PROFINET identifies devices by a device name, not primarily by IP address. Configuring a device means four things: install its GSDML file so TIA Portal knows the hardware, place it in the network view and give it a device name and IP in the project, write that same name to the physical device using Online access → Assign device name, then download the hardware configuration to the PLC. At startup the controller uses DCP to find the device by name and assigns the IP address itself. If the name in the device does not exactly match the name in the project, the device will not enter data exchange — no matter how correct the IP address is.
- Device name is the identity; IP address is assigned by the controller at startup. Get the name right first.
- Names follow DNS rules: lowercase, digits, hyphens and dots only — no spaces, no underscores, no capitals.
- Configuring the topology (which port connects to which) enables device replacement without a programming device.
- Update time and watchdog are a trade-off, not a "faster is better" setting — a too-short watchdog causes intermittent station failures on a busy network.
Device Name vs IP Address: Why PROFINET Uses Both
Engineers coming from PROFIBUS expect an address switch or a rotary selector. PROFINET has neither. Every IO device carries a device name stored in its non-volatile memory, and that name is what the IO controller looks for when the system starts.
The sequence at power-up is straightforward once you have seen it. The controller broadcasts a DCP "identify" request for each device name in its configuration. The device that holds that name answers. The controller then sends it the IP address, subnet mask and gateway from the project, and only after that does cyclic IO data exchange begin. The IP address is therefore a consequence of successful identification, not the means of it.
This design is what makes PROFINET convenient in the field: a failed remote I/O head can be swapped for a factory-fresh unit and, with topology configured, the controller writes the name and address to the replacement automatically. It is also what causes the most common startup failure — a device with the wrong name simply never answers, and the diagnostic buffer records a station failure rather than an addressing error.
Device names must follow DNS conventions: lowercase letters, digits, hyphens and dots. TIA Portal will silently convert what you type, so a station you named "Conveyor_01" in the project appears on the wire as something else, and the mismatch is invisible until you look at the assigned name in Online access.
Installing the GSDML File
Siemens devices such as ET 200SP, ET 200MP and most CPUs are already in the TIA Portal hardware catalogue. Third-party devices — a drive from another vendor, a network gateway, a smart sensor — are not, and their manufacturer supplies a GSDML file describing modules, submodules, IO sizes and diagnostic texts.
Install it through Options → Manage general station description files (GSD), pointing at the folder that contains the XML. Two practical points save time later: install the version that matches the device firmware, since a GSDML for an older firmware may not expose newer submodules; and keep the GSDML files alongside the project archive, because a colleague opening the project on another machine will get "unknown device" placeholders without them.
Practise PROFINET commissioning on real hardware
Live sessions with S7-1200/1500, ET 200SP remote I/O, managed switches and G120 drives — Pune classroom or online.
Assigning the Device Name and IP
Configuration happens in two places, and both must agree.
| Step | Where | What you set |
|---|---|---|
| 1. Project side | Network view → device properties → PROFINET interface | PROFINET device name, IP address, subnet mask, whether the IP is set in the project or by another route |
| 2. Device side | Online access → your PG interface → Update accessible devices | Right-click the device → Assign device name, and write the project name into the physical hardware |
| 3. Verify | Online access list, refreshed | The device now shows the assigned name and, after download, the project IP |
| 4. Download | PLC → Download to device → Hardware configuration | The controller receives the device list and begins DCP identification at startup |
A subtlety worth knowing: Online access shows devices found by DCP on the local network segment, which works even when the PG's own IP address is on a different subnet than the devices. That is deliberate, and it is why you can find and rename a device whose IP address you do not yet know. It is also why a device connected through a router rather than a switch will not appear — DCP is a layer 2 protocol and does not cross subnets.
Update Time and Watchdog Settings
Each IO device has an update time — how often the controller and device exchange process data — and a watchdog time, usually expressed as a number of accepted missed update cycles. TIA Portal calculates a default from the device count and network load, and the default is usually correct.
| Setting | Typical value | What goes wrong if set badly |
|---|---|---|
| Update time | 1–4 ms for most machines; 8–16 ms is fine for slow process I/O | Very short times on a network with many devices raise bus load and CPU communication load without improving control |
| Accepted update cycles without IO data | 3 (default) | Reducing it makes the system fragile — a single delayed frame triggers a station failure and stops the machine |
| Watchdog time | Update time × accepted cycles | Too short causes intermittent, hard-to-reproduce dropouts, typically blamed on cabling |
| Send clock | 1 ms default | Changing it affects every device on the IO system, not just one |
Resist the urge to set every device to the fastest possible update. A conveyor position sensor does not benefit from a 250 µs cycle, and the bandwidth is better left available. Reserve short update times for genuine motion and fast interlock signals.
Topology, Port Interconnection and Device Replacement
The topology view lets you draw which physical port on one device connects to which port on the next. Doing this is optional for basic operation but valuable in production for two reasons.
First, device replacement without a programming device. When topology is configured and the option is enabled on the controller's PROFINET interface, the controller can detect that a factory-fresh device has been plugged into a known position — identified through LLDP neighbourhood detection — and write the correct device name and IP to it automatically. A maintenance technician swaps the unit and the machine restarts. Without topology, someone must bring a laptop and assign the name manually.
Second, precise diagnostics. With topology configured, TIA Portal's online topology view shows exactly which link is broken rather than reporting a generic station failure, which turns a cable fault from a search into a pointed inspection.
