2-Wire RS-485 Wiring Diagram: PLC, Gateway and Field Devices

Direct answer

A basic two-wire RS-485 wiring diagram has one shared differential pair running as a continuous trunk through every device. At each node, one documented data conductor lands on the terminal the manufacturer defines as one side of the differential signal, and the other conductor lands on its complement. The phrase “two-wire” counts the signal pair; the approved installation may also require a signal-reference/common conductor, cable shield, protective bonding, isolation or surge protection. Termination, if required by the installed design, belongs at the two physical ends of each electrical segment. Bias or line polarisation belongs at one coordinated location unless the exact transceivers already provide the required idle-state behaviour.

Do not draw or wire the circuit by assuming that A, B, +, −, D0 and D1 mean the same thing on every product. Build a terminal-translation table from both manuals first. A selectable four-wire port can operate with a two-wire device only when its documentation explicitly permits two-wire half-duplex mode and defines how its same-polarity transmit and receive terminals are joined. Connector pin numbers are product-specific.

The safest reusable diagram is therefore a controlled template, not a universal pinout:

What this page owns—and what the broader guide owns

This page is the implementation owner for people who need to turn two or more product manuals into a connection drawing. It covers native two-wire terminals, selectable four-wire ports, terminal translation, drawing fields, diagram review and point-by-point verification. The broader RS-485 wiring, termination and troubleshooting guide owns the deeper design and fault-diagnosis decisions for termination, biasing, grounding, shielding, cable length, topology and waveform interpretation.

Keep the query variants on one diagram owner

“RS-485 2 wire,” “two-wire RS-485 connection,” “RS-485 half-duplex wiring” and “2 wire RS-485 wiring diagram” normally ask for the same installation artifact. Publishing separate keyword-swapped articles would split links and force search engines and readers to choose between nearly identical answers. This canonical answers the variants as sections, tables and FAQs while the broader physical-layer guide remains distinct by task.

RS-485 is not the application protocol

RS-485 defines electrical signalling, not device addresses, register meanings or message framing. Modbus RTU, BACnet MS/TP and proprietary serial protocols may use a two-wire RS-485 physical layer, but each adds its own role, timing and configuration rules. A correct cable can still show no application response when baud, parity, address, function or register mapping is wrong. Keep physical and protocol acceptance evidence separate.

Read the two-wire diagram correctly

Two-wire half-duplex uses one differential signal path in both directions. The controller drives a request, releases the line, and the addressed device drives a response. Other devices keep their drivers disabled. Receivers can listen while the pair is idle or active, but only the intended transmitter should control the segment at a given time.

“Two-wire” can still mean more than two conductors

The name describes the differential data pair. A cable or terminal block may also include COM, SG, SC, REF, drain/shield and protective-earth provisions. A reference conductor is not automatically the cable shield. The shield is not automatically protective earth. A drawing that shows only two coloured strokes may be adequate for a conceptual block diagram but inadequate for fabrication or commissioning.

Direction control is part of the connection contract

Many PLC modules and gateways control their driver automatically after two-wire mode is selected. Other interfaces expose a transmit-enable function or rely on adapter timing. The driver must release early enough for the response and not so early that it truncates the request. The Analog Devices AN-960 implementation guide depicts two-wire RS-485 as a multipoint half-duplex configuration with multiple drivers and receivers on the same signal path. That electrical model is why transmit ownership matters.

Build a terminal-translation table before the schematic

Terminal letters are labels, not proof of electrical polarity. Some manuals define A as inverting or negative; others associate A with positive or non-inverting. Protocol documents can introduce their own D0 and D1 convention. Treat each endpoint as a translation exercise.

Translate definitions, not typography

For every device, record the catalogue number, hardware option, manual document number/revision, connector, pin, printed terminal name, manual’s electrical description and selected port mode. Then assign neutral project conductor names such as PAIR_1 and PAIR_2. Connect only when two device definitions map to the same conductor meaning.

