Resolving PROFINET Bus Fault on S7-1515-2PN Beyond 26 G120 Drives

1. Problem Statement

A PROFINET installation built on a SIMATIC S7-1515-2PN (article 6ES7515-2AM02-0AB0, firmware V2.5) running under TIA Portal V15.1 stops accepting additional devices beyond a hard threshold of 25-26 nodes. The node list comprises two IM155-6 PN HF distributed I/O stations, five SIRIUS 3RW55 soft starters, five S7-1510 SP CPUs, five KTP700 Basic PN HMI panels, and twelve SINAMICS G120 frequency inverters (CU240E-2 PN, firmware V4.7).

The symptom is a deterministic bus-fault event: as soon as a 26th IO device is brought online, exactly three G120 inverters enter the BF (Bus Fault) state in a non-deterministic order, while the other 23 nodes continue to run unaffected. The BF condition clears only when the offending device is physically removed or the G120 is power-cycled. No fault is logged in the S7-1500 diagnostic buffer. No substitution-level error is present in the IM155 stations. The HMI panels stay in runtime.

This pattern is characteristic of a PROFINET update-time/watchdog collapse rather than a hard connection-resource exhaustion. The threshold of 26 devices is a project-specific result of the auto-calculated send clock and the G120 submodule layout, not a hard ceiling on the S7-1515-2PN IO controller.

2. Network Topology Overview

Figure 1 - Project topology. The lower line of G120 drives loses PROFINET AR whenever the controller attempts to set up the 26th AR.

3. Root Cause Analysis

The fact that only G120 drives drop, while IM155, S7-1510 SP, KTP700 and 3RW55 continue cycling, is a key diagnostic clue. The G120 is the only device class that opens the largest number of PROFINET submodules per station. A typical CU240E-2 PN with Standard Telegram 1 plus Free Telegram 1 and Free Telegram 2 exposes three IO submodules (Slot 0, Slot 1, Slot 2). Each submodule requires its own CR (Communication Relationship) on the controller side. The G120 also has the lowest acceptance watchdog tolerance in this device mix because its firmware 4.7 still defaults to a 3-cycle watchdog and a 1 ms send clock.

When TIA Portal V15.1 auto-calculates the PROFINET update time for a freshly added 26th device, the resulting calculation pushes the reduction ratio past the G120's tolerance band, so three drives fail to refresh their AR before the watchdog timer expires. The S7-1500 interprets the late AR as a broken link and de-asserts the IO controller's output CR, which manifests as BF on the drive.

There is no firmware bug and no hardware defect. The fault is an engineering-time configuration issue: TIA Portal cannot always compute the optimum update time for 30+ heterogeneous IO devices, especially when the project contains submodules that require deterministic slots. Manual tuning of the send clock, reduction ratio and watchdog time resolves the issue in nearly every case.

4. S7-1500 PROFINET IO Resource Reference

Verify the per-CPU resource ceilings before assuming a hard limit is the cause. For the S7-1515-2PN (6ES7515-2AM02-0AB0) with firmware V2.5 the published values are:

Source: S7-1500 system manual, S7-1500 Automation System / System Manual, section "PROFINET interface". The values quoted apply to firmware V2.5; later firmware releases may raise some limits.

5. SINAMICS G120 PROFINET Specifics

The SINAMICS G120 with CU240E-2 PN is a typical PROFINET IO device with the following slot map when configured for the common drive profile:

Source: SINAMICS G120 Operating Instructions, section "PROFINET communication".

For every active submodule, the controller reserves a CR. With the configuration shown (Telegram 1 + Free 1 + Free 2), each G120 opens 3 CRs. Twelve drives thus reserve 36 CRs. Once the project grows past 25-26 IO devices, the cumulative slot count crosses the threshold where the auto-calculated reduction ratio becomes insufficient.

6. Update Time, Watchdog and Reduction Ratio Tuning

The three parameters that govern PROFINET IO deterministic refresh behaviour are:

The relationship is:

T_update = T_s × R_r
T_w ≥ 3 × T_update  (recommended by Siemens for RT)

For the project in question, TIA Portal's automatic calculation at 25 devices yields T_s = 0.5 ms and R_r = 4 for the G120 devices, giving T_update = 2 ms. The watchdog defaults to 3 × 2 ms = 6 ms. As soon as the controller adds the 26th AR, the IRT phase planner extends the reduction ratio of the G120s to R_r = 8 to make room, doubling the update time to 4 ms. The watchdog is not re-evaluated automatically and the drives drop because the response is now outside the 6 ms window.

7. Step-by-Step Resolution Procedure

Apply the following actions in sequence. Each step is independently testable. Do not skip the verification step at the end of each sub-procedure.

7.1 Pre-flight check

1. Open the TIA Portal project and go offline. Confirm that the S7-1515-2PN has firmware V2.5 in the device configuration (Properties > General > Catalog). Update the controller to at least firmware V2.6 or V2.8 if you are working with a current G120. The S7-1500 firmware files are available from the S7-1500 firmware update entry page.
2. Update the G120 firmware to at least V4.7 SP14 or V5.1 SP3. SINAMICS G120 firmware is delivered via the SINAMICS Startdrive plug-in or the SINAMICS firmware update portal. Later firmware versions extend the watchdog tolerance and improve PROFINET AR retry logic.
3. Inventory the project. Note the slot count per station and the cumulative number of submodules. The 26-device symptom is rarely caused by a single device - it is a system-level effect.

