SINAMICS G120 EPOS F07490: Positioning Setpoint Not Accepted

Problem Overview: Intermittent EPOS Setpoint Rejection on SINAMICS G120

A SINAMICS G120 drive (CU250S-2 control unit) configured for the Basic Positioner (EPOS) over PROFINET will occasionally ignore a new position setpoint even though the value is being received correctly on the bus. The drive holds the existing position, reports the new setpoint as the active setpoint, and shows no torque, no following error, and no fault. Cycling the safety (OFF1 → OFF2 → ON) or issuing a brief jog pulse from the PLC clears the condition and the axis immediately traverses to the requested target.

This is a field-proven failure mode in installations where the EPOS is operated in MDI (Manual Data Input) mode with continuous PROFINET setpoint transfer. The drive enters a state where the setpoint is acknowledged but the motion command is internally latched in a non-running state. The companion fault that typically appears in the alarm buffer is F07490 — "EPOS: Enable signal withdrawn while traversing." The fault is an effect, not the cause; the root cause is the enable chain collapsing at a specific point in the EPOS state machine.

This article walks through the architecture, the parameter set that must be traced, the diagnostic steps that isolate the failing enable, and the parameter corrections that eliminate the failure. The reference drive configuration is a CU250S-2 PN (order number 6SL3246-0BA22-1FA0) running firmware V4.7 SP10 or later with the Standard EPOS option installed.

Symptom Matrix

EPOS Architecture in the SINAMICS G120 with CU250S-2

The Basic Positioner (EPOS) is a software option for the SINAMICS G120 / G120C / G120P family that requires a CU250S-2 control unit (PN, DP, or no-fieldbus variant) and firmware V4.5 or higher. EPOS adds a positioning control layer on top of the standard speed control loop. The function description, parameter list, and commissioning flowcharts are in the SINAMICS G120 CU250S-2 EPOS Function Manual.

The control structure is layered. From the bus interface inward:

1. PROFINET PZD receive words (STW1, STW2, MDI_MOD, position setpoint, velocity override)
2. PROFIdrive profile state machine (Switching On Inhibited → Ready → Switching On → Operation → Fault)
3. EPOS operating mode selection (Jog, Reference Point Approach, MDI, Traversing Blocks)
4. EPOS setpoint processing (setpoint acceptance, fine interpolation)
5. Position controller (Kp, Tn, pre-control, output limits)
6. Speed controller (below the position controller)
7. Motor / encoder feedback

Each layer has its own enable signal. If any single enable in the chain drops, the position controller clamps its output to zero and the drive decelerates to a standstill. If the EPOS was in the middle of a traversing block or an MDI setpoint, the firmware raises F07490. TIA Portal integration of the G120, including the S7-1500 motion control interface, is documented in Configuring SINAMICS G120.

F07490: "Enable Signal Withdrawn While Traversing" — Fault Anatomy

Fault code F07490 is the standard reaction when the EPOS state machine is in motion and any one of the upstream enables transitions to FALSE. The factory reaction is OFF1 (controlled ramp stop) with optional OFF2 / OFF3 follow-up if the underlying enable was OFF2 or OFF3.

Two points are critical when interpreting F07490:

1. It is a consequence fault, not a root fault. Look in r0945 / r0947 for the fault with an earlier timestamp. Common upstream triggers on the G120 are F07400 (encoder), F08501 (PROFINET sign-of-life), F07900 (motor blocked), and F30002 (DC link overvoltage).
2. It is edge-triggered. If the enable dropped for one PROFINET cycle and came back, F07490 fires once and clears, but the EPOS state machine may remain in "Operation / set" rather than "Operation / traverse." That is the "drive thinks it is at the target" condition the field is reporting.

The most common upstream trigger in this scenario is a one-cycle dropout of the OFF2 (STW1 bit 1) or OFF3 (STW1 bit 2) signal on the PROFINET telegram. Both of these are low-active — a 0 in the bit stops the drive. PLC programs that write the entire STW1 word on every cycle, including any safety logic or HMI overrides, can momentarily drop the bit to 0 without operator intent.

Setpoint Flow: PROFINET PZD to Motor Shaft

EPOS in MDI mode accepts a position setpoint, a velocity override, an acceleration/deceleration override, and a direction of travel. The standard PROFINET telegram for EPOS on a SINAMICS G120 is telegram 110 (selected via p0922 = 110). Telegram 110 provides 10 PZD words and 2 PZD double words for the MDI coupler.

PZD Receive Word Mapping for Telegram 110 (MDI)

To read the actual setpoint being delivered, monitor r2060[5] (PZD receive double word 6+7 = the position setpoint from the bus), r2665 (the CO: EPOS position setpoint after the MDI coupler), and r2521[0] (the actual position value entering the position controller after fine interpolation). If r2665 = new target but r2521[0] = previous target, the MDI setpoint was not accepted and committed.

