RS-485 Circuit

The RS-485 subsystem on the RS485 CAN HAT uses a single IC — the SP3485 half-duplex transceiver — paired with a transistor-based automatic direction control circuit. This combination allows the Raspberry Pi to communicate over RS-485 using its UART pins without any software-managed direction switching.

SP3485 — RS-485 Transceiver

The SP3485 is a low-power, half-duplex RS-485 transceiver operating at 3.3V. “Half-duplex” means the chip can either transmit or receive at any given moment, but not both simultaneously. The A and B differential lines serve double duty — carrying data in both directions, with direction controlled by the enable pins.

Pin Connections

SP3485 Pin	Connected To	Direction	Function
RO (Receiver Output)	Pi RXD	Output	Received data to Pi
DI (Driver Input)	Pi TXD	Input	Transmit data from Pi
RE̅ (Receiver Enable)	Control circuit	Input	Active LOW — enables receiver
DE (Driver Enable)	Control circuit	Input	Active HIGH — enables driver
A	Terminal A	Bus	Non-inverting differential line
B	Terminal B	Bus	Inverting differential line

RE̅ and DE are complementary in normal operation: when the driver is enabled (DE HIGH), the receiver is disabled (RE̅ HIGH), and vice versa. On this board, both pins are driven by the same control signal, so the chip is always in either transmit or receive mode — never both, never neither.

Differential Signaling

RS-485 encodes data as the voltage difference between the A and B lines:

Condition	Interpretation
V(A) - V(B) > +200 mV	Logic 1 (mark)
V(A) - V(B) < -200 mV	Logic 0 (space)

This differential scheme gives RS-485 its noise immunity. Common-mode noise — interference that shifts both A and B by the same amount — is rejected because only the difference matters. This is why RS-485 can reliably span hundreds of meters where single-ended UART signals would fail.

Bus Termination

A 100-ohm termination resistor is present on the board between the A and B lines. This resistor absorbs signal reflections at the end of the cable and prevents ringing that could cause bit errors.

Tip

The 100-ohm termination is appropriate when this board sits at one end of the RS-485 bus. If the board is in the middle of a daisy-chain (not at a physical end), the termination resistor may cause signal attenuation. In multi-drop configurations, only the two endpoints of the bus should have termination resistors.

Line Protection

A TVS (Transient Voltage Suppressor) diode on the A and B lines clamps voltage spikes before they reach the SP3485. This protects against surges from lightning strikes, ESD events, and inductive load switching — all common in industrial environments where RS-485 is deployed. The TVS response time is in nanoseconds, fast enough to suppress transients that would otherwise destroy the transceiver.

Automatic TX/RX Control

By default, the board ships with automatic direction control enabled. A small transistor circuit monitors the Pi’s UART transmit line (P_TX) and switches the SP3485 between receive and transmit modes without any GPIO involvement.

How It Works

The circuit operates on a simple principle: the UART TX line idles HIGH when no data is being sent.

Receiving data (idle state):

When P_TX is idle (HIGH), the transistor conducts. This pulls the control line to a state where RE̅ goes LOW, enabling the SP3485 receiver. Data arriving on the A/B differential pair is converted by the SP3485 and output on the RO pin, which feeds the Pi’s RXD. The Pi reads incoming bytes through its UART as normal.

Transmitting data:

When the Pi begins transmitting, P_TX drops LOW for the start bit. This causes the transistor to cut off, which drives DE HIGH and RE̅ HIGH — enabling the transmitter and disabling the receiver. The SP3485 now drives the A and B lines based on the data on its DI pin (connected to P_TX).

During the body of the transmitted byte, when a logic ‘1’ bit is sent, P_TX goes HIGH momentarily. The transistor briefly conducts, which would normally re-enable the receiver. However, during this brief window the SP3485 enters a high-impedance state on the bus side, and the bit duration is short enough that no corruption occurs. The chip returns to active transmit mode as soon as the next LOW bit (or stop/start bit) arrives.

Caution

The transistor switching circuit introduces a finite switching delay. At very high baud rates, this delay becomes significant relative to the bit period and can cause the first bit of received data to be corrupted, or the last bit of transmitted data to be clipped. If you need baud rates above 115200, consider switching to manual TX/RX control.

Timing Implications

After the last byte is transmitted, P_TX returns to idle HIGH. The transistor conducts again, switching the SP3485 back to receive mode. There is a brief dead time during this transition where the chip is neither fully transmitting nor fully receiving. For most RS-485 applications at standard baud rates (9600, 19200, 38400, 57600, 115200), this dead time is negligible and causes no issues.

Manual TX/RX Control

For applications that require precise direction control or very high baud rates, the board can be reconfigured for manual (software-controlled) direction switching.

Hardware Modification

Switching to manual control requires physically moving a 0-ohm resistor on the board. The board has two resistor positions — one for automatic mode (factory default) and one for manual mode. Moving the resistor disconnects the transistor circuit and connects the SP3485 enable pins to the RSE control line instead.

Caution

This modification requires soldering skills and a steady hand. The 0-ohm resistors are small surface-mount components (typically 0402 or 0603 size). Use a fine-tipped soldering iron and work under magnification. Refer to the Waveshare wiki for the exact resistor positions on your board revision.

Software Direction Control

With manual mode enabled, the RSE pin — routed to GPIO 4 on the Raspberry Pi — controls the SP3485 direction:

GPIO 4 State	SP3485 Mode	DE	RE̅
HIGH	Transmit	HIGH	HIGH
LOW	Receive	LOW	LOW

The software must manage the timing explicitly:

1. Set GPIO 4 HIGH before writing data to the UART.
2. Transmit the data.
3. Wait for the UART transmit buffer to fully drain (all bytes shifted out on the wire).
4. Set GPIO 4 LOW to return to receive mode.

Note

Step 3 is critical and is a common source of bugs. The Pi’s UART reports “transmit complete” when the last byte enters the hardware shift register, not when it finishes appearing on the wire. You must wait for the shift register to empty before switching direction. The tcdrain() system call (or its Python equivalent) handles this correctly. Switching too early will truncate the last byte on the bus.

When to Use Manual Control

Manual control is preferable in these situations:

• High baud rates (above 115200) where the automatic switching circuit’s transistor delay causes bit errors.
• Multi-drop networks where precise turnaround timing is required by the protocol (e.g., Modbus RTU with strict inter-frame timing).
• Low-power applications where you want to keep the transceiver in a known state (receiver disabled) to save power during idle periods.
• Deterministic timing requirements where the automatic circuit’s variable switching behavior is unacceptable.