Find its address

Every display in the box answers to 0x27, which is fine until you own two. Three solder pads change that, and the address they make is arithmetic rather than folklore.

Updated 2026-09-135 min read1602 LCD

0x27, and where it comes from

The expander's address is not a setting. It is the three address pins, read straight into the bottom of the byte: the chip answers to 0 1 0 0 A2 A1 A0, which is 0x20 plus whatever those three pins are. On this board all three are held high and the pads that could pull them low are open, so the address is 0x27.

That is why every sketch for these displays starts at 0x27, and why all three displays in the box answer there.

Three pads, and the address they make

0x27

Seven-bit address

0x27

Pads bridged

Displays on one bus

As it ships: 0x27. All three pads are open, all three address pins are held high, and 0100 111 is 0x27. Three displays out of the box all answer here, which is why the second one you plug in makes the first one misbehave. The 0x3F you will see offered as the other common address belongs to the PCF8574A, whose base is 0x38. This board carries the plain PCF8574T, so 0x20 to 0x27 is the whole of what it can be.

Three pads, eight addresses

Each pad is a pair of exposed copper rings. Bridge one with a blob of solder and that address bit goes low, so the address counts down from 0x27: bridge A0 and it is 0x26, bridge A1 and it is 0x25, bridge both and it is 0x24.

Three bezelled 1602 displays on a bench, each showing LONELY BINARY on the top row and I2C ADDR 0x25, 0x26 and 0x27 on the bottom row, all wired through one small logic level shifter on a breadboard to a single ESP32 board. — Three displays, three addresses, one pair of wires. The small black board on the breadboard is the level converter the next chapter is about.

Eight is the ceiling. If you want more than eight character displays on one bus you need a second bus or a multiplexer, and at that point a different kind of display is usually the better answer.

Run the scanner first

The scanner is the cheapest diagnostic on this bench. It asks every address in turn and prints the ones that answer, and an answer proves a lot: the display has power, the two signal wires are the right way round, and the display can pull a line low back through whatever is between you and it.

Anything in 0x20 to 0x27 is a PCF8574, and on this bench that means the display. Anything outside it is something else you have plugged in.

The code

scan_i2c.ino

A bus scanner. It prints every address that answers, once every three seconds, and flags anything in the range this display can be. Run it before you suspect your code of anything.

// Wiring (Arduino Uno):
//
//   LCD GND -> GND
//   LCD VCC -> 5V
//   LCD SDA -> A4
//   LCD SCK -> A5
//
// Arduino IDE: Tools > Board "Arduino Uno", Serial Monitor at 9600.
// No library needed — Wire is built in.

#include <Wire.h>

void setup() {
  Serial.begin(9600);
  Wire.begin();
  Wire.setClock(100000);   // the expander is specified to 100 kHz
  Serial.println("scanning");
}

void loop() {
  int found = 0;
  for (uint8_t addr = 1; addr < 127; addr++) {
    Wire.beginTransmission(addr);
    if (Wire.endTransmission() == 0) {
      Serial.print("  0x");
      Serial.print(addr, HEX);
      if (addr >= 0x20 && addr <= 0x27) Serial.print("  <- a PCF8574, so probably the display");
      Serial.println();
      found++;
    }
  }
  if (found == 0) Serial.println("  nothing answered - check VCC, GND, SDA and SCK");
  Serial.println();
  delay(3000);
}

One line and no more means one display. Nothing at all means the wiring, not the address — the scan talks to a chip that is powered and connected, and it cannot fail politely. Two displays both printing 0x27 is two boards fighting over one address, and the pads are the fix.

When it does not work

The scan finds nothing at all

That is a wiring fault, not an address problem — the scan asks all 126 addresses, so it cannot miss one. Check 5 V on the display's own VCC pin, a shared ground, and that SDA and SCK are not swapped. On a 3.3 V board, check the level converter is powered on both sides.

It finds 0x27 but the screen stays blank

Then the bus is fine and the contrast is not. Turn the trimmer through its whole travel; “Power, and the contrast trimmer” has the detail. The scan succeeding proves the display can pull a line low, which is most of the wiring.

Two displays and only one address appears

Both are answering at 0x27 and the scan cannot tell them apart. Bridge one address pad on the second display, rescan, and give each one its own address in your sketch. Until you do, writing to either one writes to both.

Tutorials say to try 0x3F and it does not work

0x3F belongs to the PCF8574A, a different part whose base address is 0x38. This board carries the plain PCF8574T, so it can only be 0x20 to 0x27. If nothing in that range answers, the problem is not the address.

The address changed after I soldered a pad

That is what the pad does. Bridging pulls that bit low, so the address counts down from 0x27. Re-run the scan and put the new number in the LiquidCrystal_I2C constructor.

Alongside this page

Where an I²C address comes from, and what a clash looks like What the scanner is actually putting on the wire

Where this goes next

The one chapter to read before this display meets an ESP32.

Why it wants five volts →

Edit this page — content/books/lcd1602/find-its-address.mdx

Community

Questions about this product

See what other owners have asked, and read their solutions.

Ask a question ↗

3-Pack 1602 LCD Display Module, I2C 16x2 Blue

Loading discussions…

View all on the community ↗Read here. Sign in to the community to post or reply.

Discuss this article

Ask about this page. The answer stays here, on the page it belongs to, for whoever hits the same wall next.

Browse Modules and blocks on the forum →

← 03Power, and the contrast trimmer 05 →Why it wants five volts