SINAMICS S120: Fixing Lost Homing After DRIVE-CLiQ Unplug

David Krause7 min read
Motion ControlSiemensTroubleshooting
Licensed PE Working through this on a live machine? A Maine-licensed engineer can take it from here — included with IMD hardware, by the hour for everything else. Book an engineer

Problem Details

A SINAMICS S120 axis driving a linear car with an absolute encoder holds position correctly across a full DC-link power cycle, but loses its referenced position whenever the DRIVE-CLiQ cable is pulled from the SMC30 Sensor Module and plugged back in.

Component Order number (family)
Control Unit CU320 (6SL3040-0MA00-0Axx)
Smart Line Module 6SL3130-6TE25-5Axx
Single Motor Module 6SL3120-1TE31-3Axx
Sensor Module Cabinet SMC30 6SL3055-0AA00-5Cxx

Event sequence captured when the DRIVE-CLiQ cable is removed during operation (motor at standstill throughout):

Code Type Text / detail
F7453 Fault Position actual value preprocessing error
F31885 Fault Encoder 1 DRIVE-CLiQ (CU): Cyclic data transfer error — Component number 4, Cause 33
A1507 Alarm BICO: Interconnections to inactive objects present (0)
A1315 Alarm Drive object not ready for operation
A1482 Alarm Topology: Comparison Sensor Module missing in the actual topology (Component number 4)
A7441 Alarm Save the position offset of the absolute encoder adjustment (appears after reconnect + acknowledge)

After acknowledging all faults, the actual position value is wrong and r2684.11 (homed / reference point set) reads No. Recovery requires a fresh reference-point set followed by RAM-to-ROM save (p0977 = 1), because p2525 (offset of the encoder adjustment) no longer matches the encoder raw value.

Key distinction: a pure encoder communication fault (F31885 / F7453 alone) does not clear the homed status on the Basic Positioner. What clears it here is the component leaving the topology — the alarm to watch is A1482, not F31885.

Root Cause

Pulling the DRIVE-CLiQ cable does two different things at once:

  1. Cyclic data loss → F31885 and the resulting position preprocessing fault F7453. On its own this is a recoverable communication fault; the drive keeps its adjustment data and the homed bit.
  2. Component drop-out of the actual topology → A1482 reports the Sensor Module (component 4) missing versus the target topology. The drive object referencing that encoder is no longer operable (A1315), and BICO interconnections point at an inactive object (A1507). At this point the encoder data set is treated as invalid/re-initialised, the absolute adjustment (p2525) reference is discarded and r2684.11 falls to 0.

When the module reappears, the encoder is re-initialised as a "new" encoder instance. The alarm A7441 is the drive telling you exactly this: the absolute encoder adjustment offset must be re-established and stored non-volatile. A RAM-to-ROM save performed before the disconnect does not help, because the offset that was saved is no longer considered valid against the freshly initialised encoder component.

This also explains why a DC supply disconnect behaves correctly: on a normal power cycle the topology is complete on start-up, the absolute encoder raw value is read, p2525 is applied, and the axis comes up homed.

Other causes to rule out first

  • Encoder or Sensor Module physically replaced — a new absolute encoder always requires a new absolute adjustment.
  • Axis moved while the drive was powered down beyond the tolerance window monitored by p2722, which produces F07449. Check the fault buffer for F07449 before blaming the cable.
  • Encoder configuration changed (resolution, multi-turn range, gear ratio) — any change invalidates the stored offset.

Solution: Deactivate the Encoder Component Before Unplugging

The behaviour is reproducible and the working countermeasure is to take the encoder component out of the topology in a controlled way using the encoder activation/deactivation parameter p0145, instead of yanking a live DRIVE-CLiQ connection.

With the encoder interface deactivated via p0145 before the cable is removed, and reactivated after it is reconnected, the position and homed status survive the operation.

Maintenance procedure

  1. Stop the axis and remove the enables (OFF1/OFF2). Confirm zero speed.
  2. Deactivate the encoder interface for the affected encoder via p0145.
  3. Verify the drive no longer signals the encoder as an active topology member (no A1482 pending, no F31885 storm).
  4. Unplug the DRIVE-CLiQ cable at the SMC30 and perform the maintenance work.
  5. Reconnect the DRIVE-CLiQ cable. Wait for the Sensor Module LED to indicate normal operation.
  6. Reactivate the encoder interface via p0145.
  7. Acknowledge any residual alarms and check r2684.11 = 1 and the actual position value.

