One conductor per valve, all carried in a single multi-core cable to one connector on the terminal.

Inside the terminal, the connector's pins are wired individually to the solenoid coils - pin one to station one, pin two to station two, and so on - with one or more conductors serving as the common return.

At the other end, the cable terminates at the control cabinet, where each conductor lands on a digital output of the PLC. Energising that output energises that valve. There is no encoding, no addressing and no protocol.

The cable is normally a pre-made assembly with a moulded connector, which makes installation quick and reliable - a single plug replaces the individual wiring of every valve.

Because the arrangement is purely parallel, the terminal has no electronics beyond the coils themselves. That makes it robust and cheap, and it means there is nothing to fail electronically.

The conductor count in the cable and connector - each valve needs its own wire, so the count is a hard limit.

Common multi-pin connectors provide a fixed number of pins, and with allowance for the common return that translates directly into a maximum number of solenoids. Double-solenoid valves consume two.

Beyond that, either a second cable and connector is added or the arrangement must change.

There is also a practical limit before the theoretical one. As the conductor count rises the cable becomes thick, heavy and expensive, its minimum bend radius grows, and routing it through a machine's cable management becomes difficult - particularly on anything that moves.

And every one of those conductors occupies a PLC output, so the controller's output card count grows with the valve count.

That combination of cable bulk and output cards is what makes fieldbus cheaper above a certain number of valves, and it is worth calculating rather than assuming.

Simplicity, cost at low valve counts, and independence from any network.

There is nothing to configure. No addresses to set, no network parameters, no protocol compatibility to confirm between the controller and the terminal, and no software involved in getting an output to operate a valve.

Fault-finding needs only a multimeter: check for voltage at the pin, check continuity to the coil, and the problem is located. No diagnostic software, no network analyser and no specialist knowledge.

At small valve counts it is also cheaper - a multi-pin terminal has no bus node electronics, and a pre-made cable costs less than a fieldbus interface.

And it is independent: the terminal has no dependency on a network being healthy, on the bus master, or on a particular controller family. A machine built with multi-pin can be moved between controllers without touching the pneumatics.

For small standalone machines and jigs, those are real advantages.

No - and that is its principal weakness compared with a fieldbus terminal.

The controller sets an output and assumes a valve has operated. It has no way of knowing whether current actually flowed, whether the coil is intact, whether a conductor has broken, or whether the connector has worked loose.

So a failure presents as the machine misbehaving - a cylinder that does not move - and the diagnosis starts from the mechanical symptom rather than from an electrical report.

Some mitigation is possible: most terminals have an indicator per station showing the coil is energised, which lets a technician see at a glance whether the signal reached the valve. That separates electrical from pneumatic faults quickly, though only when someone is standing at the machine.

For machines where downtime is expensive or the terminal is inaccessible, the absence of remote diagnostics is a strong argument for fieldbus, quite apart from the wiring question.

To a documented pin-to-station map, with the coil polarity and the common arrangement confirmed before energising.

The terminal manufacturer publishes which pin operates which station, and that map should be reproduced in the machine documentation and reflected in the PLC's output labelling. Without it, a maintenance technician has a block of identical valves and no way to relate them to outputs.

Coil polarity matters on DC terminals with electronics or indicators, and the common arrangement - whether the shared conductor is positive or negative - must match the PLC's output type. Getting that wrong generally means nothing operates, but it can also damage the terminal.

Surge suppression is normally built into the terminal's stations; where it is not, it should be provided at the outputs.

Label both ends of the cable, and keep a spare cable assembly for machines where a damaged loom would be difficult to replace quickly.

When the valve count grows, when diagnostics are wanted, or when the machine is part of a larger networked installation.

The crossover on cost is driven by two things: the cable and the PLC output cards. Both scale with valve count on multi-pin and barely at all on fieldbus, so above a certain number of valves fieldbus is simply cheaper - and the calculation should include the output cards, which are often overlooked.

Diagnostics are the other driver. Where the machine is expected to report its own faults, where downtime is expensive, or where the terminal is physically hard to reach, the ability to see a broken coil from the control system is worth a great deal.

And where a machine sits on a plant network already, adding the valve terminal as another node is natural.

For a small standalone machine with a handful of valves, none of that applies and multi-pin remains the sensible choice.

Connector integrity and cable condition, since everything depends on one plug and one loom.

The single connector that makes installation easy is also a single point of failure. Check it is properly seated and its seal intact, particularly in washdown or vibrating environments, and confirm the retaining mechanism is secure - a connector working loose produces intermittent faults across several valves at once, which is confusing to diagnose.

The cable is exposed on the machine and can be damaged by abrasion, crushing or repeated flexing. Where it runs to a moving part, use cable rated for continuous flexing and manage it properly, since a broken conductor inside an intact sheath is a difficult fault.

Keep the station map available at the machine.

And hold a spare valve for each function. Because the valves are modular, a single spare covers every station of that type, and a valve change is a two-screw job once the block is exhausted.