The setpoint acceptance edge is critical. In MDI mode, the firmware only latches a new position setpoint when:

1. The setpoint acceptance signal (configured by p2650 and interconnected through p2669 / p2670) sees a 0→1 edge, OR
2. The drive is configured for continuous setpoint transfer (where every new PZD value is treated as a target without a discrete acceptance edge).

If the setpoint acceptance edge is missing, the EPOS setpoint register holds the previous value. The PLC sees the new value on the bus, r2665 may or may not update (it depends on the wiring of p2660), and the drive does not commit it. The apparent contradiction between the bus and the position controller is the classic symptom.

Enable Signal Chain: Why a Toggle Recovers the Drive

The momentary "shake" that an operator reports is the only diagnostic step needed to confirm the diagnosis. The EPOS state machine gets unstuck when a control bit toggles. That bit is almost always one of:

The PLC operator who "shakes" the drive by jogging the axis or by re-toggling the safety outputs is incidentally re-asserting one of these edges. The fault is cleared because the state machine is forced to re-enter the traversal state. The most likely candidate is the p2650 (setpoint acceptance) edge or a momentary dropout of STW1.1 / STW1.2 from a poorly-handled bit-write to the control word.

To prove the suspect, read the ZSW1 (status word 1) bits from the drive: bit 0 = Ready to switch on, bit 1 = Ready to operate, bit 2 = Operation enabled, bit 3 = Fault present, bit 4 = Coast stop not active (OFF2 inactive), bit 5 = Quick stop not active (OFF3 inactive), bit 6 = Switching on inhibited. If the drive reports the axis as "Operation enabled" (ZSW1.2 = 1) but r2522 = 0, the EPOS is enabled but the position controller is producing zero output. That isolates the fault to the EPOS layer, not the speed controller layer.

Diagnostic Parameter Reference

Use Startdrive (TIA Portal V15.1 or later), STARTER, or the integrated web server (for CU250S-2 PN with firmware V4.7 SP6+) expert list to read the following parameters live. Group them by function so the trace data is coherent across the failure event.

Bus Layer

EPOS Setpoint Layer

Position Controller Layer

EPOS Operating Mode and Status

Key Configuration Parameters

Root Cause Decision Tree

Start at the bus layer and work down. Each branch is a decision that points to a specific check.

1. Is r2060[5] updating every PROFINET cycle with the new target?
   • No → PROFINET or PLC issue. Check the PLC hardware configuration for the telegram, the IO controller cycle time, and the PZD update on the PLC side. Use the PROFINET diagnostics in r2054 to see if any bus errors are being recorded.
   • Yes → Continue to step 2.
2. Is r2665 (EPOS position setpoint) equal to the new target?
   • No → MDI coupler is not seeing the new value. Check p2660 (signal source for EPOS position setpoint) — should be r2060[5] for the standard mapping. If the BICO is wrong, reconnect it.
   • Yes → Continue to step 3.
3. Is the setpoint acceptance edge toggling 0→1 for every new target?
   • No → The MDI setpoint is not being latched. The drive holds the previous setpoint. This is the most common cause of the reported symptom. Fix: confirm p2650 / p2670 (setpoint acceptance signal source) is wired to the correct STW bit (typically STW2.13) and that the PLC toggles it on every new target.
   • Yes → Continue to step 4.
4. Is r2522 (position controller output) non-zero during the traverse?
   • No → The position controller is producing zero output. Check the following error window (p2546) and the position controller output limit (p2540). If r2526 (following error) is within p2547 but r2522 = 0, the position controller has reached the target internally and the drive considers the move complete even though the position is wrong.
   • Yes → The position controller is driving. Check the speed controller (r1500, r1508) and the torque limit (p1520, p1521).
5. Does the speed controller saturate or does the torque limit engage?
   • Yes → Mechanical issue or undersized drive. Check the load, the gear ratio, and the torque limit settings.
   • No → Check the encoder feedback (r0481, F3xxx faults).

For the F07490 path, additionally read the alarm buffer (r0945, r0947) sorted by time. The fault that occurred immediately before F07490 is the trigger.