For availability, media redundancy (MRP) can close the line into a ring with one device acting as redundancy manager. MRP recovery is fast enough for most process applications but is not a substitute for correct cable routing and strain relief.
Diagnostics: LEDs, TIA Portal and PRONETA
Work from the cheapest diagnostic to the most expensive.
| Tool | Best for | What it tells you |
|---|---|---|
| Device LEDs (RUN, ERROR, LINK) | First 30 seconds at the panel | Whether the link is physically up and whether the device has an active fault |
| Online access → Accessible devices | Name and address problems | Every DCP-reachable device with its current name, IP and MAC — the fastest way to spot a name mismatch |
| TIA Portal online & diagnostics | Configured systems | Diagnostic buffer entries with timestamps, module-level fault detail, topology comparison of set vs actual |
| SIEMENS PRONETA | Sites without the project file | Network scan, name and IP assignment, port statistics and cable test, without needing TIA Portal |
| Managed switch port statistics | Intermittent faults | CRC error and discard counters that reveal a marginal cable long before it fails completely |
The diagnostic buffer is the single most underused resource on an S7-1500. It timestamps every station failure and return, which converts "it drops out sometimes" into a pattern you can correlate with a shift change, a welder starting, or a specific machine cycle.
The Six Errors That Cause Most Commissioning Delays
| Symptom | Likely cause | Fix |
|---|---|---|
| Device visible in Online access but station failure in the PLC | Device name does not match the project exactly | Re-assign the device name; compare character by character, including case |
| Device not visible at all | No link, wrong port, or PG connected through a router | Check LINK LED, connect the PG to the same switch, verify the cable at both ends |
| Works from the PG but not after power cycle | Name written temporarily rather than stored permanently | Re-assign and confirm the name persists after a device power cycle |
| Intermittent station failures under load | Watchdog too short, or a marginal cable | Restore default accepted update cycles, then check switch CRC counters |
| Modules show as configuration error | Physical module arrangement differs from the project | Compare set and actual configuration online; slot order matters |
| Duplicate IP warning | An engineering laptop or spare device left on the network | Use PRONETA or Accessible devices to find the second holder of the address |
Step-by-Step Lab: Configure an ET 200SP on PROFINET
Hands-on- TIA Portal V17 or later, an S7-1200 or S7-1500 CPU, and an ET 200SP interface module with at least one digital input and one digital output module.
- A standard Ethernet switch and two patch cables, or a direct PLC-to-device connection.
- Your PG connected to the same physical network segment as the devices.
- Estimated time: 30 minutes.
Find what is actually on the network
Before configuring anything, open Online access, select your PG's Ethernet adapter and double-click Update accessible devices. Note every device found, with its current name, IP and MAC address.
Build the project-side configuration
In the network view, add the ET 200SP interface module, drag a PROFINET connection from the CPU port to the device, then populate the slots with the exact I/O modules present on the rail, in the same order.
Set a clean device name and IP
Open the device's PROFINET interface properties. Set the name to something like et200sp-line1 — lowercase, hyphen, no underscores — and set the IP address on the same subnet as the CPU.
Write the name to the physical device
Return to Online access, right-click the ET 200SP and choose Assign device name. Pick the project name from the list and assign it. Refresh the accessible devices list.
Download and verify data exchange
Download the hardware configuration to the CPU and go online. Open a watch table with one input byte and one output byte from the ET 200SP and toggle a physical input.
Break it deliberately, then read the diagnostics
Unplug the ET 200SP cable for five seconds, reconnect it, then open the CPU's diagnostic buffer.
The configuration is genuinely correct if the device recovers on its own after a cable pull and a power cycle, with no manual intervention and no entries in the buffer other than the expected failure and return pair. If you had to re-assign the name after a power cycle, the name was written to volatile memory only — repeat step 4. As a next step, configure the topology and enable device replacement without a programming device, then test it by swapping in a factory-reset module.
Frequently asked questions
Why does PROFINET use a device name instead of just an IP address?
The controller finds each device by name using DCP at startup and then assigns the IP address from the project. This means a replacement device needs no pre-configured address — the controller gives it one — which is what allows a failed unit to be swapped without a programming device when topology is configured.
What are the rules for a PROFINET device name?
Names follow DNS conventions: lowercase letters, digits, hyphens and dots only. Spaces, underscores and capital letters are not permitted. TIA Portal converts non-conformant input automatically, which can create a mismatch between what you typed and what is actually assigned, so always verify the name in Online access.
My device appears in Online access but the PLC reports a station failure. Why?
Almost always the device name stored in the hardware does not exactly match the name in the project. DCP discovery finds the device regardless of name, so seeing it in the accessible devices list proves only that the physical link works. Compare the two names character by character and re-assign.
What update time should I use?
Accept the default TIA Portal calculates unless you have a specific reason not to. Values of 1–4 ms suit typical machine I/O and 8–16 ms is ample for slow process signals. Setting every device to the shortest possible time raises network and CPU communication load without improving control quality.
Do I have to configure the topology?
No, the system runs without it. Configure it when you want device replacement without a programming device, or when you want the online topology view to identify exactly which link has failed instead of reporting a generic station failure. Both benefits matter more in production than during commissioning.
Can I configure a PROFINET device without TIA Portal?
Yes. SIEMENS PRONETA scans the network, shows every reachable device, and can assign device names and IP addresses without the project file. It is the practical choice for a service visit to a site whose project you do not have.