S7-315-2PN/DP Profibus Troubleshooting: G120 CU240E-2DP Dropout

1. Problem Overview

A SIMATIC S7-315-2PN/DP CPU is configured as a Profibus DP class 1 master and exchanges cyclic I/O with seven SINAMICS G120 inverters fitted with CU240E-2DP control units. The drives are wired in a single Profibus segment, daisy-chained with Profibus DP connectors. After commissioning, only the first three drives (Profibus addresses 1, 2, 3) come online. Drives 4 through 7 remain in a bus-fault state with the drive LEDs indicating no Profibus handshake. Pulling the Profibus connector on drive 4 makes drive 5 appear but causes drive 6 to drop out, confirming that nodes past the third drive are not seeing a clean differential signal.

The pattern is characteristic of one of three faults, in order of probability for this exact symptom set:

1. Missing or incorrect bus termination at one end of the segment, causing signal reflection and reduced receive voltage at slaves beyond roughly the third node.
2. Baud rate / segment length mismatch where the configured transmission speed forces a maximum trunk length shorter than the actual physical segment.
3. Shield discontinuity or split-pair routing in the cable between drive 3 and drive 4, producing attenuation or common-mode noise on the RS-485 lines.

Individual point-to-point communication with each drive works because the segment reduces to a single-node bus when only one slave is attached, masking the termination and reflection faults that appear as the segment length grows. The remainder of this article walks through each suspect systematically and provides the verification steps to prove the root cause.

2. Affected System Components and Versions

The components involved in this fault pattern are listed below. The exact catalog numbers should be confirmed against the hardware configuration in STEP 7 (HW Config) or TIA Portal (Device Configuration) before ordering spares or replacement cabling.

3. Profibus DP Topology Fundamentals

Profibus DP is an RS-485 based token-passing protocol. The S7-315-2PN/DP cycles as a class 1 master, polls each configured slave, and expects a deterministic reply within a watchdog window. The physical layer is a single twisted pair with overall shield, terminated at both ends with a defined resistor network, and supports multiple speeds from 9.6 kbit/s up to 12 Mbit/s. Three physical limits define every Profibus DP segment:

• Maximum of 32 nodes per segment (one node per address). The S7-315 master counts as one node, leaving 31 slave addresses free. Seven drives fit easily.
• Maximum of 9 repeaters between any two nodes across the entire network.
• Maximum trunk length per segment as a function of baud rate (see Section 5).

Critically, only the two physical ends of the segment may have termination active. Any active terminator in the middle of the daisy chain creates a low-impedance stub that reflects energy and pulls the differential voltage below the 1.0 V peak-to-peak noise-free threshold that RS-485 transceivers need to cleanly distinguish marks from spaces. This is the most common cause of "first three slaves work, the rest don't" symptoms when the cable is otherwise intact.

The signal-quality drop-off with active middle termination is cumulative: each correctly terminated end node absorbs reflections, but a mid-segment terminator reflects incoming edges back into the trunk. With seven drives at typical 1.5 Mbit/s and modest cable lengths, the reflected energy from a single mid-segment terminator is sufficient to corrupt bit timing at slave 4 and beyond while slaves 1-3 still receive clean waveforms.

4. Root Cause Analysis

The four causes that reproduce this exact pattern are ranked by frequency of field occurrence for 7-node segments:

The original poster's symptom - drive 5 appearing when drive 4 is removed - is the textbook signature of cause 1 (a mid-segment terminator). Each Profibus DP connector has a switch with three positions: OFF (no termination, signal passes through), ON (termination active, no signal pass-through), and on some variants a middle position. The convention is that the two physical ends of the segment sit at ON, and every node in between sits at OFF. A single mid-segment connector inadvertently switched to ON will leave the drive visible to the master (it still responds to the first poll request that reaches it before the reflection destroys bit timing) but renders downstream slaves invisible.