Step-by-Step Diagnostic Procedure

1. Connect to the drive with Startdrive (TIA Portal V15.1 or later) or STARTER. Read the firmware version in r0018; EPOS on G120 requires firmware V4.5 or later. The most stable EPOS behaviour on the CU250S-2 PN is firmware V4.7 SP10 and later. Upgrade to the latest service pack if the drive is below this revision.
2. Open the trace function. Record the following parameters on a 4 ms tick for at least 30 seconds: r2060[5], r2665, r2676, r2521[0], r2522, r2526, r2527, r2680.0, ZSW1, ZSW2. A 4 ms tick is the right resolution for typical PROFINET IRT cycle times of 1–4 ms.
3. Trigger the failing condition from the PLC (force the new setpoint).
4. Read the alarm buffer. Note the timestamp of F07490. Look one entry above (chronologically earlier) for the trigger fault. If no earlier fault is present, the OFF1/OFF2/OFF3 enable dropped directly. Check the PLC's write to STW1 — search for a code path that writes 0 to STW1.0/1/2 even momentarily.
5. Examine the trace. If r2676 (setpoint acceptance) did not toggle on the new target, the PLC did not pulse the setpoint acceptance bit. The drive never latched the new value. This is the dominant cause. Fix the PLC code (toggle the bit on every new target, or use continuous transfer mode with a one-cycle pulse from the PLC).
6. If r2676 toggled but r2522 (position controller output) is still 0, the position controller is internally in "target reached." Check r2527.0 (position reached bit) and r2526 (following error). If r2527.0 = 1 immediately on the new target, the positioning window (p2546) is too large relative to the new target delta. Reduce p2546 to a value smaller than the smallest commanded increment.
7. If r2522 is non-zero but r2523 (actual speed) is zero, the speed controller is not acting. Check STW1 bits 0–3 and ZSW1 bits 0–6 to confirm the speed controller is enabled. If ZSW1.6 (Switching on inhibited) is set, the drive is in the inhibit state; check p0840, p0844, p0848, p0852 for the enable sources.
8. Examine the PROFINET cycle time. For EPOS MDI, the recommended send clock is 1–4 ms with an IO cycle of 1–4 ms. Slower cycles (8 ms or more) can cause setpoint acceptance edges to be missed. Confirm the controller's send clock (p2051, p2052) matches the PLC configuration. Mismatched send clocks on the PLC and drive are a common source of intermittent position drops.
9. Replace the safety OFF→ON cycling test with a PLC-controlled one-cycle pulse on STW1.0 (ON/OFF1) instead of the safety input. This isolates whether the recovery is from the safety drop or the EPOS state machine re-entry.
10. Confirm the resolution of the issue by running the application for at least 10× the historical MTBF interval (typically 24–72 hours) with the trace recording active.

Solution Matrix

Verification Procedure

1. After applying the fix, capture a 60-second trace of r2060[5], r2665, r2676, r2522, r2523, r2526, r2527, r2680, ZSW1, ZSW2, r2681 (error word).
2. Confirm: every new target on the PLC side is followed by a 0→1 edge on the setpoint acceptance bit within one PROFINET cycle.
3. Confirm: r2522 ramps from 0 to the commanded velocity (r2666 scaled by p2571) within 50 ms of the acceptance edge.
4. Confirm: r2526 (following error) is bounded by ±p2547 and decays to 0 within the position controller settling time.
5. Confirm: r2527.0 transitions to 1 only when r2526 falls within p2546, and r2527.0 is 0 during the traverse.
6. Confirm: r2680.0 (traversing active) is 1 during the motion and 0 at the end.
7. Confirm: no F07490 in the alarm buffer. If F07490 reappears, re-read the upstream fault and iterate.
8. Run the application under full production load for at least the historical MTBF interval. The default target is 30 days without a single F07490 occurrence.

A useful trace configuration in Startdrive for this verification is:

Trace configuration (Startdrive V15.1+, CU250S-2 PN, FW V4.7 SP10)
  Sample time:           4 ms
  Trace duration:        60 s
  Trigger:               r2522 deviation from 0 > 5 rpm
  Pre-trigger:           20 %
  Signals (in order):
    1. r2060[5]            -- PZD position setpoint (32-bit)
    2. r2665               -- EPOS position setpoint (LU)
    3. r2676               -- MDI setpoint acceptance
    4. r2090               -- STW1 bit decode
    5. r2093               -- STW2 bit decode
    6. r2521[0]            -- Actual position value
    7. r2522               -- Speed setpoint (position controller output)
    8. r2523               -- Speed actual value
    9. r2526               -- Following error
   10. r2527               -- Position controller status
   11. r2680               -- EPOS status
   12. r2053[0]            -- ZSW1 (status word to PLC)
   13. r2053[1]            -- ZSW2 (status word to PLC)
   14. r1526               -- Torque limit active

Related Fault Codes and Edge Cases

Several other EPOS faults share the same enable-chain sensitivity and can produce the same "drive thinks it is there" symptom. Cross-reference them when triaging similar reports.

For PROFINET configuration in TIA Portal, see the Configuring SINAMICS G120 with TIA Portal documentation. For full EPOS function details, parameter list, and commissioning flowcharts, see the SINAMICS G120 CU250S-2 EPOS Function Manual.

FAQ