LM2596 Output Voltage Calculator
Adjustable version, based on VOUT = 1.23 V × (1 + R2 / R1)
| Ideal R2 | – |
|---|---|
| Nearest E24 value for R2 | – |
| Actual output with that R2 | – |
| Next E24 value down / up | – |
–
The adjustable LM2596 can regulate outputs from 1.23 V to 37 V.
Keep the input voltage a few volts above the target output so the regulator has enough headroom to stay in regulation.
The LM2596 is a nonsynchronous step-down (buck) switching regulator that converts an input of up to 40 V down to a lower, regulated output at load currents up to 3 A. It runs from a fixed 150 kHz internal oscillator, needs only four external components in its basic configuration, and ships in four versions: fixed 3.3 V, 5 V, and 12 V outputs, plus an adjustable version (LM2596-ADJ) that covers 1.2 V to 37 V. The adjustable version is the one you have almost certainly seen: it is the regulator at the center of the low-cost “blue module” buck converter boards sold everywhere.
| Parameter | Value |
|---|---|
| Input voltage range | 4.5 V to 40 V (45 V absolute maximum) |
| Output versions | 3.3 V, 5 V, 12 V fixed; adjustable 1.2 V to 37 V |
| Output tolerance | ±4% maximum over line and load |
| Output current | 3 A |
| Switching frequency | 150 kHz ±15% |
| Feedback reference voltage | 1.23 V (adjustable version) |
| Switch current limit | 4.5 A typical |
| Quiescent current | 5 mA typical operating; 80 µA typical in shutdown |
| Efficiency | 73% (3.3 V out, 12 V in, 3 A); 80% (5 V out, 12 V in, 3 A); 90% (12 V out, 25 V in, 3 A) |
| Packages | 5-pin TO-220 (NDH), 5-pin TO-263 (KTT) |
Every number above comes from the current TI datasheet, revision SNVS124G. Keep the original open while you design: download the LM2596 datasheet (PDF)
One naming note before anything else. Texas Instruments and onsemi both sell a part called LM2596 with the same pinout and the same basic ratings, but they are separate products with separate datasheets. This article follows the TI document. If your board uses the onsemi part, verify against the onsemi datasheet before relying on any limit.
LM2596 Pinout Explained
Both packages use the same five-pin assignment.

| Pin | Name | Function |
|---|---|---|
| 1 | VIN | Positive supply input. Needs a low-ESR bypass capacitor close to the pin. |
| 2 | Output | The internal switch. Swings between roughly VIN and −0.5 V at 150 kHz. Keep the copper on this node small. |
| 3 | Ground | Circuit ground. |
| 4 | Feedback | Senses the output voltage and closes the regulation loop. |
| 5 | ON/OFF | Logic shutdown input. |
The ON/OFF pin trips up more first builds than any other detail, because its logic is inverted from what the name suggests. Pulling the pin below about 1.3 V turns the regulator on; pulling it above 1.3 V (up to a 25 V maximum) shuts it down and drops the supply current to about 80 µA. If you do not need shutdown, tie the pin to ground or leave it open — either way the regulator runs. A board that stays dead until you ground pin 5 is behaving exactly as designed.
How the LM2596 Buck Converter Works
A buck converter regulates by switching, not by burning off the excess voltage as heat. The internal transistor chops the input at 150 kHz with a duty cycle of roughly V<sub>OUT</sub>/V<sub>IN</sub>. The inductor and output capacitor average that chopped waveform into a smooth DC level, and the catch diode carries the inductor current during the portion of each cycle when the switch is off.

The feedback pin watches the output — directly on fixed versions, through a resistor divider on the adjustable version — and the control loop adjusts duty cycle to hold it at the target.
This is why a switcher beats a linear regulator so decisively when the voltage step is large. A linear regulator dropping 12 V to 5 V at 1 A dissipates 7 W in the pass element, no exceptions; the LM2596 doing the same job runs around 80% efficient and dissipates a fraction of that. The trade is switching ripple on the output and more care in layout. The LM2596 is nonsynchronous — the low-side element is a diode, not a second transistor — which caps its efficiency below modern synchronous parts, but keeps the external circuit down to four components.
The 150 kHz switching frequency is also what keeps the filter parts small. Compared with older 52 kHz regulators of the same family, the inductor and capacitors can be roughly a third the size for the same ripple target.
Typical Application Circuits
Fixed output version
The fixed 5 V circuit from the datasheet is the complete part count: input capacitor, catch diode, inductor, output capacitor.

