Troubleshooting Siemens MM440 F0001 and F0021: Root Cause Repair

Overview: F0001 and F0021 on the Siemens MicroMaster 440

The Siemens MICROMASTER 440 (MM440) is a sensorless and closed-loop vector drive widely deployed on conveyors, pumps, fans, extruders, and hoists in the 0.37 kW to 250 kW range. Two of the most disruptive faults an MM440 operator can encounter are F0001 (Overcurrent) and F0021 (Earth Fault), particularly on frame sizes D, E, F, and Gx where the drive uses three phase-current sensors for closed-loop vector control and earth-fault detection.

This reference consolidates the diagnostic logic required to isolate the root cause when these faults:

• Recur immediately after a power cycle (i.e., not a one-time transient).
• Disappear when the vector / CUSA card is removed (suggesting sensing or control electronics, not the IGBT stack).
• Reappear after 20 to 30 minutes of run time (suggesting thermal drift in a sensor, capacitor, or IGBT module).
• Are accompanied by audible motor noise (a high-frequency whine or shrill tone) only when running on the VFD and not on direct-on-line (DOL).

These are the canonical symptoms of a current-transformer (CT) failure, a defective choke, or a developing motor insulation weakness — not a parameter-tuning problem. The faults will not be cleared by a factory reset, and they will not respond to a motor-cable swap once the root cause is internal to the drive or motor winding.

Fault Code Reference

Both faults share a common diagnostic pathway because both rely on the phase-current feedback path (r0069). The F0021 trigger — the vector sum of the three phase currents — is effectively a residual-current measurement. Under normal balanced operation, IA + IB + IC ≈ 0. A non-zero sum greater than 5% of nominal current indicates either a real earth leakage path or a CT mismatch / saturation event.

Current-Sensing Architecture on Frame Size D and Larger

The MM440 uses two different current-sensing topologies depending on frame size:

The 37 kW / 400 V MM440 used in the field case sits in FS D, which is the smallest frame size with full three-CT earth-fault detection. This is why F0021 is meaningful on this drive but may not behave the same way on smaller units.

Root Cause Analysis: Why the Faults Recur After Power Cycle

F0001 and F0021 that survive a power-on reset are not parameter faults. The drive has already proven that the IGBT gate-drive path, the control board, and the BOP are functional — the drive runs, accelerates, and supports load for tens of minutes before tripping. A persistent trip after power cycle points to one of four root-cause families:

1. Insulation breakdown in the motor or cable — either a hard short to ground (resistance < ~50 kΩ phase-to-earth) or a developing partial-discharge / thermal-fuse weakness that increases leakage as the winding heats up. The 25-minute delay before the second trip is highly diagnostic of thermal-driven insulation failure.
2. Current transformer (CT) failure or saturation — a CT with degraded secondary insulation or a partial short between secondary turns will produce a DC-biased or scaled output. The vector sum drifts away from zero even with a perfectly healthy motor, the drive computes a "leakage" current, and trips F0021. CT failures are the most common cause of F0021 on FS D-Gx that disappears when the vector card is removed, because the vector card hosts the CT signal conditioning on some MM440 hardware revisions.
3. Input or output choke (line reactor) winding failure — a partial inter-turn short on the choke shifts the impedance balance of the three phases, producing a non-zero sum in r0069 and a non-zero leakage current. Choke failures are often accompanied by audible 100/120 Hz hum even at no-load.
4. Drive-internal power-stage defect — IGBT module gate-emitter leakage, DC-link capacitor imbalance, or a defective current-sense amplifier on the CUSA / vector card.

