Troubleshooting SINAMICS S120 CU320 PROFIBUS DP Communication

1. Problem Overview

Communication breakdowns between a SIMATIC S7 CPU with integrated PROFIBUS DP master and a SINAMICS S120 drive line-up controlled by a CU320 (Control Unit) are not uncommon during commissioning and early production ramp-up. The pattern that drives this article is the most common one observed in the field:

1. Initial PROFIBUS DP cyclic data exchange between the PLC and the CU320 is healthy. The drive follows setpoints and returns actual values without faults.
2. STARTER (Siemens' commissioning tool) suddenly loses online access to the CU320 while the S7 still exchanges process data successfully. Browsing the bus with "Accessible nodes" reports the CU320 at its configured PROFIBUS address, but STARTER cannot identify the device type or load the online configuration.
3. After a power cycle, all PROFIBUS DP communication is lost. Neither the S7 nor STARTER can reach the CU320, even though the green power LED on the Control Unit is on and the drive appears to run its internal boot sequence.

This cascading failure almost always points at one of two root causes: (a) a corrupted parameter image on the CU320 CompactFlash card, or (b) a PROFIBUS configuration mismatch between PG/PC interface and the DP master in the STEP 7 hardware configuration. Both can be diagnosed and corrected without replacing hardware, and the procedures below cover both cases.

2. Affected Components and Identification

The recovery procedures below assume the following hardware and software set, which is the configuration most often reported for this failure mode. Substitute the version fields with whatever is installed on the machine before applying the steps.

If the project has been migrated to TIA Portal, the same diagnostic flow applies, but the access point and PG/PC interface are configured in the TIA Portal device access list instead of the legacy "Set PG/PC Interface" control panel applet.

3. Root Cause Analysis

There are five recurring root causes for the symptoms described. Walk through them in the order shown before opening the cabinet, because most incidents are resolved at step 1 or 2.

3.1 PROFIBUS Baud Rate Mismatch

The DP master configured in STEP 7 hardware (HW Config) sets the bus baud rate. If STARTER is launched with a different baud rate on its PG/PC interface — for example, "Auto" or "1.5 Mbps" while the master is running at "12 Mbps" — the PG can see the slave's request frames (because it listens passively) but cannot hold a token and cannot complete parameterization. The drive shows up at the right PROFIBUS address but STARTER fails to identify it.

3.2 PG/PC Access Point Set to the Wrong Interface

The "Access point of the application" must be set to S7ONLINE for STEP 7 and STARTER to use the CP5512 as the active PROFIBUS interface. If the access point is left on a non-existent Ethernet adapter, the tool loops at startup with a "no online interface found" error, even though the CP5512 is physically present and the bus is healthy.

3.3 Corrupted Parameter Image on the CompactFlash Card

The CU320 stores its parameter set on the CF card under \USER\SINAMICS\DATA as files with the pattern PS00xxxx.ACX and under \USER\SINAMICS\DATA\RT_DIR\CFG\TOCON as files matching TO00xxxx. Partial writes from a power loss, an aborted firmware download, or a mismatched STARTER version can leave these files in an inconsistent state. The CU320 may still power up, but the PROFIBUS slave can fail to complete its startup because the parameter set cannot be loaded. The PLC then sees no slave response at all.

3.4 Firmware / STARTER Version Drift

Older CU320 firmware (V2.x) ships with a GSD file that lists a specific device identity. If STARTER has been upgraded past a service pack that no longer recognizes that device ID, "Accessible nodes" will list the slave without being able to classify it. The drive still exchanges data, because the GSD on the master side is unchanged, but STARTER cannot open the project online.

3.5 Physical Layer Faults

Less frequent, but still common: a damaged PROFIBUS connector with a missing terminating resistor, a broken shield, a connector plugged into the wrong socket, or a daisy-chain cable that has been pulled out at an intermediate node. These typically drop the entire segment, not just one slave.

4. Diagnostic Sequence

Run the checks in this order. Each step is designed to isolate one root cause with the smallest possible disturbance to the running process.