The datasheet’s selection rules for each part matter more than the specific values, because they are what you scale when your input voltage or load current differs:
Catch diode (D1). Current rating at least 1.3× the maximum load current, reverse voltage rating at least 1.25× the maximum input voltage, and it must be fast. Schottky diodes give the best efficiency, especially at 5 V outputs and below. Slow rectifiers like the 1N4001 or 1N5400 series must not be used — their recovery time is far too long for 150 kHz switching.
Input capacitor (C<sub>IN</sub>). Selected by RMS ripple current rating, not capacitance: it needs to handle at least half the DC load current. Voltage rating around 1.5× the maximum input. A low-ESR electrolytic close to pin 1 with short leads.
Output capacitor (C<sub>OUT</sub>). Low-ESR electrolytic or solid tantalum between 82 µF and 820 µF — and not larger than 820 µF. ESR has both an upper and a lower limit here: too high and output ripple grows, too low and the feedback loop can go unstable and oscillate. Staying within the datasheet’s recommended capacitor tables keeps you inside that window.
Inductor (L1). Chosen from the datasheet’s selection nomographs for your input voltage and load current, sized for continuous-mode operation. Exceeding the inductor’s current rating saturates the core, its inductance collapses, and the regulator hits cycle-by-cycle current limit.
Adjustable version
The adjustable circuit adds two resistors. R1 sits between the feedback pin and ground, R2 between the output and the feedback pin, and the output settles where the divider delivers 1.23 V to the feedback pin:
V<sub>OUT</sub> = 1.23 V × (1 + R2/R1)

The datasheet’s procedure: pick R1 between 240 Ω and 1.5 kΩ, use 1% metal film for stability, then solve R2 = R1 × (V<sub>OUT</sub>/1.23 − 1) and take the nearest standard value. Its own worked example targets 20 V with R1 = 1 kΩ, which gives an ideal R2 of 15.26 kΩ and a chosen value of 15.4 kΩ. The calculator at the top of this page runs the same equation both directions and snaps R2 to standard E24 values, so you can skip the arithmetic.
Two additions for higher outputs. Above roughly 10 V, the datasheet calls for a feedforward capacitor across R2 (C<sub>FF</sub> = 1/(31 × 10³ × R2), typically 100 pF to 33 nF) to keep the loop stable. And keep both divider resistors physically close to the IC with the feedback trace routed away from the inductor — the feedback node is high impedance and picks up switching noise easily.
Using the LM2596 Module
The ubiquitous buck module is the LM2596-ADJ application circuit laid out on a small board, with the fixed R2 replaced by a multiturn trimmer potentiometer.

Set it up in this order: connect the input with nothing on the output, put a multimeter across OUT+ and OUT−, and turn the pot until you read the target voltage. Only then connect the load. The pot is a multiturn part — expect to turn it several full rotations before the voltage starts moving, especially on a fresh board. Turning it with the load already attached risks sweeping the output through voltages the load was never meant to see.
Keep the input a few volts above the target output. The internal switch drops about 1.16 V at 3 A, and as the input approaches the output plus that drop, the regulator runs out of headroom and falls out of regulation. The datasheet’s dropout voltage curve shows how the required margin grows with load current.
Heat and real-world current limits
The 3 A rating is a silicon rating, not a promise about any particular board. The datasheet is direct about this: in the TO-220 package the LM2596 requires a heat sink under most conditions, and in the surface-mount TO-263 package the PCB copper is the heat sink — TI specifies at least 0.4 in² of copper under the tab and recommends two or more square inches for good dissipation. A compact module offers limited copper, so how much continuous current it sustains depends on input-output differential, ambient temperature, and airflow. The chip protects itself with thermal shutdown and current limiting, so overload usually means the output cuts out rather than the part dying — but a converter cycling through thermal shutdown is not regulating. For sustained loads well above 1–2 A, check the IC temperature in your actual setup and add a heat sink or airflow before trusting the full rating.
Powering an Arduino or ESP32
Dropping a 7.4 V or 12 V battery to logic voltage is the single most common hobbyist use of these modules. The reliable pattern: set the module to 5 V and feed the board’s 5 V input (the 5V pin on an Arduino, or VIN/5V on an ESP32 dev board), letting the board’s onboard regulator produce the clean 3.3 V rail. Feeding a bare ESP32’s 3.3 V pin directly from a switching converter skips that final filtering stage and passes the switcher’s ripple straight into the chip’s supply — it can work, but the two-stage arrangement is more robust for the cost of nothing. Set and verify the output voltage with a meter before the microcontroller ever touches the module.
LM2596 FAQ
Can the LM2596 charge lithium batteries?
Not by itself. A bare LM2596 regulates voltage only. Lithium charging requires controlled current during the bulk phase and a precise voltage limit during absorption, which is why the module variants sold for charging add a constant-current circuit and adjustment pot alongside the voltage-setting one. For anything beyond bench experiments, a dedicated charger IC with proper termination is the correct tool.
What is the difference between the LM2596 and the LM2577?
Opposite directions. The LM2596 is a step-down (buck) regulator: output lower than input. The LM2577 is a step-up (boost) regulator from the same SIMPLE SWITCHER family: output higher than input. If your source voltage is below your target voltage, no amount of adjustment on an LM2596 will get you there.
Why is my output unstable or oscillating?
Three usual causes, all covered in the datasheet. First, an output capacitor with extremely low ESR (ceramic-only or some solid tantalums) can push the feedback loop into oscillation — the ESR window has a lower bound, not just an upper one. Second, on adjustable builds above 10 V, a missing feedforward capacitor across R2 leaves the loop short on phase margin. Third, open-core (bobbin or rod) inductors spray magnetic flux into nearby traces; if the feedback trace runs close to one, the induced noise destabilizes regulation. Re-route the feedback away from the inductor or switch to a toroid or E-core part.