The current Schneider PM8000 help, for example, describes − as data minus/inverting, + as data plus/non-inverting and C as common for that product family. That is useful product evidence, not permission to reinterpret another vendor’s A and B without its manual.

Do not trust cable colour or connector shape

RS-485 does not assign a universal conductor colour. Nor does a DB-9, RJ-style or removable terminal connector imply a universal RS-485 pinout. The same shell can carry RS-232, RS-422, two-wire RS-485, four-wire RS-485, power or manufacturer-specific signals. Draw the connector from the equipment-facing or cable-facing perspective explicitly; confusing those views mirrors the pins.

Draw a native two-wire multidrop connection

Start with one horizontal trunk, then place devices in physical cable order. Show the pair entering and leaving intermediate terminals or junctions without creating long branches. Mark the two physical segment ends. Add the documented reference and shield paths separately. Only after the topology is clear should the drawing show termination and bias components.

A practical drawing sequence

1. Identify every device, connector and port mode from the approved manuals.
2. Choose neutral project names for the two signal conductors.
3. Draw one continuous trunk in real physical order, including repeaters as boundaries.
4. Add each node’s very short approved connection to the trunk.
5. Add common/reference, shield and protective bonding as distinct paths.
6. Mark actual electrical ends and every internal or external termination network.
7. Mark the single coordinated bias/polarisation source or documented internal fail-safe strategy.
8. Add cable type, route/segment length, stub length, connector view and drawing revision.
9. Add a note that each terminal definition is controlled by the cited device manual.
10. Review the drawing against the installed configuration before any energised test.

Termination is a design decision, not a decorative symbol

Parallel termination commonly approximates the cable’s differential characteristic impedance and is placed at the physical ends when required. Two nominal 120-ohm end resistors are common in industrial RS-485, but the installed cable, edge rate, device instructions and integrated networks control the real value and need. The Texas Instruments termination guidance explains the relationship between characteristic impedance, reflections and termination. Do not add a resistor at every node or assume the logical controller is always a physical end.

Bias must be coordinated with termination and receiver behaviour

External pull-up/pull-down networks establish an idle differential state in designs that need them. Several devices with enabled bias networks can overload or shift the bus. Modern receivers may provide open, short or idle fail-safe behaviour. The schematic should identify the approved strategy, not simply repeat generic resistor values copied from the internet.

Connect a selectable four-wire port to a two-wire device

A four-terminal serial interface often exposes separate transmit and receive pairs. Some products permit those same-polarity paths to be joined so the port can operate as a two-wire half-duplex interface. Others require a switch or software selection; some RS-422-only ports cannot release their driver and are unsuitable for a multidrop two-wire bus.

Use the manufacturer’s two-wire mode instructions

NI’s current support procedure for its applicable RS-485 hardware instructs users to join TXD+ with RXD+, join TXD− with RXD−, connect the resulting pair to the two-wire device, select the two-wire automatic mode and verify with a supported command. NI also publishes different connector pin pairs for DB-9, DB-25 and RJ-50 hardware, which demonstrates why a generic drawing must not supply universal pin numbers.

The reusable logic is “same documented polarity together,” not “pin X always joins pin Y.” Confirm that the product is an RS-485 interface capable of tri-stating its driver, not merely a four-wire RS-422 transmitter.

Add reference, shield and bonding without conflating them

Differential signalling rejects noise common to the pair, but receivers still have common-mode limits. A signal-reference conductor can keep endpoint voltages within that range when required. A shield intercepts or redirects electromagnetic coupling according to the EMC design. Protective earth and equipotential bonding serve safety and installation functions. They are related design elements, not synonyms.

Follow the most specific approved system guidance

The current Schneider PM8000 wiring guidance calls for a data +, data −, a C common and a shield connection for that family. The Modbus Serial Line implementation guide presents a common conductor and shield/polarisation guidance for its ecosystem. Other isolated products can differ. Use the product and system documents that actually govern the installation.

Verify the drawing before energisation

A good schematic is inspectable. A second qualified reviewer should be able to trace every conductor from the controller through each device to the far end, identify the source for every terminal definition and state which components depend on device power. Resolve contradictions while the design is still paper, not while swapping conductors on live equipment.

