RS-485 is an electrical interface standard widely used for serial data transmission between industrial devices. It uses balanced differential signalling and is commonly found on PLCs, HMIs, variable frequency drives, meters, controllers, sensors and distributed I/O systems. RS-485 does not define device addresses, registers, commands or error-checking rules. Those functions belong to higher-level communication protocols such as Modbus RTU.

In simple terms: RS-485 provides the electrical path that allows industrial devices to exchange data over a cable. It defines how the electrical signals travel on the line, while a protocol such as Modbus determines what the messages mean, which device is being addressed and what information is being requested or returned.

How does RS-485 work?

RS-485 does not determine a logic state by measuring the voltage on a single conductor against ground. Instead, the receiver evaluates the voltage difference between two conductors. This is known as differential signalling. When both conductors pick up similar electrical noise, much of that common noise can be rejected because the receiver is interested mainly in the difference between the two signals.

This is one reason RS-485 remains useful in industrial environments containing contactors, motors and variable frequency drives, where electrical interference can be considerably higher than in office or residential installations.

RS-485 defines the electrical signalling layer rather than the application protocol. A device marked “RS-485” therefore does not automatically mean that it supports Modbus RTU. The same RS-485 interface may carry Modbus RTU, a manufacturer-specific protocol or another serial communication system. Both devices must support the same physical interface and the same higher-level protocol before they can communicate.

Two-wire and four-wire RS-485

The most common industrial arrangement is two-wire RS-485. The same differential pair is used for transmission and reception, so this arrangement normally operates in half-duplex mode. One device transmits while the other devices listen, and two stations should not drive the same pair at the same time.

RS-485 can also be implemented with separate transmit and receive pairs. A four-wire full-duplex arrangement allows transmission and reception to take place simultaneously when the hardware and communication scheme are designed for it. For this reason, describing RS-485 itself as “always half-duplex” is inaccurate.

How should the A and B terminals be connected?

RS-485 terminals are often labelled A and B, but manufacturers may also use D+, D−, 485+, 485−, D0 and D1. Unfortunately, A/B naming has not always been applied consistently across manufacturers. It is therefore unsafe to assume that “A is always positive” and “B is always negative.”

The terminal definitions in the manuals of both devices should be compared before wiring the connection. If communication parameters are correct but no traffic can be established, reversed differential conductors are one of the first physical wiring issues worth checking.

What is the correct RS-485 network topology?

A conventional RS-485 network is normally installed as a bus or daisy-chain. The main cable continues from one station to the next, while stubs should be kept as short as practical. Long star branches and uncontrolled junctions can produce reflections and signal-integrity problems, particularly as communication speed and cable length increase.

When several PLCs, HMI panels, drives or field instruments share the same serial network, the physical order of the devices, cable routing and branch lengths should be considered during system design rather than after communication problems appear.

Where a star arrangement cannot be avoided, a properly designed RS-485 repeater, hub or isolated interface may be more appropriate than simply splitting the passive cable into several long branches.

Why are 120-ohm termination resistors used?

At sufficient cable length and signalling speed, an RS-485 cable behaves as a transmission line. If the impedance at the end of the cable does not reasonably match the cable's characteristic impedance, part of the signal can be reflected back along the line. These reflections may distort signal transitions and lead to intermittent communication errors.

Industrial twisted-pair cables used for RS-485 are commonly specified around 120 Ω characteristic impedance. For this reason, termination resistors of approximately 120 Ω are commonly installed at the two physical ends of the bus. A termination resistor should not automatically be fitted to every device on the network, because doing so can excessively load the line.

Termination is not something to add blindly. A short, low-speed network may operate correctly without termination, and many industrial devices already include switchable or built-in termination. The equipment documentation and cable characteristics should always be checked.

What are bias and fail-safe resistors?

On a half-duplex RS-485 network there can be periods when none of the transmitters is actively driving the bus. Without a defined idle state, electrical noise may cause a receiver to interpret false transitions. Some systems therefore use bias, polarization or fail-safe resistors to establish a known differential voltage while the bus is idle.

Many modern RS-485 transceivers also provide internal fail-safe receiver behaviour. Whether external biasing is required depends on the equipment being used. When an external bias network is necessary, it should be designed for the complete bus rather than duplicated at every station without calculation, since excessive biasing can load the network.

How many devices can be connected to RS-485?

Older documentation often states that an RS-485 network supports a maximum of 32 devices. The traditional specification is more accurately related to 32 unit loads, not an absolute 32-device limit for every modern installation.

Modern receivers may present only 1/2, 1/4 or 1/8 of a traditional unit load. Appropriate transceivers can therefore allow substantially more than 32 physical nodes on one segment. The actual limit depends on transceiver loading, termination, biasing, repeaters, cable conditions and the manufacturer's network specifications.

The physical RS-485 node capacity must also be distinguished from the addressing capacity of the protocol. A physical network may electrically support many stations while the higher-level protocol imposes a different address limit.

