Siemens SINAMICS F31115 F7452 F7453 Choke Encoder Fault Fix

1. Problem Scope and System Topology

The reported fault cluster — F31115, F7452, F7453, and F7490 — appears on a SINAMICS S120 drive controlling a choke / servo motor / encoder assembly for pressure control. The choke position is set by an absolute encoder (EnDat 2.1) and the drive is operated in EPOS (Basic Positioner) mode. The configured position range is 0–1,000,000 LU, which corresponds to 0%–100% choke opening.

The drive trips "randomly" both while the motor is moving toward a new setpoint and when the motor is stationary. On inspection the trips are not random; they correlate with the choke position crossing a specific point in the encoder's bipolar signal range. Two chokes with identical drive/encoder configuration fail at the same kind of position, which initially suggested a duplicate hardware failure but is now understood to be a position-related signal-quality phenomenon rather than two simultaneous encoder failures.

The relevant system topology elements are:

• Drive: SINAMICS S120 (CU320-2 or compatible control unit) with EPOS option active.
• Motor: synchronous servo motor with integrated absolute encoder (EnDat 2.1, singleturn or multiturn).
• Encoder interface: DRIVE-CLiQ or SMC/SME module feeding encoder channel 1.
• Position resolution: 1,000,000 LU over the full choke stroke.
• Closed-loop chain: position controller → speed controller → current controller → motor → encoder feedback.

2. Fault Cluster Decoded

Each fault has a different origin within the drive firmware, but in this application they are all symptoms of the same underlying encoder-signal-quality issue. The exact SINAMICS list-manual descriptions are summarised below.

In short: F31115 is the root, F7453 is a side-effect, F7452 may or may not appear, and F7490 is the EPOS layer detecting that the drive dropped out underneath it. Treat the F7490/F7452 chatter as a symptom log, not as independent problems to solve.

3. Why the Fault "Looks Random" — Bipolar Encoder Position

The Endat encoder fitted to these motors reports its absolute position in two overlapping ranges. The output therefore shows a bipolar characteristic: a slow linear segment followed by a slope reversal, repeated every 180° of mechanical rotation, which corresponds to one SIN period. Plotting encoder actual position value (r0483) against the mechanical stroke produces a saw-tooth or triangular shape rather than a straight line.

Tripping is observed at the intersecting point of the two ranges — the position where the encoder transitions from the positive half of the SIN/COS signal to the negative half (or vice versa). At that transition the differential signal amplitude is at its lowest in absolute terms, the SIN/COS sub-signals are closest to zero crossing, and any small loss of signal margin (cable, connector, EMC, controller gain pushing the controller into saturation near setpoint) tips the Sensor Module below the F31115 detection threshold.

This is also why the fault is observed at the same setpoint on the same motor and at the same setpoint on the second motor of identical construction — it is not a statistical hardware failure, it is a position-driven signal-to-noise problem.

4. Configuration vs. Hardware — Decision Path

Field experience with the SINAMICS platform shows that the F31115 / F7453 pair on EnDat 2.1 encoders resolves into one of three causes, in order of likelihood:

1. Signal-quality degradation on the encoder cable path. Damaged cable, loose DRIVE-CLiQ connector, missing shield bond, mixed encoder cable types, or cable length approaching the maximum.
2. Controller tuning too aggressive for the load. Kp of speed controller (p1460) or position controller (p2538) drives the current loop to saturation near setpoint. The current controller's transient then modulates the supply rail and dips the encoder sub-signal momentarily below threshold exactly at the SIN/COS zero crossing.
3. Genuine encoder failure. LED/photodiode degradation, contamination, mechanical shock to the encoder disk, or bearing wear that increases eccentricity. This typically shows as a uniformly low SIN/COS amplitude across the rotation, not as a localised dip.

Do not change the following-error window (p2546), the position monitoring time (p2545), or the standstill window before verifying the SIN/COS amplitude with the SINAMICS trace. Raising the limit only delays the fault, it does not fix the cause. The Siemens support guidance and the discussion in this case both confirm that tightening or loosening the LR parameters "makes no sense" if the encoder signal is the underlying problem.

5. Diagnostic Procedure — Read the SIN/COS Tracks Live

SINAMICS firmware exposes the raw SIN/COS sub-signals for the encoder. This is the single most useful diagnostic step and should be completed before any parameter is changed.

5.1 Parameters to read

• p0496 — encoder 1 actual incremental signal amplitude (Track A, peak amplitude in mV).
• p0497 — encoder 1 actual incremental signal amplitude (Track B, peak amplitude in mV).
• p0498 — encoder 1 actual amplitude of the C/D track on a square-wave encoder, otherwise the zero-mark or absolute-track diagnostic.
• r0499 — encoder 1 zero-mark status, also used as a "signal OK" flag.
• p0438, p0439, p0431, p0432, p0433 — encoder configuration (interface, signal level, protocol, sampling rate).
• p4962 / p4963 / r4964 — on firmware 4.7 and later, the raw SIN and COS samples (scaled) can be traced directly. They appear as a 1 Vpp sine/cosine pair on the trace when the encoder is healthy.

Healthy signal levels: nominal 1 Vpp (i.e. ±500 mV around 2.5 V bias). The Sensor Module alarm threshold is normally ~700 mV peak-to-peak. The trip threshold is typically ~500 mV peak-to-peak. The fault values shown in the case (−1624, −6744 in the internal 16-bit signed representation) indicate the raw ADC samples, not millivolts. To convert: the Sensor Module scales 1 Vpp into roughly ±14745 counts on a typical SMC20, so a value of 6744 corresponds to ~228 mV peak amplitude on that channel — clearly below the alarm threshold and the reason the fault is raised.