Perform a controlled document and continuity review

Under the site’s approved isolation and test procedures, compare the approved drawing, cable schedule, installed terminal markings and manufacturer manuals. Verify connector orientation. Confirm pair continuity and absence of unintended shorts using methods permitted by the equipment. Do not apply an ohmmeter to an energised network. Do not megger through connected electronics unless the equipment and test plan explicitly allow it.

Commission the connection with layered evidence

Start with the smallest approved segment: controller or test interface, one target, correct two-wire mode and documented end networks. Prove the request and response before adding nodes. Then add devices or segments in controlled increments. A change that causes errors is evidence; do not erase it by changing several configuration and wiring variables together.

Capture four boundaries for a no-response fault

Observe whether the request exists at the controller, reaches the target, produces a response at the target and returns to the controller. This four-boundary method separates a missing request, damaged outbound path, target rejection, target driver problem and damaged return path. Use suitably rated, isolated instruments and approved connection practices for the environment.

Close with an as-built acceptance record

Record the final node order, connector/pin mapping, cable and segment lengths, stub lengths, termination and bias locations, port modes, serial parameters, shield/reference design, device/firmware identities and baseline captures. Remove temporary adapters and test jumpers. Restore covers, protections and normal control ownership. The drawing is not finished until it matches the installed and accepted state.

Common diagram mistakes and corrective decisions

The most dangerous drawings look plausible while hiding assumptions. A straight line labelled A/B can conceal reversed vendor conventions. A shield symbol can conceal an unintended current path. A termination symbol beside the controller can conceal the fact that the controller sits halfway along the trunk.

Diagnose the drawing before diagnosing the bus

When a network fails, annotate the as-built diagram with measured evidence rather than redrawing it from memory. Mark the first boundary where the request disappears or the last boundary where the response exists. Record which node addition changes the failure. This makes the next test falsifiable.

Diagnostic answer map for search and AI-assisted troubleshooting

These expanded questions keep related two-wire connection intent on one evidence-rich canonical surface.

Frequently asked questions

What is the simplest two-wire RS-485 wiring diagram?

It is one continuous twisted differential pair from one physical end of the segment through each node to the other physical end. Each device terminal is mapped from its own manual to the two project conductors. Add the documented reference/common and shield paths separately, and show termination and bias only where the approved design requires them.

Can I connect A to A and B to B?

Only after both manuals prove those labels use compatible conventions. A/B labelling is inconsistent across manufacturers. Map each terminal to its stated electrical meaning—such as plus/minus, inverting/non-inverting or D0/D1—then connect equivalent project conductors. Do not use wire colour as proof.

Is two-wire RS-485 actually a two-conductor cable?

Not necessarily. “Two-wire” counts the differential signal pair. The installation may require a third signal-reference/common conductor, a shield/drain and protective bonding. Some products are isolated and specify a different reference strategy. Use the governing product and system drawings.

Where should 120-ohm resistors appear on the drawing?

Common RS-485 designs use parallel termination near the two physical ends of a segment when required, often near the cable’s nominal differential impedance. That is not a universal instruction for every length, edge rate or product. Show all internal and external networks and follow the exact device/cable design guidance.

Does every device need a termination resistor?

No. A resistor at every node can overload the driver and reduce differential amplitude. Intermediate devices normally do not become transmission-line ends simply because they have a terminal switch. A repeater creates separate segments; each side must be reviewed as its own electrical network.

How do I connect TX+/TX− and RX+/RX− to a two-wire device?

If the manufacturer explicitly supports two-wire mode, its procedure commonly joins the same documented polarity transmit and receive terminals, then connects the resulting pair to the device and selects half-duplex automatic direction. Use the exact connector pins and switch/software setting from that product’s manual; there is no universal pinout.

Why does the RS-485 adapter transmit but never receive?