Making it usable by the maintenance team

Direct parameter access is unacceptable for maintenance staff. Wrap the sequence so no one edits parameters by hand:

  • HMI button: write the deactivate/activate value to p0145 over PROFIBUS/PROFINET acyclic parameter access (Read/Write Record) from the PLC, triggered by a password-protected "Encoder Service Mode" button.
  • Digital input: interconnect a CU320 digital input to the control word bit driving the deactivation routine, so a keyswitch in the cabinet arms service mode.
  • Interlock the routine against drive enable: never allow deactivation while pulses are enabled or the axis is commanded.
  • On leaving service mode, automatically issue p0977 = 1 (RAM to ROM) only after r2684.11 confirms the homed state is still valid.
Warning: deactivating an encoder interface removes the position feedback from the closed loop. Only execute with pulses inhibited and the axis mechanically safe (brake applied / vertical axis secured). Verify the exact permissible values and the effect of p0145 for your CU firmware in the SINAMICS S120/S150 List Manual before deploying the routine.

If Re-Homing Is Unavoidable

Where field wiring will still be disturbed without the controlled sequence, build a fast, repeatable recovery:

  1. Acknowledge F7453 / F31885 and confirm the topology is complete (A1482 gone, component 4 present).
  2. Move the axis to a physical, repeatable datum (hard stop, proximity datum plate, or laser-verified mark).
  3. Execute the absolute encoder adjustment / set reference point at that datum.
  4. Immediately save non-volatile: p0977 = 1. The A7441 alarm clears only after the offset is stored.
  5. Record the resulting p2525 value in the machine documentation so a technician can compare it after future service events — a changed p2525 for the same physical datum indicates the encoder or its mounting changed.

Verification

Check Where Expected
Homed / reference point set r2684.11 1 (Yes) after reconnect and acknowledge
Absolute adjustment offset p2525 Unchanged versus documented value
Offset save pending A7441 Not present
Topology match A1482 Not present; Sensor Module component 4 in actual topology
Drive object ready A1315 Not present
BICO to inactive objects A1507 Not present
Standstill position drift Actual position value Matches pre-disconnect value within encoder resolution

Functional test: park the axis at a marked mechanical position, record the actual position value, run the controlled service sequence, and confirm the value returns identically after reactivation. Repeat once with a full 24 V / DC-link power cycle to confirm both recovery paths.

Controller-Side Behaviour for Comparison

If the axis is handled as a Technology Object on an S7-1500 rather than by the on-board Basic Positioner, the homed status of an absolute encoder is reset only in these cases:

  • CPU replacement
  • Change of the encoder configuration
  • Reset of the CPU to factory settings
  • Download of a different project to the controller
  • Installation of a new absolute encoder (requires a new absolute adjustment)

A CPU memory reset or a project upgrade does not require a new absolute encoder adjustment. Knowing this split matters during commissioning: if the homed status is lost after a PLC download but the drive-side p2525 is intact, the problem is in the TO configuration, not in the DRIVE-CLiQ chain.

Preventive Measures

  • Strain-relieve and lock every DRIVE-CLiQ connector at the SMC30; the latch must click. Accidental partial unplugging produces the same component drop-out.
  • Keep DRIVE-CLiQ routing separate from motor cables and honour the cable-length limits for the type used.
  • Log r2684.11 in the PLC and alarm the operator immediately if it drops to 0, so a lost home is detected before a positioning move is commanded.
  • Block automatic positioning commands in the PLC while r2684.11 = 0.
  • Back up the drive parameter set (including p2525) after every successful adjustment.

Why does my SINAMICS S120 absolute encoder lose homing when I unplug the DRIVE-CLiQ cable?

Because the Sensor Module leaves the actual topology (alarm A1482), the encoder is re-initialised on reconnect and the absolute adjustment offset in p2525 is treated as invalid, clearing r2684.11. The communication faults F31885 and F7453 alone would not reset the homed status.

What does alarm A7441 mean on SINAMICS S120?

A7441 "Save the position offset of the absolute encoder adjustment" means the adjustment offset p2525 is only in RAM and must be stored non-volatile. Perform the absolute encoder adjustment, then execute p0977 = 1 (RAM to ROM); the alarm clears once the value is stored.

How do I remove an SMC30 for service without losing the referenced position?

Inhibit the pulses, deactivate the encoder interface with p0145 before unplugging the DRIVE-CLiQ cable, do the work, reconnect, then reactivate the encoder with p0145 and verify r2684.11 = 1. Verify the permitted p0145 values for your firmware in the SINAMICS S120/S150 List Manual.

Which parameter shows whether the axis is still homed?

r2684.11 (reference point set / homed) on the Basic Positioner. Map it into the PLC and block automatic positioning whenever it reads 0.

The axis reports F07449 after a power-down — is that the same problem?

No. F07449 relates to the absolute-position monitoring configured with p2722 and indicates the axis was moved beyond the tolerated window while the drive was off. Check the fault buffer for F07449 before attributing lost position to the DRIVE-CLiQ cable.

Back to blog