PROFIBUS DP Class 1 vs Class 2 Master

Back to Blog

The Class 1 Master Runs the Process

The DPM1 is the active heart of a DP bus. It holds the configuration of the slaves it owns, exchanges cyclic input and output data with each of them every bus cycle, and monitors them for faults. This is the master doing the real control work, and on a typical line it is the PLC. Each DP slave is parameterized and owned by exactly one Class 1 master, which is what gives the bus a single, clear controller of record for its process data.

Because the DPM1 drives the cyclic exchange, it is also what detects a slave that fails to respond, reports a diagnosis, or falls out of the expected configuration. It runs continuously as long as the process runs. Its counterpart in the PROFINET world is the IO controller, and the two play the same owning-and-driving role in their respective networks.

The Class 2 Master Configures and Observes

The DPM2 is the engineering and diagnostic role. It is typically a laptop or a programming device that connects to the bus temporarily to set slave addresses, download configurations, read diagnostics, and monitor live values during commissioning or troubleshooting. It does not run the cyclic control loop, so you can attach it to a running bus to look at what is happening and detach it again without stopping the process.

This is exactly where token passing earns its keep: the Class 2 master shares the bus with the Class 1 master because the token arbitrates access, so a diagnostic tool can coexist with live control instead of colliding with it. The distinction maps cleanly onto PROFINET, where the DPM2 role corresponds to the IO supervisor used during commissioning. The practical rule is the same in both worlds: the Class 1 master or IO controller owns and drives; the Class 2 master or supervisor configures and observes and is not required for the process to keep running.

The Two Roles Side by Side

	Class 1 master (DPM1)	Class 2 master (DPM2)
Typical hardware	The PLC or process controller	An engineering laptop, programming device, or asset-management station
Presence on the bus	Permanent; runs as long as the process runs	Temporary; attached for commissioning or diagnostics
Relationship to slaves	Parameterizes, owns, and exchanges cyclic I/O data	Reads diagnostics and monitors; acyclic access only
Required for production	Yes	No

The table also explains a habit worth keeping: never assume a bus has exactly one master just because only one PLC is in the drawing. Any time an engineering tool is plugged in, the bus is a multi-master network, and its timing parameters were - or should have been - configured with that in mind.

Token Timing and Addressing When a DPM2 Joins

Both master classes are active stations: they take turns holding the token, and only the token holder may initiate traffic. That has two practical consequences for plugging in an engineering station. First, the address you give the DPM2 must be a free station address at or below the configured highest station address - the HSA determines which addresses participate in the token ring, and a master parked above it never receives the token and simply cannot talk. The mechanics are covered under the PROFIBUS station address and HSA.

Second, an extra token holder adds to the rotation. The bus parameters set at configuration time, including the target rotation time, determine how much slack exists for a visiting master's traffic; a bus configured with no allowance for a second master can see its cycle stretch when a tool attaches. How the token circulates among masters and how the cyclic exchange fits inside each token holding is described under the DP bus cycle and token passing. Whether a given bus has that slack is a property of its configured parameters, so check the project settings rather than assuming.

The Acyclic Services a DPM2 Actually Uses

A Class 2 master does its work through acyclic services rather than the cyclic exchange. It establishes a temporary connection to a slave alongside the slave's cyclic relationship with its owning DPM1, reads or writes parameter records, then releases the connection. These acyclic read and write services arrived with the DP extensions - the capability split across protocol generations is laid out in DP-V0, DP-V1, and DP-V2 - and they are what make it possible to configure an intelligent field device in a running system without disturbing its cyclic I/O.

This is also why the DPM2 role matters so much on installations with PROFIBUS PA instruments hanging off the DP backbone: device parameterization tools reach those instruments through exactly this acyclic path, while the DPM1 keeps exchanging the cyclic process values the whole time. Owner and visitor use different doors into the same slave, and neither blocks the other.

Before You Plug In: A Short Field Checklist

1. Confirm a free station address for the tool, at or below the HSA, and set it before connecting.
2. Match the configured baud rate; a mismatched station can disturb a live bus.
3. Connect at a proper bus connector or dedicated service socket rather than improvising a tap.
4. After attaching, watch bus health for a moment - slave diagnostics and communication status - before starting any downloads.
5. When finished, disconnect cleanly and confirm the bus returns to its normal state.

None of this is exotic, but the failure stories all start the same way: someone attached a tool with a conflicting address or the wrong baud rate to a bus that was running production. The token mechanism makes coexistence possible; the checklist makes it uneventful. On anything feeding a live process, coordinate with operations before attaching, per site procedures.

Frequently Asked Questions