The adapter may be in four-wire mode, may hold its driver enabled, may release it at the wrong time, or may have incorrect same-polarity jumpering. First prove a complete request at the target. Then determine whether the target creates a response and whether that response returns to the adapter.

Should RS-485 cable be daisy-chained or wired as a star?

A continuous trunk with controlled short device connections is the normal multipoint topology. A passive star creates multiple propagation paths and reflections. Product-specific low-speed or active-star systems can differ, but they require explicit design authority rather than a generic star drawing.

How should the shield and common be shown?

Show them as separate conductors or boundaries with their intended connection points. The signal common controls circuit reference where specified; the shield serves the EMC design; protective earth serves safety/bonding. Do not silently combine them or bypass an isolation boundary.

What should I test first when the completed link has no response?

Verify the exact port and two-wire mode, then observe whether a valid request exists at the controller and reaches the target. If the target receives it, check serial format, address and protocol validity, then look for a response at the target and follow it back. This localises the failure before changing several variables.

Sources, review scope, and limitations

Direct primary and official sources reviewed

1. Analog Devices AN-960 — RS-485/RS-422 Circuit Implementation Guide, accessed 2026-08-29. Used for two-wire half-duplex topology, driver/receiver sharing, termination, stubs, data-rate/length and isolation context.
2. Analog Devices — Guidelines for Proper Wiring of an RS-485 Network, accessed 2026-08-29. Used for balanced-pair, cable and physical wiring context.
3. National Instruments — Set Up 2-Wire (Half-Duplex) Communication with RS-485 Port, updated 2023-08-03, accessed 2026-08-29. Used for product-specific same-polarity TX/RX jumpering, connector-specific pins and automatic two-wire mode.
4. Schneider Electric PM8000 — RS-485 wiring, updated 2026-04-27, accessed 2026-08-29. Used as a current device-family example for plus, minus, common, shield, cable and bus distance guidance.
5. Schneider Electric — Wiring of RS485 Communications Networks, FA221785, accessed 2026-08-29. Used as manufacturer guidance on two-/four-wire selection, twisted pair and product-family electrical limits.
6. Modbus Organization — Modbus Serial Line Protocol and Implementation Guide V1.02, accessed 2026-08-29. Used for Modbus-over-RS-485 cable, D0/D1/common, trunk, line polarisation and serial implementation context.
7. Texas Instruments — When Termination Is Necessary and How to Do It Properly, revised 2026-03, accessed 2026-08-29. Used for transmission-line and characteristic-impedance context.
8. Texas Instruments — Two Ways to Fail-safe Bias Your Network, accessed 2026-08-29. Used for integrated and external receiver fail-safe context.
9. Digi International — RS485 DB9 on Connect Products, accessed 2026-08-29. Used as a product-specific example showing why two-/four-wire connector pinouts are not universal.
10. Advantech — RS-485 Connections FAQ, accessed 2026-08-29. Used for manufacturer connection examples and the requirement to determine two- versus four-wire mode from device specifications.
11. Phoenix Contact EMpro — Modbus/RTU, revision 2025-07-15, accessed 2026-08-29. Used as an installed-device example for pair, shield, termination, bias and serial-parameter requirements.
12. OSHA 29 CFR 1910.147 — The control of hazardous energy, accessed 2026-08-29. Used for the United States hazardous-energy-control boundary during covered servicing and maintenance.

Scope and limitations

This guide is vendor-neutral educational material and a drawing-review framework. It is not a construction drawing, engineered EMC/bonding design, intrinsic-safety assessment, hazardous-location design, surge/lightning design, live-measurement procedure or substitute for the TIA/EIA-485 standard, applicable electrical rules and exact installed-equipment manuals. The generated illustrations communicate relationships and must not be used for terminal labels, pin numbers, conductor colours or resistor values.

Qualified personnel must control hazardous energy, assess arc-flash/electrical and process hazards, approve test instruments and methods, and authorise wiring or configuration changes. Browser simulation can help rehearse topology, half-duplex timing and frame evidence, but it cannot reproduce the installed cable impedance, ground potential, EMC environment, isolation or protection system.