skip to content
arjun nair
projects
completed

Adjustable AC/DC Power Converter

Circuits class · Stanford

I built an adjustable AC/DC power supply from a 120 V AC wall outlet, finishing at roughly 0–20 V of adjustable output. The central problem was making an adjustable voltage source hold its output when a load is connected — solved with a MOSFET source-follower buffer, then verified with a multimeter and oscilloscope across several load resistances.

transformer + AC measurements

I soldered the power connector onto a prototype board for a 120 V to 24 V AC wall transformer. Using a function generator and oscilloscope, I tested its response at frequencies from 10 Hz to 1 MHz.

Supplied circuit block diagram showing a transformer, bridge rectifier, filter capacitor, and adjustable voltage regulator.
Conceptual circuit path from the supplied image, not the built schematic. Its approximately 12 V transformer label differs from my 24 V AC wall adapter.

At 60 Hz, a 10 V peak-to-peak sinusoidal input produced a 4.62 V peak-to-peak output: a measured ratio of about 2.16:1 under that test setup, separate from the transformer's nominal 120 V to 24 V rating.

Underside of a prototype board showing two soldered wires
The underside of the board: the transformer leads soldered through and dressed in, before the rest of the circuit went on top.
Lab 3 oscilloscope capture of sinusoidal transformer output at 59.975 Hz
Transformer output with the wall adapter powered, from Lab 3, Q11: 59.975 Hz and a 16.674 ms period. The scope reads 79.6 V peak-to-peak, corresponding to the report's 39.8 V amplitude; this is separate from the function-generator test above.

full-wave bridge rectifier

I soldered four 1N4002 silicon diodes into a bridge so both halves of the AC waveform produce the same output polarity. With a 10 kΩ load, the oscilloscope recorded 120.01 Hz, an 8.3328 ms period, and 39.4 V peak-to-peak. The roughly 120 Hz output comes from rectifying both halves of the 60 Hz input.

Across an individual diode, I measured a maximum forward voltage of approximately 1.0 V and a peak reverse voltage of 39.6 V.

Front of a prototype board with a barrel jack and four diodes in a bridge
The front of the board: the input jack for the transformer, with the four 1N4002 diodes soldered in as the bridge.
Lab 3 oscilloscope capture of full-wave rectifier output at 120.01 Hz
Voltage across the 10 kΩ load, from Lab 3, Q12: 120.01 Hz, an 8.3328 ms period, and 39.4 V peak-to-peak. Both halves of the AC input appear as single-polarity pulses.
Lab 3 oscilloscope capture of voltage across one bridge diode with negative peaks at minus 39.6 V
Voltage across one bridge diode, from Lab 3, Q13. The report records approximately +1.0 V maximum forward voltage and −39.6 V maximum reverse voltage, giving a 39.6 V peak inverse voltage.

capacitor filtering

I added a 1000 µF capacitor across the rectifier output to store charge between waveform peaks and reduce ripple. I checked its polarity, discharged it safely before modifying the circuit, and examined the output with the oscilloscope.

I've also worked through how capacitance and load resistance affect ripple — the tradeoff that drove the capacitor choice and, later, the output buffer.

Supplied illustrative capacitor-smoothed waveform.
Conceptual filtering illustration from the supplied image, not measured output. Its voltage scale does not establish the circuit's ripple or output voltage.

Zener regulation + adjustable output

With a nominal 24 V DC supply established, I added a 24 V Zener diode and a 1 kΩ series resistor, then integrated a 10 kΩ potentiometer as a voltage divider so the wiper position sets the output voltage.

The limitation was load regulation. The potentiometer made the voltage adjustable, but its relatively high Thévenin resistance meant a low-resistance load pulled the output voltage down — the supply couldn't hold its setpoint once it had to deliver real current.

Prototype board with rectifier bridge, electrolytic capacitor, Zener diode, and series resistor
The board after filtering and regulation went in: the 1000 µF capacitor sits below the bridge, with the 24 V Zener and 1 kΩ series resistor wired after it.
Prototype board with the 10 kΩ potentiometer added for adjustable output
Lab 5: the 10 kΩ potentiometer (blue, lower right) added to the board, making the output adjustable through its wiper position.

MOSFET output buffer

To fix the load regulation problem, I added an IRF530A N-channel MOSFET configured as a common-drain amplifier — a source follower. Its high input impedance isolates the potentiometer from the external load, while its low output impedance lets the circuit supply current without the voltage collapsing.

A 470 µF output capacitor suppresses ripple at the output, and a 10 kΩ output resistor gives the capacitor a discharge path. The completed design targets an adjustable range of roughly 0–20 V with much better stability under changing loads.

Completed prototype board with MOSFET, output capacitor, and potentiometer
Lab 6: the completed assembly — the IRF530A MOSFET and 470 µF output capacitor joined the earlier stages, with the potentiometer still setting the voltage.

measured performance

I evaluated the final circuit with a Keysight digital multimeter and oscilloscope, connecting external resistive loads through a breadboard and recording how output voltage and ripple changed as the load drew more current.

At an 18 V unloaded setpoint, the output held up well across three decades of load resistance:

loadoutput voltagepeak-to-peak ripple
10 kΩ17.929 V2.225 mV
1 kΩ17.587 V2.275 mV
100 Ω17.256 V9.175 mV

The same pattern held at lower setpoints: at 10 V unloaded, the output ran from 9.562 V at 10 kΩ down to 9.199 V at 100 Ω, and at 5 V unloaded, from 4.952 V down to 4.547 V. The voltage still droops as load current rises — this isn't a perfectly regulated supply — but the source follower keeps the droop to a few hundred millivolts where the potentiometer alone would have lost volts.

Bench with a Keysight multimeter reading 17.678 V DC connected to the converter and a breadboard load
Load testing on the bench: the Keysight multimeter reads 17.678 V DC with a resistive load connected through the breadboard.
Close-up of the converter board wired to a breadboard with a 100 Ω load resistor and oscilloscope probe
The experimental setup: the converter board in a vise, wired to the breadboard load (a 5 W, 100 Ω resistor) with the oscilloscope probe on the output.
arjun nair · stanford