How far can RS-485 communicate?

There is no single cable length that can be guaranteed for every RS-485 installation. Maximum practical distance depends on baud rate, cable capacitance, conductor characteristics, impedance, transceiver performance, topology, number of loads and the electromagnetic environment.

As a general engineering rule, longer distances become easier to achieve as the data rate is reduced. Higher signalling rates normally require shorter cables and tighter control of termination, topology and stub length.

Values such as 1000 or 1200 metres are often quoted in industrial literature, but they should be treated as application references rather than universal guarantees. The distance and baud-rate limits specified by the actual equipment manufacturer take priority.

Which cable should be used for RS-485?

RS-485 should normally be wired using a suitable twisted-pair communication cable with electrical characteristics appropriate for the network. Shielded cable can be useful in electrically noisy installations. Communication wiring should not be routed for long distances directly alongside motor leads, contactor wiring or PWM output cables from variable frequency drives.

Differential signalling provides strong noise immunity, but it does not make the interface immune to unlimited ground-potential differences. If the common-mode limits of the transceiver are exceeded, communication can fail and hardware may be damaged. Galvanically isolated RS-485 interfaces can be particularly valuable between separate panels, distant sections of a plant or equipment with significant potential differences.

The cable shield, protective earth and signal reference are not interchangeable conductors. Their connection should follow the EMC and wiring recommendations of the equipment manufacturers. Using a shield as an improvised signal-reference conductor is not a sound installation practice.

Are RS-485 and Modbus the same thing?

No. This is one of the most common misunderstandings in industrial serial communication. RS-485 defines the electrical signalling interface. Modbus defines how devices structure and interpret messages.

A Modbus RTU message may contain a device address, function code, data field and CRC. None of those concepts is defined by RS-485 itself. Expressions such as “RS-485 address” usually refer to the address assigned by the protocol running over the RS-485 physical network.

For applications using Modbus over an RS-485 physical interface, see the Modbus RS-485 communication guide. This distinction is especially important during troubleshooting: a perfectly wired RS-485 bus can still fail if the connected devices do not use compatible protocols or communication settings.

RS-232, RS-422 and RS-485: what is the basic difference?

RS-232 is primarily intended for point-to-point serial communication and uses single-ended signalling. RS-422 uses differential signalling and is commonly associated with one driver communicating with multiple receivers. RS-485 extends the concept to multipoint bus systems in which multiple drivers can share the same network when they are controlled correctly.

These electrical standards should not be confused with application protocols. RS-232, RS-422 or RS-485 describe how electrical serial signals are transmitted, while protocols define the meaning and organisation of the data being carried.

Why does an RS-485 network fail to communicate?

A large proportion of RS-485 problems should be investigated at the physical layer before changing PLC or HMI software. Common causes include reversed differential conductors, excessive star branching, incorrect termination, termination fitted to every device, unsuitable cable, poor grounding practices, electromagnetic interference and excessive ground-potential differences between connected equipment.

Once the physical bus has been verified, the settings required by the higher-level protocol should be checked. In a Modbus RTU installation this typically includes baud rate, parity, data and stop-bit format, and device addresses. Two devices using the same protocol address on the same segment can create reply conflicts.

If an SCADA system or HMI cannot read any station, a sensible troubleshooting sequence is to verify the physical line first, then the serial-port parameters, and finally the protocol messages and register mapping.

A multimeter alone may not reveal intermittent signal-integrity problems. For difficult faults, an oscilloscope or suitable serial communication analyser can help identify reflections, noise, timing problems and malformed protocol traffic.

Where is RS-485 used?

RS-485 is widely used in machine automation, process control, building automation, energy metering, motor drives, temperature controllers, weighing systems, remote I/O, PLC networks and operator interfaces. It remains a practical option where a simple and economical field connection is required and an Ethernet network is unnecessary.

An RS-485 port alone does not guarantee interoperability. Before connecting an HMI to a PLC, drive or controller, check the supported protocol, baud rate, framing parameters and the required address or register structure as well as the physical interface.

Practical RS-485 field checklist

1. Do both devices actually have RS-485 interfaces and support the same higher-level protocol?

2. Have the A/B, D+/D− or D0/D1 terminal definitions been checked in both manuals?

3. Is the cable installed as a bus or daisy-chain without unnecessarily long branches?

4. Is termination installed only where required, and have built-in termination switches been checked?

5. Are baud rate, parity and serial framing settings identical at both ends?

6. Are protocol addresses unique where the protocol requires device addressing?

7. Is the communication cable separated from motor and VFD power wiring?

8. Has galvanic isolation been considered for long links or equipment with different ground potentials?

Related Industrial Communication Guides

Industrial Communication · Modbus FAQ · Modbus RS-485 Communication

Related Automation Categories

PLCs and Modules · HMI Panels · SCADA · AC Drives and VFDs