7.2 Reduce the slot footprint of the G120

1. Open the device view of one G120 in TIA Portal. Go to Device view > G120 > Properties > PROFINET interface > Telegram configuration.
2. Confirm that only Standard Telegram 1 is selected. Disable Free Telegram 1 and Free Telegram 2 unless they are actually used in the program.
3. Compile the hardware configuration. Re-download to the controller.
4. Verify with the online G120 that the BF LED is off. The drive should be in RUN state with the PROFINET "LINK" LED steady green.

7.3 Manually configure the update time and watchdog

1. In TIA Portal, navigate to the G120 device in the network view. Open Properties > PROFINET interface > Real-time settings > Update time.
2. Disable "Automatically calculate update time". Set the update time to a fixed value. Recommended starting point: 4 ms (T_s = 0.5 ms, R_r = 8).
3. Open the same device's Properties > PROFINET interface > Watchdog time. Set it to 12 ms (3 × 4 ms).
4. Repeat the same fixed update time and watchdog on all twelve G120 devices, all five 3RW55 soft starters and both IM155-6 PN stations. HMI panels can stay on the default (typically 50-100 ms).
5. Save and compile. In the project tree, right-click the S7-1500 station and choose Compile > Hardware (rebuild all). This regenerates the PROFINET IRT phase plan and the AR table.
6. Download the hardware configuration to the S7-1500. Cycle power to the controller if TIA Portal prompts you to do so.

7.4 Verify the new AR budget

1. After download, go online with the controller. Open Online & Diagnostics > PROFINET diagnostics > IO devices.
2. Confirm that all IO devices are in state "OK". The actual send clock, reduction ratio and update time are reported per device in this view. They must match the values you set.
3. Right-click the controller and choose Online & Diagnostics > Diagnostics buffer. There should be no PROFINET IO error entries.
4. Open Web Based Management of the SCALANCE switches (https://<switch IP>) and check Information > PROFINET > Statistics. There must be no cyclic frame loss or out-of-order frames.

8. Optional Hardening for Larger Installations

When the project grows beyond 30 IO devices or when the G120 drives are distributed across multiple line topologies, apply these additional measures:

8.1 Topology editor in TIA Portal

Use the TIA Portal Topology editor (Network view > Topology view) to declare every port-to-port link. The IRT phase planner uses this topology to compute the most efficient scheduling. Without an explicit topology, TIA Portal assumes a worst-case star layout and over-allocates time slots.

8.2 Use PROFINET IRT with TOP synchronization

For G120 firmware 4.7 and later, switch the controller to IRT mode with TOP (Time Optimized Planning). TOP requires an explicit topology but removes the reduction-ratio interference that drives the random BF behaviour seen in pure RT mode. Procedure:

1. Open the controller's PROFINET interface properties. Set Real-time setting > RT class to IRT.
2. Enable Synchronization role > Sync master (1515-2PN is the only sync master on the subnet).
3. Open the Topology view and connect the ports end-to-end. Mark every port as "is partner port".
4. Recompile. The IRT phase plan is regenerated and the G120 update time is calculated deterministically.

8.3 Segment the network with SCALANCE XC-208

Split the twelve G120 drives and the five soft starters onto a dedicated SCALANCE switch pair. Each segment is then a 12-device broadcast domain, well below the practical AR limit. Use managed SCALANCE XC-208 (6GK5208-0BA00-2AC2) or SCALANCE XC216 (6GK5216-0BA00-2AC2) switches that support PROFINET diagnostics. Reference: SCALANCE XC-208 Operating Instructions.

9. Verification Matrix

10. Preventive Checklist for Future Expansions

• Lock the send clock and reduction ratio for every G120 drive from the first commissioning. Do not let TIA Portal's auto-calculation run unattended on a large project.
• Keep G120 firmware on a uniform release across the entire fleet. Mixed firmware (4.7 and 5.1) in the same line can produce divergent watchdog behaviour.
• Add free telegrams only when the application actually reads or writes the corresponding PZD words. Every free telegram reserves an additional CR and a slot in the IRT schedule.
• Document the slot count of each device class. Maintain a per-segment budget: do not exceed ~30 active CRs per controller port in RT mode without enabling IRT/TOP.
• For controllers that need more than 30 IO devices, switch to a CPU 1516-3 PN/DP (6ES7516-3AN02-0AB0) or 1517-3 PN/DP (6ES7517-3AP00-3AB0). These support 256 and 512 IO devices respectively and provide larger IRT phase windows.

11. Related Fault Patterns

The same BF pattern is observed when a different controller parameter is the root cause. Use the table below to map symptoms to causes:

12. Field Notes for Engineers Commissioning This Configuration

• Always check the S7-1500 diagnostic buffer and the online PROFINET view together. The diagnostic buffer logs controller-side events; the online view shows the device-side state. The two are not identical.
• The "BUS FAULT" text on a G120 OP panel (e.g. BOP-2 or IOP-2) is sometimes F08501 (PROFINET communication fault) or F08502 (sign-of-life failure). Both are symptoms of a watchdog trip, not the root cause.
• If the 1515-2PN is also acting as a PN/PN coupler for legacy PROFIBUS, the number of available ARs is unchanged but the IRT phase window is reduced. Do not run a PN/PN coupler on the same subnet as a 30+ device RT line; place it on a separate physical interface or use a dedicated IE/PB link.
• When a soft starter is the failed device, remember that the SIRIUS 3RW55 itself contains an internal PROFINET stack with its own watchdog. Its default is 100 ms - well above the G120 default. The 3RW55 will not normally fail first.