180) to force the input current waveform to track the input voltage waveform, pushing the PF to >0.95. Without knowing the PF, your breaker sizing and wire gauge calculations for the AC side will be dangerously undersized.
For comprehensive schematic examples of both passive and active filtering stages, the Electronics Tutorials full-wave rectifier guide remains a benchmark reference for visualizing the ripple waveforms.
Frequently Asked Questions
Why does my multimeter read lower DC voltage than the formula predicts?
Standard multimeters read the average or RMS equivalent of a rippled DC waveform, not the absolute peak. If your filter capacitor is too small (e.g., 100µF on a 1A load), the voltage drops significantly between the 120Hz peaks. To measure the true peak, you need an oscilloscope with AC+DC coupling enabled.
Can I use a bridge rectifier directly on 120V AC without a transformer?
Electrically, yes; it will output ~168V DC. Practically, this is extremely dangerous for hobbyist projects because the resulting DC bus is not galvanically isolated from the lethal AC mains. A single fault can energize your low-voltage microcontroller chassis with 168V DC. Always use an isolation transformer for bench testing.
How do I calculate the required smoothing capacitor size?
Use the formula: C = I_load / (f × V_ripple). For a 1A load on a 60Hz full-wave bridge (f = 120Hz) where you can tolerate 10V of ripple: C = 1 / (120 × 10) = 0.00083 Farads, or 830µF. You would select the next standard size up, such as a 1000µF electrolytic capacitor rated for at least 200V.