Acoustic Symptom Analysis: Why the Motor Whines Only on VFD

The shrill, high-frequency tone emitted by the motor when operated from the VFD — and its complete absence on direct-on-line operation — is a critical diagnostic clue. The motor is electrically and mechanically healthy on DOL, so the noise is not a bearing or rotor-bar problem. Possible sources of VFD-only motor noise that correlate with F0001/F0021:

• Common-mode voltage-induced bearing currents — the VFD's PWM output has a high dv/dt common-mode component that drives current through the motor's parasitic capacitances to ground. With a damaged CT or failing choke, the common-mode path is asymmetric, exciting audible shaft currents and stator-end-winding vibration at the PWM switching frequency (~4 kHz on MM440 default) and its harmonics.
• Pulse-width-modulated torque ripple — if the drive has lost one phase of current feedback (CT failure on one leg), the controller over-modulates the remaining two legs, producing severe torque ripple at low frequency. This is perceived as the shrill "siren" tone.
• Resonance with the supply-side choke — a choke with partial winding failure has shifted inductance, creating an LC resonance with the cable capacitance near the switching frequency. The motor acts as a sounding board.

On DOL, the motor sees a clean 50/60 Hz sine wave, the choke and CTs are bypassed, and the offending common-mode and torque-ripple mechanisms are eliminated — hence the silent operation.

Diagnostic Parameters

The MM440 exposes a comprehensive set of read-only parameters for diagnosis. The key ones for F0001/F0021:

Trip threshold reference for F0021:

F0021_trip = 0.05 × r0207
For a 37 kW, 400 V MM440: r0207 ≈ 75 A
F0021_trip ≈ 3.75 A residual current

A residual current above 3.75 A on a 30 m cable run with healthy chokes is uncharacteristic — the cable capacitance alone (typical 0.3 nF/m) at 4 kHz switching produces only ~0.1 A of capacitive leakage. Any reading above 1 A residual under steady-state conditions should be considered a hard fault.

Step-by-Step Troubleshooting Procedure

1. Capture fault history before reset. Navigate to r0947 (last fault) and r0948 (fault time) on the BOP. Confirm F0001 and F0021 both appear. Note r0947[0..7] to see the fault sequence.
2. Read r0069 with the motor decoupled. Disconnect the motor leads from the drive output terminals (U, V, W). Run the drive at 50 Hz, no-load. Read r0069[0], r0069[1], r0069[2]. With no motor connected, all three should read < 0.5 A of capacitive noise and the vector sum should be near zero. A residual > 2% of r0207 with no motor is a definitive indicator of CT failure.
3. Megger the motor and cable. With the motor leads isolated from the drive, perform insulation-resistance testing phase-to-phase and phase-to-earth at 500 VDC (or 1000 VDC for 690 V class). Acceptable: > 100 MΩ per phase at 40 °C. Warning: < 5 MΩ indicates definite insulation failure; 5–100 MΩ indicates contamination, moisture, or incipient failure.
4. Megger the cable. Same procedure on the motor cable. Pay particular attention to the screen / armor earth path. A damaged screen creates a leakage path even when the conductors are intact.
5. Check chokes. With the drive powered down and the chokes isolated, measure inductance of each phase of the input and output reactors with an LCR meter. Compare to nameplate (typically ±5% for the three phases). A shorted turn will drop inductance by 30–90%.
6. Verify the vector / CUSA card seating. The fact that the fault "vanished" when the vector card was removed is significant. Re-seat the card, clean the edge connectors with isopropyl alcohol, and re-test. If the fault returns, the card itself is suspect.
7. Test the IGBT modules. With the drive isolated, use a diode-test mode on a DMM to check each IGBT module: emitter-to-collector should show ~0.3–0.7 V in one direction and OL in the other, on all six switches. A shorted module will read < 0.1 V in both directions. A leaky module will read < 1 MΩ in the reverse direction.
8. Factory reset and re-commission only after hardware clearance. A factory reset (P0010 = 30, P0970 = 1) is appropriate only after all hardware tests pass. A reset that "fixes" F0001/F0021 is masking a developing fault.