1. Verify physical connectivity. Check the bus terminators on the first and last nodes. Measure the voltage between pin 6 (VP) and pin 5 (DGND) on a spare connector: a healthy segment reads between 4.75 V and 5.25 V. Lower than 4.5 V indicates an overloaded bus or a short. Look at the LED on the CP5512 or on the CPU 314C-2DP: the "BF" (bus fault) LED should be off once a slave is found.
2. Read PLC diagnostics. In STEP 7, open the online view of the CPU 314C-2DP and check the diagnostic buffer for DP slave errors. A repeating "Slave not found" or "Station failure" entry confirms that the master cannot see the slave at all. If instead the buffer shows "Configuration error", the issue is on the parameter side rather than the physical layer.
3. Test PG/PC access independently of the PLC. Disconnect the PLC from PROFIBUS or temporarily remove its bus connector, leaving only the CU320 and the CP5512 on a short stub. Try STARTER's "Search for online drives" or "Accessible nodes" from this minimal segment. This proves whether the PG can reach the CU320 at all.
4. Inspect STARTER online state. If STARTER can browse the bus but cannot identify the CU320, the issue is baud rate or device identity. If it cannot even browse the bus, the issue is access point or driver.
5. Confirm CF card state. If STARTER cannot connect after a power cycle but the green RDY LED on the CU320 is on, power down, remove the CF card, mount it in a card reader, and check that the directories listed in section 3.3 are populated and the files are non-zero in size.

5. Step-by-Step Recovery: PROFIBUS and PG/PC Configuration

Complete the steps below before any drive-side intervention. They solve the most common incidents without touching the CU320 firmware.

5.1 Re-add the S120 to the STEP 7 Hardware Configuration

If the S120 has been added to the STEP 7 project with the wrong GSD revision, delete the slave from HW Config and reinsert it from the catalog. The GSD file must match the firmware version of the CU320.

1. Open the SIMATIC Manager and the S7 project that drives the machine.
2. Open HW Config and double-click the DP master system on the CPU 314C-2DP.
3. Right-click the SINAMICS S120 slave and select "Delete". Confirm the prompt. Save and compile.
4. Open the hardware catalog, navigate to "PROFIBUS DP > Drives > SINAMICS > SINAMICS S120" and drag the correct slave type onto the DP master system.
5. Set the PROFIBUS address to 5 (or whatever address matches the DIP switch on the CU320), then assign the same PZD telegram layout the drive was originally commissioned with (typically Standard Telegram 1, 2, 3, or 4, or a Siemens-specific free telegram 999).
6. Save, compile, and download the hardware configuration to the CPU 314C-2DP.

5.2 Match the PROFIBUS Baud Rate

STEP 7 hardware config sets the master baud rate. Make STARTER use the same value.

1. Open "Set PG/PC Interface" from the Windows Control Panel (or from STARTER: Options > Set PG/PC Interface).
2. Select "CP5512 (PROFIBUS)" as the interface parameterization used.
3. Open the Properties dialog and set "Transmission rate" to the value configured in HW Config. Common values: 1.5 Mbps, 3 Mbps, 6 Mbps, 12 Mbps. If the project uses "Auto" in the master, pick 1.5 Mbps as a stable starting point and only step up after the segment has been validated with an oscilloscope or BT200 tester.
4. Confirm the highest station address and the bus profile match the master. Click OK.

5.3 Set the Correct Access Point

In the "Set PG/PC Interface" dialog, switch to the Access Point of the Application tab and set the access point S7ONLINE to the CP5512 entry. The same setting is required for STARTER's Options > Settings > Access point. Some installations also require setting STARTER to the same CP5512 entry; the two are independent and both must point to the same physical interface.

5.4 Use the "Search for Online Drives" Fallback

If "Accessible nodes" still does not work, use the more aggressive scanner inside STARTER:

1. Launch STARTER and open the existing offline project (if it matches the machine).
2. Select Project > Connect to target system > Search for online drives.
3. Choose the CP5512 entry and click "Start search".
4. STARTER will broadcast on all supported baud rates and report every drive it can read. If the CU320 appears with a yellow warning triangle, the firmware/STARTER version mismatch is confirmed. If it does not appear at all, the issue is at the physical layer or the CF card.

6. Step-by-Step Recovery: CU320 CompactFlash Card

If the PLC can no longer see the CU320 after a power cycle, and the green RDY LED is on but the DP slave is absent, treat the CF card as suspect. The recovery has two levels: clean the user data, then reflash the firmware if the user-data clean does not bring the slave back.

