Few topics cause more field head-scratching than sinking versus sourcing. A sensor that reads perfectly on the bench refuses to turn on an input, and the culprit is almost always a mismatch in current direction rather than a broken part. The confusion is made worse by a second vocabulary - PNP and NPN - that describes the same idea from the sensor's side. This guide untangles both: what sinking and sourcing actually mean in terms of which way current flows, how PNP and NPN sensors map onto them, and the simple rule for matching a sensor to a DC input card so it works the first time.
Sinking vs Sourcing I/O in one line: Sinking and sourcing describe the direction DC current flows through an I/O point. A sourcing device supplies (sources) positive current out to the load, while a sinking device provides the return path, pulling current down to the negative common. To work, a sourcing sensor must connect to a sinking input, and a PNP sensor pairs with a sinking-input card while an NPN sensor pairs with a sourcing-input card - always a source paired with a sink so a complete current path exists.
The whole idea comes down to which side of the DC supply a device connects the current to. Think of a simple loop: current has to leave the positive terminal of the supply, pass through the device, and return to the negative terminal. A sourcing device is the one that provides the positive current going out - it sources current into the load. A sinking device is the one that provides the path back to the negative common - it sinks the current down. For a loop to work, exactly one end has to source and the other has to sink; two sourcing devices or two sinking devices leave no complete path, and nothing turns on.
This applies to both inputs and outputs. A sourcing output pushes current out to a load that returns to common, so it needs a load that sinks. A sinking input provides the return path for a signal, so it needs a source feeding it. The single mental model - one side sources, the other side sinks - covers every case. Where people get lost is that the same physical wire can be described from either end, so the same connection is called sourcing by one person and sinking by another depending on which device they are standing at.
Sensor manufacturers usually label three-wire DC sensors as PNP or NPN, named for the transistor that does the switching. A PNP sensor switches the positive supply to its output - when it activates, it connects its signal wire to positive, so it sources current. An NPN sensor switches the negative side - when it activates, it connects its signal wire to the negative common, so it sinks current. Once you translate PNP to sourcing and NPN to sinking, the two vocabularies collapse into the same picture.
The pairing rule follows directly. Because a source must meet a sink, a PNP (sourcing) sensor must feed a sinking input, and an NPN (sinking) sensor must feed a sourcing input. Regional habits add to the confusion here: PNP wiring is dominant in much of Europe while NPN has been historically common in parts of Asia, so imported machinery can arrive expecting the opposite of what a plant standardizes on. When a new sensor does nothing, the fastest check is often to confirm whether it is PNP or NPN and whether the input card it is wired to expects the matching sink or source - a mismatch there explains a dead-but-undamaged input more often than any actual failure.
Sinking and sourcing decisions ripple all the way up to the SCADA screen. If a proximity sensor is wired against the input card's polarity, its status simply never changes, and an operator watching a Merobix dashboard sees a piece of equipment that appears permanently idle or permanently tripped for no obvious reason. The data looks like a process problem when it is really a wiring problem, and troubleshooting it remotely without knowing the sensor and card types can waste a lot of time.
Standardizing on one convention across a facility is the practical defense. Many plants pick either all-PNP or all-NPN for DC field devices and specify input cards to match, so a technician wiring a new instrument does not have to rediscover the polarity each time. Clear drawings that show whether inputs are sinking or sourcing, labeled commons, and sensors chosen to fit the standard all reduce the odds of a silent mismatch. For unmanned sites feeding a cloud SCADA, that discipline is what keeps a status point on the screen honest, because there is no one on location to notice that the sensor is fine and only the wiring is wrong.
A PNP sensor is sourcing. When it activates it switches its output to the positive supply, pushing positive current out toward the load. Because a source must meet a sink, a PNP sensor must be wired to a sinking input card so a complete current path exists.
Nothing turns on. Two sinking devices both try to provide the return path to negative, so there is no source pushing current through the loop and no complete path. The input never sees current and stays off, even though both the sensor and the card are perfectly healthy - it is a polarity mismatch, not a failure.
The same physical connection can be described from either end, so one device sources while the other sinks and both terms refer to the same wire. On top of that, sensors use the PNP and NPN labels for the same idea, and regional wiring conventions differ. Translating everything into current direction - one side sources, the other sinks - removes the confusion.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.