5. Cable, Connector, and Transmission Specifications

The Profibus DP cable is a single twisted pair (red = B, green = A) inside an overall foil and braid shield. Specific electrical parameters that affect segment length are:

Maximum segment length by baud rate (per EN 50170 / IEC 61158):

For seven G120 drives in a single cabinet or a row of cabinets, total segment length rarely exceeds 50 m, so the 1.5 Mbit/s default will work if the cable meets spec. If the site uses a 12 Mbit/s configuration across the same physical span and is past the 100 m limit, the symptom appears identical to a mid-segment termination fault, with slaves 1-3 surviving on edge cases and 4-7 dropping out.

6. Termination Rules and Signal Reflection

Each Profibus DP segment must be terminated at both physical ends with the following resistor network, drawn between the A (green) and B (red) cores and the 5 V reference supplied by the master or the active terminator:

• 220 ohm between A and B (the differential termination)
• 390 ohm from B to +5V (the B-side pull-up)
• 390 ohm from A to GND (the A-side pull-down)

These three resistors are integrated into the 6ES7972-0BA12-0XA0 and similar Siemens connectors and selected by the slide switch. The three valid switch states are:

To verify termination without disconnecting any cable, an oscilloscope across the A and B cores at the far end of the segment will show:

• Clean square edges with amplitude 4.0-5.0 Vpp (driven high and low by the master during idle) when both ends are terminated and no traffic is on the bus.
• Smaller amplitude or ringing when only one end is terminated.
• Rounded edges and reduced amplitude (typically 1.5-2.5 Vpp) when a middle node is mis-terminated, with ringing amplitude growing toward the far end.

7. Step-by-Step Diagnostic Procedure

1. Inventory the segment. Identify the physical order of nodes: PLC (typically Profibus address 2 with the OP/TP on address 1) -> drive 1 -> drive 2 -> drive 3 -> drive 4 -> drive 5 -> drive 6 -> drive 7. Confirm the chain order matches the addresses (i.e., the node with address 3 is the third physical node from the master). A mis-wired address sequence produces intermittent drops that look like a termination fault.
2. Walk each connector and read the termination switch. The first Profibus connector at the master (X2 port) and the last connector at the last drive should be ON. Every connector between must be OFF. Photograph or mark each position to document state.
3. Power down and reseat every connector. Pin 6 (VP, +5V termination power) must be present at every node so the on-board resistors can pull up. A connector with a bent pin 6 will leave that node's termination resistors underpowered and the bus partially unterminated at that point.
4. Measure the segment resistance from A to B with all drives powered off. Disconnect both ends of the segment from the master. A correctly terminated segment reads 110 ohm (220 ohm || 220 ohm at the two ends). Anything significantly below 100 ohm indicates a mid-segment terminator or a cable short. Anything significantly above 130 ohm indicates an open termination.
5. Measure shield continuity end to end. With the segment disconnected, the shield should read less than 1 ohm from one end to the other, and the shield should be bonded to ground at one point only (typically at the cabinet entry or at the master end).
6. Capture Profibus traffic with an analyzer. Use a Softing PB-T3, Inpro, or Siemens BT200 hardware tester on the segment. Look for: repeated token retries, slave "not reached" diagnostics on addresses 4-7, and an excessive gap-time histogram. Save the trace for the manufacturer's support case.
7. Check baud rate in HW Config. Right-click the DP master interface in STEP 7 or TIA Portal and inspect the Profibus properties. Confirm 1.5 Mbit/s is the default; if the project was changed to 12 Mbit/s, drop it to 1.5 Mbit/s or 500 kbit/s and re-download to the CPU.
8. Verify drive address uniqueness. On each CU240E-2DP, read parameter r2050 or use parameter P2040 (Profibus address) to confirm unique addresses 1 through 7. A duplicate address forces one of the two slaves off the bus.

8. SINAMICS G120 CU240E-2DP Parameter Verification

The CU240E-2DP parameter set required for Profibus DP communication is small but easy to overlook in a commissioning rush. Verify each value from the BOP-2 panel, IOP, or via STARTER / Startdrive connected to the drive over the same Profibus segment using routing.