6.1 Clean the User Data on the CF Card

1. Power the drive line-up down completely. Wait for the RDY LED on the CU320 to extinguish.
2. Eject the CF card from the CU320 slot and mount it in a USB or PCMCIA card reader on the PG.
3. Open the card in Windows Explorer. Confirm that the following paths exist:
   • \USER\SINAMICS\DATA — contains parameter backup files named PS00xxxx.ACX
   • \USER\SINAMICS\DATA\RT_DIR\CFG\TOCON — contains topology files named TO00xxxx
4. Delete the PS00xxxx.ACX files in \USER\SINAMICS\DATA and the TO00xxxx files in \USER\SINAMICS\DATA\RT_DIR\CFG\TOCON. Do not delete the KEYS.TXT license file, the \FIRMWARE directory, or the \OEM directory.
5. Reinsert the CF card into the CU320 and power the cabinet back up.
6. The CU320 will boot with factory defaults. The PROFIBUS DP slave should reappear in the master at its configured address. Drive commissioning must be repeated: motor data, encoder data, BICO interconnections, and the PROFIBUS telegram configuration.

6.2 Reinstall the Firmware

If cleaning the user data does not bring the DP slave back, the firmware image itself is suspect. Reinstall from the original firmware package.

1. Locate the firmware matching the V2.6 release used on the machine. The firmware is delivered as a ZIP file containing a .bin or .upd file and a tool such as the STARTER firmware loader or the Web server's "Update" page.
2. Insert the CF card into the PG's card reader.
3. Copy the firmware files to the \FIRMWARE directory on the card, preserving the directory structure expected by the loader.
4. Reinsert the card in the CU320 and power on.
5. The CU320 will detect the firmware file on the card and update automatically. RDY and DP LEDs will flash in a programmed pattern; do not interrupt power during this sequence.
6. After the update completes, power-cycle again. Confirm the firmware version in the Web server or STARTER's online view.

7. Hardware Reconfiguration in the PLC Project

After the drive side is recovered, return to the STEP 7 project and re-establish the cyclic data exchange. For a CU320 with V2.6 firmware, the standard Siemens telegrams 1 through 4 are usually the right choice:

Sample STEP 7 hardware configuration snippet for the DP slave properties, expressed as it would appear in the HW Config dialog:

PROFIBUS address : 5
Diagnostic address: 2046 (configurable)
Assigned slot     : 0 (control unit, CU320)
Telegram          : Standard Telegram 1
I address range   : PIW 256..PIW 259  (4 bytes = 2 PZD words)
Q address range   : PQW 256..PQW 259  (4 bytes = 2 PZD words)
Consistency       : Total length (use SFC14/SFC15 in the user program)

For DP-V0 slaves such as the CU320 with V2.6 firmware, use the SFC14 "DPRD_DAT" and SFC15 "DPWR_DAT" system function blocks to read and write the process data with the configured consistency:

// OB1 - read 4 bytes of actual values from the drive
CALL  "DPRD_DAT"
      LADDR  := W#16#100     // I address base 256 = 0x100
      RET_VAL:= MW100        // error code, 0 = no error
      RECORD := P#M 110.0 BYTE 4

// OB1 - write 4 bytes of setpoint to the drive
CALL  "DPWR_DAT"
      LADDR  := W#16#100     // Q address base 256 = 0x100
      RECORD := P#M 120.0 BYTE 4
      RET_VAL:= MW102        // error code, 0 = no error

The LADDR values use the byte address divided by 8 plus an offset; for the address range starting at 256, the LADDR is 256/8 = 32 = W#16#20 in some S7 conventions, but the example above uses the higher-level mnemonic in which the I/Q base address is the LADDR. Refer to the CPU 314C-2DP manual for the exact conversion if your build uses a different convention.

8. Verification

Run this checklist after the recovery to confirm the system is back to a known-good state.

9. Prevention and Commissioning Recommendations

1. Document the PROFIBUS baud rate per segment. The bus profile, baud rate, and highest station address belong in the cabinet documentation and on the cover of the electrical drawing set. A later service visit will thank you.
2. Keep the access point consistent. PG/PC interface and access point should be set once at the start of a project and locked. Many "mystery" offline failures are caused by an engineer switching the access point to an Ethernet adapter that does not exist on the PG.
3. Back up the CF card after every commissioning step. Use the Commissioning > CF card > Save to PC function in STARTER at every milestone. The backup is the difference between a 15-minute recovery and a full recommissioning.
4. Avoid mixed firmware and STARTER versions on the same PG. If the project must support both V2.6 and V4.x or V5.x CU320 firmware, keep the matching STARTER installations in separate virtual machines. The "missing device identification" symptom is almost always version drift.
5. Use a stabilized UPS for the cabinet. A CF card corruption is the most common firmware side effect of an uncontrolled power-down during a write. The cost of a small 1 kVA UPS is trivial compared to a service visit.

10. Troubleshooting Matrix

11. FAQ