NPN (Sinking) and PNP (Sourcing) are two types of transistors used for digital outputs. The terms Sinking and Sourcing refer to current flow with respect to the terminal pin on the IO card. A device is called sinking if current flows into the terminal, and is called sourcing if current flows out of the terminal.

• • PNP transistor outputs are called “sourcing outputs” because they source current from the output to the load.

• NPN transistor outputs are called “sinking outputs” because they sink current from the load into the output.

The question of sinking vs. sourcing and NPN vs. PNP causes confusion throughout the controls industry. Often engineers working on different parts of a system will not communicate the chosen configuration. These errors will inevitably be discovered, the question is when and at what cost?

This paper will define these terms and what they mean with respect to digital input and digital output circuits used in industrial control systems. Then we will explore how these components are used and some of the issues to be aware of when selecting components for a system.

Definition of Terms

The terms NPN and PNP are commonly used as synonyms for sinking and sourcing respectively. This is a reasonable assumption but strictly speaking they refer to different concepts.

The terms NPN and PNP refer to how a transistor is constructed. A transistor is a three layer sandwich of two different types of material, the two outside layers are the same and the middle layer is different. The magic of the transistor is that the layer in the middle can be used to control the flow of current between the outside layers. The two types of material that can be used to build a transistor are n-type and p-type. Thus, there are two possible types of transistor: NPN and PNP. The practical difference between these two transistor types is in how current flows through the circuit. In a bipolar junction transistor (BJT) the control leg, the middle of the sandwich, is called the base. The other two connection points are called the collector and the emitter. In a PNP transistor a small current flowing from the emitter into the base causes a larger current to flow from the emitter to the collector.

On a circuit diagram this is indicated by a small arrow on the emitter Pointing iN to the base. 

In an NPN transistor a small current flowing from the base to the emitter causes a larger current to flow from the collector to the emitter.

On a circuit diagram this is indicated by a small arrow on the emitter pointing away from the base (or Not Pointing iN to the base).

Note that in a PNP transistor current flows from the emitter to the collector. While for an NPN transistor current flows from the collector to the emitter.

So, how does the choice of an NPN or PNP transistor relate to the issue of sinking vs sourcing? The terms sink and source are used to indicate which way current is flowing with respect to some reference point. In the controls world, the reference point is typically a screw terminal on an output card. If current flows into the screw terminal the device is said to sink current. If current flows out of the screw terminal it is said to source current. The connection between sinking vs. sourcing and NPN vs. PNP has to do with which type of transistor is used to create the circuit and where it is placed relative to the load being controlled.

In a PNP (BJT) transistor, to get current to flow from the emitter to the collector, a small control current must flow from the emitter into the base. Thus the base must be held at a lower voltage than the emitter.
If the load were connected between the power supply and the emitter the voltage
at the emitter will be pulled down as the transistor begins to conduct. That is
when no current is flowing through the load there is no voltage drop across the
load. As current begins to flow the voltage drop across the load is given by
Ohm’s law: VL = IRL. This voltage drop makes it harder to
maintain a voltage at the base that is lower than the voltage at the emitter.

Thus for a PNP transistor the load is usually connected to
the negative side of the transistor, which is the collector in the case of a
PNP transistor. In a digital output the load is connected to the output pin on
the IO module, so current flows out of the pin and into the load, hence this
type of output is said to source current.

In an NPN transistor current must flow from the base into
the emitter to activate the transistor. This requires that the base is held at
a higher voltage than the emitter. If the load were connected between the
emitter and 0V this would become harder to do as the transistor began to
conduct current and the voltage drop across the load begins to rise. So, when
using an NPN transistor the load is placed before the transistor, between the
power supply and the collector. In an output module the load is connected to
the output pin so current flows from the load into the terminal and this type
of output is called a sinking output.

For our purposes we will follow convention and assume NPN =
Sinking and PNP = Sourcing.

Putting Things Together

Let’s explore the different configurations possible using
NPN / Sinking and PNP / Sourcing, inputs and outputs and some of the issues
involved. In a typical digital control circuit there are three components that
need to be connected together, in the correct order, to make the circuit work.
These are; the power supply, the output and the input.

Since we are discussing DC circuits the power supply will
have positive and negative connections, typically 24VDC and a 0VDC.