9. S7-315-2PN/DP Profibus Master Configuration

The CPU 315-2PN/DP has two interfaces: X1 is Profinet (industrial Ethernet) and X2 is Profibus DP. This fault is always on the X2 (DP) interface; the X1 PN port is not involved. Verify the following in STEP 7 (HW Config) or TIA Portal (Device Configuration):

• The CPU is configured as a DP master on X2 with the correct Profibus address (typically 2). If the address collides with a drive, that drive will not appear.
• The bus profile is "DP" not "Standard" or "Universal". Profile "DP" sets the master to Profibus DP class 1 cyclic polling; "Standard" leaves it as FMS-capable which most drives do not support.
• The baud rate is 1.5 Mbit/s (default for drives).
• The configured slave list contains seven G120 entries at addresses 1 through 7, with the matching GSD (SIEM8175.GSD for current firmware).
• The I/O mapping for each slave matches the drive's PROFIdrive telegram. Telegram 1 gives 4 words PZD (2 control / 2 status); Telegram 20 gives 6 words; Telegram 350/352/353 are Siemens-specific mappings for SINAMICS with extended status word.
• The DP diagnostics address (typically the input start address of the diagnostic area for each slave) is reserved in the process image and not used by other I/O.

A mismatch between the telegram configured in HW Config and the telegram selected on the drive (P0922) does not produce the dropout pattern described. It produces a per-slave fault F01910 ("PROFIdrive telegram configuration error") with the affected drive still appearing in the diagnostic buffer. Because drives 1-3 are coming online cleanly, the telegram configuration is unlikely to be the fault.

The S7-300 organization blocks relevant to Profibus diagnostics are OB82 (diagnostic interrupt), OB86 (rack / DP slave failure), OB100 (restart), and OB122 (I/O access error). Add an empty OB82 and OB86 in the project so the CPU does not go into STOP on the first slave dropout. Then read the diagnostic buffer (online -> PLC -> Module Information -> Diagnostic Buffer) for "Slave 4 not found" or "Diagnostic interrupt from slave 4" messages that pinpoint which slave is failing first.

10. Shielding, Grounding, and EMC

The original poster reports that the cable shield is grounded at all nodes. This is the second most common mistake behind termination, and the two together explain almost every intermittent Profibus fault. The Profibus guideline is ground the shield at both ends of the segment but at each node use a low-impedance bond to the local cabinet ground, and ensure there is no ground loop path through the shield that carries load current. If the shield is grounded at all nodes and the cabinet grounds are not equipotential (which is common across long cable trays that span multiple cabinets), 50/60 Hz ground potential differences drive substantial current through the shield. That current couples into the data pair and adds common-mode noise that the receiver's common-mode rejection can no longer cancel.

Correct installation for a single-segment Profibus DP network:

1. Use Profibus DP cable with both foil and braid shield (not single-shield types).
2. At every Profibus connector, clamp the cable shield 360 degrees around the connector backshell using the integrated strain relief. The shield must continue through the connector, not terminate at the connector body alone.
3. Bond the shield to protective earth (PE) at both physical ends of the segment. Mid-segment nodes may either pass the shield straight through to the next connector or locally bond it to the cabinet ground bar - both are acceptable if the local ground is equipotential.
4. If equipotential bonding cannot be guaranteed (long inter-cabinet runs, separate buildings), install a Profibus fiber-optic repeater (e.g., 6GK1500-0FC10 or 6GK1500-3AB10 for OLMs) to break the electrical continuity.
5. Keep Profibus cable at least 200 mm from VFD output power cables (motor leads) to avoid capacitive coupling. Cross at right angles if crossing is unavoidable.
6. Avoid routing Profibus and 24 VDC control in the same bundle; the 24 V return currents can couple noise into the data pair.

11. Implementing the Fix

Based on the symptom ("first three slaves work, the rest drop out"), the order of operations is:

1. Check termination switches first. Walk the segment and set switches to ON at the master and the last drive, OFF at every node in between. Power-cycle the segment and observe the PLC diagnostic buffer. If drives 4-7 now appear, the fault was a mid-segment terminator.
2. If termination was correct, reduce baud rate. Drop from 12 Mbit/s to 1.5 Mbit/s in HW Config, download to the CPU, and observe. If drives 4-7 now appear, the cable is too long or too lossy for the higher rate; 1.5 Mbit/s at < 100 m is the recommended steady state.
3. If baud rate was correct, replace connector at the boundary node. Replace the Profibus connector at the third drive (the last slave that was working) - this is the connector feeding the failing chain. Inspect for bent pins (especially pin 6, VP), broken shield clamps, or cracked housing.
4. If the connector was good, replace the cable between drive 3 and drive 4. This is the most likely cable damage point if the cables are routed through a cable tray that has been disturbed since commissioning. Use a known-good Profibus DP cable with both shields intact.
5. If the cable was good, verify drive-by-drive with a stub. Disconnect all drives except drive 1, confirm drive 1 comes online. Then add drive 2, confirm, and so on. The first drive that fails to come online is the failing link. Replace that connector and the cable to its upstream neighbor.
6. If individual stub testing passes, reconnect the full segment and rerun with a Profibus analyzer logging token retries and gap times.

12. Verification and Acceptance Test

After applying the fix, perform the following to confirm a permanent resolution:

1. All seven drives come online. Read r2050 on each drive to confirm the expected address. Read r2043 (bus status) - all should report "Operational".
2. Diagnostic buffer clean. In STEP 7 / TIA Portal, read the CPU diagnostic buffer and confirm no OB86 (slave failure) events in the last hour of operation.
3. Cyclic I/O exchange. Force a control word 0x047E (OFF1=0, OFF2=1, OFF3=1, inhibit operation, pulse enable) to each drive and read back status word 0x0631 (ready to switch on, switched on, operation enabled, no fault). All seven drives must echo the expected status.
4. Watchdog test. Use parameter P2041 (display only) or STARTER / Startdrive trace to log the bus error counter over a 24-hour period. A clean installation reports zero retries; an installation with marginal signal quality reports retries in the single digits.
5. Mechanical / EMC test. With the system running, gently tap each Profibus connector and tug each cable section. No drive should drop offline. Any drive that drops during this test indicates a marginal connector or cable that needs replacement before commissioning sign-off.
6. Long-run stability. Leave the system in operation for at least 24 hours and recheck r2043 on every drive. The Profibus error counters should remain at zero.

13. Common Adjacent Faults to Rule Out

After the termination and cable faults above have been eliminated, these adjacent causes occasionally appear in the field and are worth a one-line check:

• PG (programming device) cable attached to a middle connector with PG port. Some PG-port connectors feed the bus back to a programming port. If a laptop is left connected with its Profibus interface enabled, it can act as a second class 1 master and steal the token. Unplug all PG cables from middle connectors before final commissioning.
• CM1243-5 / DP/DP coupler left configured. If a Profibus coupler or another CPU is physically attached to the segment but not in the master's slave list, it still loads the bus. Remove or address it explicitly.
• Drive firmware mismatch with the GSD. A G120 with firmware 5.x requires the corresponding 5.x GSD. Using a 4.x GSD on a 5.x drive works for basic cyclic I/O but some extended diagnostic words will not map correctly, occasionally producing false-positive communication faults in the PLC.
• Star topology. Profibus DP is a bus topology. If the drives have been wired in a star from a junction box, each branch looks like a stub. Stubs longer than 1 m at high baud rates cause reflections. Use a repeater (6GK1500-0AB10) if a star topology is unavoidable.
• Power quality on the drive. A G120 CU240E-2DP with weak 24 V supply (less than 21 V under load) will intermittently reset its Profibus interface. Measure the 24 V at the drive terminals during the fault to rule out power supply sag.

14. Related Siemens Documentation

For deeper reference, the official Siemens documentation set covers each component in this article:

• S7-300 CPU 315-2 PN/DP manual: SIMATIC S7-300 CPU 31xC and CPU 31x Operating Instructions
• SINAMICS G120 CU240E-2DP operating instructions: SINAMICS G120 Control Units CU240B/E-2 Operating Instructions
• SINAMICS G120 parameter list: SINAMICS G120 List Manual
• SINAMICS G120 GSD files download: SINAMICS G120 PROFIBUS GSD files
• PROFIBUS DP installation guideline: PROFIBUS Installation Guideline for Cable Routing (PI order number 8.032)
• SIMATIC Profibus network diagnostics with FB125: SIMATIC S7-300/400 Profibus Diagnostics with FB125

15. FAQ