5.2 Capture a trace

1. Open STARTER or Startdrive and connect to the drive.
2. Open the Trace function and select the trace that is currently recording the position controller variables.
3. Add the following trace signals:
   • r0061 (actual speed, unsmoothed)
   • r0479 (encoder 1 diagnostic, including zero-mark counter)
   • p4962 / p4963 (raw SIN/COS where available)
   • r0483 (position actual value of encoder 1, Gn_POS)
   • r2521 (position actual value, EPOS)
4. Trigger the trace on F31115 itself (fault-driven trigger) and let the drive run through the fault.
5. Save the recording and inspect: the SIN/COS pair will show a clear amplitude dip exactly at the moment the encoder position crosses the bipolar boundary.

5.3 Mechanical-electrical checklist

1. Verify encoder cable length is within the maximum allowed for the Sensor Module (typically 100 m for SMC20 with EnDat 2.1; 50 m for higher sampling rates).
2. Verify shield bonding at the drive end and at the motor end; the shield must be clamped, not pig-tailed.
3. Verify DRIVE-CLiQ connectors are fully seated and the DRIVE-CLiQ cable is not running parallel to power cables for more than 200 mm.
4. Measure DC supply to the encoder (typically 5 V ±5%); droop under load indicates cable resistance or a weak Sensor Module regulator.
5. Verify the encoder cable is the Siemens-recommended type (e.g. 6FX2002-2EN00-….) and not a generic substitute.
6. Verify the motor and encoder are properly grounded to the same equipotential bonding bar.

6. Configuration Fixes (After the Signal is Confirmed Healthy)

If the SIN/COS trace shows clean amplitudes (>800 mV peak) and the fault persists, the next layer is controller tuning. Aggressive speed- and position-controller gains can be addressed without changing hardware.

The 30% reduction is a starting point taken from the Siemens field-recommended practice for EnDat 2.1 encoders with marginal signal amplitude. The correct end value is the largest Kp that does not produce a visible transient in the SIN/COS trace at the bipolar crossover.

7. Firmware and Encoder-Mode Considerations

SINAMICS firmware V4.7.9 and later introduced a fallback for encoders that are physically EnDat 2.1 but where the absolute protocol cannot be reliably decoded. The encoder can be reparameterised as an SSI-only encoder:

• Set p0400.0 (encoder type) to "SSI absolute" if the encoder is SSI-based, or keep "EnDat" if the protocol is genuine EnDat 2.1.
• Set p0425 (encoder protocol) accordingly.
• Verify the resulting SIN/COS amplitude: if the protocol is the cause of the F7453 / F31115 pair, switching to SSI may bring the signal amplitude up to nominal.

Before making this switch, confirm the encoder model is actually SSI-compatible, because the Siemens EnDat protocol carries additional safety and diagnostics channels that an SSI-only configuration will lose.

8. Hardware Verification — When the Signal is Genuinely Low

If the trace in Section 5 shows the SIN/COS amplitude is below 500 mV peak-to-peak across the entire rotation, treat the encoder (or the cable) as the prime suspect.

1. With the motor isolated, measure the encoder cable resistance on each conductor end-to-end. Acceptable values are <1 Ω per conductor for cable lengths up to 50 m.
2. Measure insulation resistance between conductors and shield, >10 MΩ at 500 V.
3. Swap the Sensor Module (SMC20 or SME) with a known-good unit and repeat the trace.
4. If the signal is still low on a known-good cable and module, the encoder itself is degraded. Replace the encoder (and the encoder mounting, if it is a hollow-shaft design — bearing play causes eccentricity that modulates the SIN/COS amplitude with mechanical rotation).
5. On the new encoder, run a one-hour motion cycle across the full stroke. If the trip is not reproduced, the original encoder was the cause.

9. Commissioning Parameters — Recommended Baseline

The following baseline is a defensible starting point for a 1,000,000-LU choke with EnDat 2.1 absolute encoder, EPOS mode, and pressure-control dynamics. Adjust to match the specific motor and load.

10. Verification After the Fix

1. Run a continuous-traverse profile that sweeps the full 0–1,000,000 LU range at 25%, 50%, 75%, and 100% of the rated speed for at least 1,000 cycles each.
2. Trigger the SINAMICS trace on F31115 and confirm that no record is produced.
3. Monitor r2521 (position actual) and r0061 (speed actual) for any transients exceeding the expected acceleration.
4. Capture p0496 and p0497 over a full stroke; both should remain >700 mV peak across the rotation.
5. Capture the encoder position value (r0483) over a full mechanical revolution and confirm the bipolar saw-tooth is a clean, repeatable waveform. A deformed waveform indicates mechanical eccentricity.
6. Run a 24-hour soak test under normal production setpoint cadence before declaring the fix complete.

11. Troubleshooting Matrix

12. Related Faults to Recognise

• F31110 — encoder 1 communication error. Pair of F31115; look at the same cable path.
• F31111 — encoder 1 internal error (ASIC / EnDat protocol error). Often firmware-related.
• F31117 — encoder 1 inversion error / A-B swap. Configuration, not signal-quality.
• F31130 — encoder 1 zero-mark failure. Can co-occur with low SIN/COS amplitude.
• F7450 — LR: standstill monitoring tripped. Setpoint reached but controller saturated.
• F7451 — LR: positioning monitoring tripped. Following error window violated at end of motion.
• F7481 — EPOS: setpoint output error. Look at the EPOS jog/teach-in wiring.

Each of these is documented in the SINAMICS S120/S150 List Manual; treat the list manual as the source of truth for thresholds and parameter numbers, as the firmware revision you are running may have shifted specific values.