Insulation Resistance Test (Megger) Procedure

1. Lock out and tag out the drive and motor disconnect.
2. Disconnect the motor leads from the drive output terminals.
3. Disconnect any motor-side surge capacitors, dV/dt filters, or output reactors that would be damaged by DC test voltage.
4. Connect a 500 V (or 1000 V for 690 V class) insulation tester between phase U and earth. Read after 60 seconds (PI test) or after 30 seconds for a single-point reading.
5. Repeat for V-earth and W-earth.
6. Repeat U-V, V-W, W-U for phase-to-phase tests.
7. Record results. Compare against IEEE 43 / IEC 60034-27 minimum values:

Current Transformer (CT) Verification

For FS D-Gx MM440, the phase CTs are typically:

• Mounted on the output busbars inside the drive.
• Connected to the vector / CUSA card via a 3-pin or 6-pin connector.
• Rated for the nominal drive current with a turns ratio that produces ~50 mA secondary at full load.

To test a suspect CT in-circuit:

1. With drive isolated, disconnect the CT secondary lead from the CUSA card.
2. Measure the CT secondary resistance with a DMM — should be a few ohms to a few tens of ohms. A reading of 0 Ω indicates a shorted secondary winding; OL indicates an open.
3. Compare all three CTs. They should match within ±5% on both resistance and (if an LCR meter is available) inductance.
4. If a CT is suspect, replace as a set — never mix old and new CTs on the same drive, as matching is critical for F0021 to read correctly.

On some MM440 hardware revisions, particularly in the 18.5 kW class and below, the CTs are integrated into the power module itself and are not user-replaceable. In those cases, the entire power module must be replaced.

Choke Inspection: Input and Output Reactors

The user's installation includes both input and output chokes. The purpose of each:

A failed output reactor is a strong candidate when F0021 appears together with audible noise: the inductance imbalance creates both the false earth-fault reading and the resonant tone in the motor.

Vector Card / CUSA Module Handling

Removing the vector card appeared to clear the fault in the field case. The two possible interpretations are:

• Interpretation A — CT signal conditioning is on the vector card. A defect on the card (failed op-amp, cracked solder, contaminated PCB) causes the CT signal to be misread, and removing the card breaks the feedback path. This is a card-level failure, not a CT failure.
• Interpretation B — Removing the card reverts the drive to V/F-only control without closed-loop torque regulation. The drive still runs, but the inner current loop is no longer closing on r0069, so F0021 is not evaluated. The fault reappears after 25 minutes because the underlying hardware problem is still present and the trip eventually fires in V/F mode as F0001 (overcurrent) when the IGBT current limit is reached.

Do not interpret the temporary clearing as a fix. The card should be sent for repair or replaced. A factory reset will not resolve a card-level fault.

Repair vs Replace Decision Matrix

Verification and Commissioning After Repair

1. With the motor reconnected, run the drive from the BOP in V/F mode at 10 Hz, 25 Hz, and 50 Hz with no load. Monitor r0069 — the vector sum should be < 2% of r0207 at every frequency.
2. Apply 50% nameplate load, then 100% load, for 30 minutes each. Monitor r0027 (output current) and r0035 (motor temperature model). Confirm no F0001 or F0021.
3. Run the drive for a full thermal soak (≥ 2 hours at full load) and confirm trip-free operation.
4. Capture a parameter set (P0010 = 0, then upload via DriveMonitor / STARTER) and store in the maintenance records.
5. If a vector card or CT set was replaced, re-run the motor identification (P1910 = 1) and rotating measurement (P1960 = 1) to re-tune the controller to the actual motor characteristics.

Preventive Measures

• Schedule annual insulation-resistance testing of motors on drives > 15 kW.
• Verify choke temperature under load — chokes running above 110 °C have shortened life expectancy.
• Keep drive interior clean; dust contamination on the CUSA card and CTs is a documented F0021 cause and can be cleared by careful cleaning in many cases.
• Maintain clean earth bonding at the motor frame, the cable screen, and the drive chassis. A floating screen can develop induced voltages that couple into the CT secondary loop.

FAQ