BCD Binary Explained: Converting Decimal to 4-Bit Logic for Displays and RTCs

Binary-Coded Decimal (BCD) is a digital encoding system where each individual decimal digit from 0 to 9 is represented by its own distinct four-bit binary sequence. In a real circuit or installation, BCD fundamentally changes how microcontrollers interface with human-readable hardware; it eliminates the need for CPU-intensive base-2 to base-10 mathematical division when driving displays or reading timekeeping registers, allowing the logic to simply shift and mask 4-bit nibbles. Instead of treating a byte as a single value from 0 to 255, BCD treats a byte as two separate decimal digits, each capped at 9.

The BCD Binary Lookup Table and Conversion Rules

The most common variant used in digital logic is 8421 BCD. The name refers to the binary weights of the four bits in a nibble (from most significant to least significant: 8, 4, 2, and 1). Because four bits can technically count up to 15 (1111), the states from 10 to 15 are considered "invalid" or "don't care" states in strict BCD logic. Feeding these invalid states into a hardware decoder will result in unpredictable or blank outputs.

As shown in the table, the BCD nibble for 5 is 0101 (4 + 1). If you were using pure binary to count to 5, the result is identical. The divergence happens the moment you cross the decimal threshold of 9.

Worked Numeric Example: Pure Binary vs. BCD Binary

To understand why BCD exists, let us look at how a microcontroller handles the decimal number 42 when writing to an 8-bit hardware register (like PORTD on an ATmega328P or an ESP32 GPIO expander).

Scenario A: Pure Binary Encoding

In standard base-2 math, 42 is calculated by summing powers of two: 32 + 8 + 2.
Pure Binary: 0010 1010 (Hex 0x2A).
If you send 0x2A directly to a BCD-to-7-segment decoder chip, the lower nibble (1010, which is 10) is an invalid BCD state. The decoder will likely blank the display or show a garbage symbol, because it expects a digit between 0 and 9.

Scenario B: BCD Binary Encoding

In BCD, we ignore the overall mathematical value and encode the characters "4" and "2" individually.
The digit 4 is 0100.
The digit 2 is 0010.
BCD Binary: 0100 0010 (Hex 0x42).

To convert a standard binary variable to BCD in C++ before sending it over I2C or SPI, you use bitwise shift and mask operations:

uint8_t decimal_val = 42;
uint8_t bcd_val = ((decimal_val / 10) << 4) | (decimal_val % 10);
// Result: bcd_val is 0x42 (0100 0010 in binary)

Where You Meet BCD Binary in Practice

While modern software handles base-10 string formatting effortlessly, BCD remains deeply embedded in hardware interfaces where silicon cost, pin count, and legacy compatibility matter.

1. Real-Time Clocks (RTCs) like the DS1307 and DS3231

If you read the time from a DS1307 RTC via I2C, the chip returns time data in BCD. The seconds register (Address 0x00) does not return a value from 0-59 in pure binary. If the time is 45 seconds, the I2C bus returns 0x45. If your Arduino code treats 0x45 as a standard integer, it will display "69 seconds" on your LCD, because 0x45 in pure hex is 69 in decimal. You must unpack the BCD nibbles to get the correct time.

2. 7-Segment Display Decoders

Hardware decoder ICs like the Texas Instruments 74LS47 (for common-anode displays) or the CD4511 (for common-cathode displays) take 4 BCD input pins (A, B, C, D) and translate them into the 7 individual LED segment pins (a-g). This saves 3 microcontroller GPIO pins per digit. By daisy-chaining the Ripple Blanking Input (RBI) and Ripple Blanking Output (RBO) pins, you can suppress leading zeros on multi-digit displays without writing a single line of software.

3. Industrial PLC Thumbwheel Switches

On older or heavy-duty industrial control panels, operators set parameters (like motor RPM limits or timer delays) using physical BCD thumbwheel switches. These switches output 4-bit BCD directly to the PLC's digital input cards, ensuring that a physical dial set to "8" physically cannot output a binary "1010" (10) due to internal mechanical diode steering.

Common Confusions and Troubleshooting BCD Circuits

When debugging digital logic or writing firmware, BCD is frequently confused with other encoding schemes. Here is how to separate them and fix common hardware lockups.

BCD vs. ASCII Encoding

ASCII is a 7-bit or 8-bit character encoding standard used for serial text transmission. The ASCII character for the digit "5" is 0x35 (Binary 0011 0101). The BCD encoding for the digit "5" is simply 0x05 (Binary 0000 0101). A common beginner mistake when parsing serial data from a GPS module or a digital scale is attempting to feed raw ASCII bytes directly into a BCD hardware decoder, resulting in completely incorrect numerical outputs.

BCD vs. Gray Code

Gray code is an encoding where only one bit changes state between any two consecutive numbers, used primarily to prevent errors in rotary encoders and mechanical position sensors. BCD does not share this property; transitioning from BCD 3 (0011) to BCD 4 (0100) requires three bits to flip simultaneously.

Hardware Lockup: Invalid States on the 74LS47

If your 7-segment display is showing fragmented, dim, or asymmetrical patterns (like just the 'g' and 'e' segments lit), check your microcontroller's output with a logic analyzer or multimeter. You are likely feeding an invalid BCD state (10 through 15) into the decoder. This usually happens if your software counter increments past 9 without resetting the lower nibble and carrying over to the upper nibble. Ensure your software implements a modulo-10 rollover for the ones digit before incrementing the tens digit.