 The output could be several different types of devices; a
switch, a relay contact, or a sinking or sourcing transistor output. A transistor
output will be either a sinking (NPN) output that can only be used after the
load (the load being the input in this case). Or, it will be a sourcing (PNP)
output that can only be used before the load.

 All of the outputs on
the IO card will have their negative terminals tied together inside of the card
and connected to a common pin. This common pin can only be connected to 0V in
the case of sinking (NPN) outputs or only to 24V in the case of sourcing (PNP)
outputs:

Relays and switches are more flexible, they can be wired to
source or sink current.

The final piece of the puzzle is the input. For this
discussion, the input is modeled as an LED and current limiting resistor.

This is a simplification of an optically isolated input. To
turn the input on you allow current to flow through the diode so that it emits
light. That light activates a photo transistor, which in turn sets the input
bit to true. With this type of input circuit there really isn’t a good reason why
it can’t be connected directly to 24VDC on the power supply when using a
sinking (NPN) output to drive it. Or, directly to 0v on the power supply, after
the output, when using a sourcing (PNP) output. As long as sufficient current
is flowing through the diode it will turn the input on.

The problem with this type of input is that there are
several inputs on one IO module and they all have one side tied to a common terminal
pin. A sourcing only input will have the 24VDC connection made internally. And
a sinking input will have the 0V connection made internally.

If you put the switch (output) on the common pin you would
switch all of the inputs at once. If you try to reverse the polarity of the
power supply the diode will not let current flow backward through the input,
regardless of the state of the output. One way to solve this problem is to use
two diodes for each input. These are placed in parallel facing opposite
directions.

 With this configuration, if the power supply is reversed one
of the diodes will continue to function. Indeed, many manufacturers (including
Maple Systems) supply bidirectional inputs that can function in either
direction. Note that the power supply will be connected to all of the inputs sharing
a common terminal. As a consequence the entire module can be used in sinking or
sourcing configuration but cannot have a mix of sinking and sourcing inputs
using the same common pin.

There are good design reasons to prefer a sourcing output
with sinking input in one situation and a sourcing output and sinking input in
another situation.

When purchasing components for a system it is wise to
maintain the most flexibility possible. For example, when deciding on an output
card or sensor that could be used in many different situations it is a good
idea to use relays with Normally Closed and Normally Open contacts. These can
be wired to sink or source depending on the situation. That said, there are reasons
to use transistor outputs that often trump the need for flexibility. Chief
among these are; speed, cost and space requirements.

When sourcing inputs use bidirectional inputs whenever
possible. This will maintain the most flexibility possible. It really does not
increase the cost of the Input card at all, so unless there is a specific
reason to get a sourcing only or sinking only input bidirectional is the way to
go.

So what happens, when despite the best intentions, a need
arises to connect a sinking output to a sinking only input or a sourcing output
to sourcing only input? For example, a systems integrator always buys sinking
only inputs but a particular sensor manufacturer only has sensors with sinking outputs.
No one caught the conflict in the design phase and now it’s time to commission
the system. What can be done? It is possible to use a resistor to convert a
sinking output into a sourcing output, or a sourcing output into a sinking one.
In the case of a sinking output simply use the resistor as the load instead of
the input:

When the output is off, current only flows through the
input, turning it on. When the output is on, current flows through the
transistor, pulling the voltage at the low side of the external resistor down
to nearly 0V and turning the input off. The transistor turns off the input by
steeling all the available current.

To turn a sourcing output into a sinking output again, the
resistor replaces the input as the load. This time the resistor is placed
between the output pin and zero volts on the power supply.

 When the output is off, current can only flow through the
input; turning it on. When the output is on a larger amount of current flows
through the transistor to the external resistor, bringing the high side of the
resistor to nearly the power supply voltage, and turning the input off.

In both cases the resistor value must be selected so that,
when the output is off, enough current flows through the external resistor to
turn the input on but, when the output is on, the output can sink or source
enough current through the resistor to lower or raise the voltage enough to
turn the input off. Inputs typically have very high impedance so this is not a
problem.

This can be done in a pinch, but there will be a discrete
resistor floating around between screw terminals somewhere. This makes for a
less tidy installation and will probably cause confusion when someone tries to
upgrade or maintain the system, so it’s generally not recommended. Better to
plan ahead and use components with as much flexibility as possible.