Resolving Siemens Masterdrives 6SE7021 A017 Alarm: Field Guide

Overview of Fault A017 on 6SE7021 Series

The SIMOVERT MASTERDRIVES 6SE7021 family (Vector Control and Motion Control Plus) reports alarm A017 "SAFE STOP alarm active" under two fundamentally different conditions. When the drive is fitted with the safe-off / safe-stop option board (K80 or equivalent), A017 indicates that the hardware interlock at terminal X9/5-6 has been released. When the drive is a Compact-Plus frame size without the safe-off hardware, A017 is generated by the control firmware in response to a detected anomaly in the IGBT gate firing path, and the alarm is latched until the fault is cleared and the inverter is re-initialised.

The visible symptom is identical in both cases: the unit powers up, runs the start-up self-test, transitions through the operating-state word to state 8, and then flashes A017 on the PMU display instead of advancing to state 9 (Ready to Run, ON command accepted, pulses enabled). The motor remains de-energised; no enable is possible until the underlying cause is removed and the alarm is acknowledged.

Affected Drives and Frame-Size Variants

The 6SE7021 designation covers a wide range of MASTERDRIVES units. The failure pattern described here is most common on the Compact-Plus mechanical frame, of which the 6SE7021-0TP50 is a representative example. Compact-Plus units differ from the larger frames in three ways that are relevant to A017 troubleshooting:

1. There is no separate SL (Signal Logic) board and no separate IVI (Inverter Interface) board. The drive contains only the CUCP (Control Unit, Compact Plus) and a single power electronic board on which the rectifier, brake chopper, IGBT module, gate-drive stage, and current transducers are integrated.
2. There is no internal fibre-optic link between a control board and a power board; all gate-fire commands are transmitted electrically on the power board itself.
3. Because the safe-off option requires a dedicated daughter board and connector plate, Compact-Plus units are never delivered with the safe-stop hardware fitted. The connector footprint (X4) is present on the power board but is normally occupied by a jumper or left open depending on configuration.

Root Cause 1 - Safe-Stop Hardware Trigger (Drives with K80 Option)

On MASTERDRIVES that do carry the safe-stop option, A017 is generated when the normally-closed loop wired to the inverter-pulse-blocking input is opened. Per the Vector Control documentation (Siemens publication SIMOVERT MASTERDRIVES Vector Control, Operating Instructions, Edition 04/2003), the relevant terminals are X9 pin 5 and pin 6. The X9 strip on the power / interface section is the standard MASTERDRIVES pluggable terminal block, and the safe-stop channel is wired across pins 5 and 6 of that block.

When the contact between X9/5 and X9/6 opens - either because the safety relay has dropped out, because the wiring is broken, or because the connector is loose - the control unit sees the "pulse inhibit" signal and raises A017.

Corrective Action for Root Cause 1

1. Power down the drive and lock out the supply.
2. Inspect the X9 connector for proper seating; pull it off and re-mate it to clear any oxide on the contacts.
3. With the drive still isolated, measure continuity between X9/5 and X9/6. The reading must be less than 1 ohm; the safe-stop channel is a dry loop that expects a closed contact.
4. Trace the wiring from X9/5-6 back to the safety relay or jumper. On units that have the option board populated but are running without a safety system, the factory-fitted jumper between X9/5 and X9/6 may have been removed and not replaced; reinstall a shorting link.
5. Re-apply control power, then power, and acknowledge the alarm via the PMU or parameter P398 = 0 followed by re-initialisation.

Root Cause 2 - IGBT Gate-Firing Fault on Compact-Plus

On a Compact-Plus frame drive such as the 6SE7021-0TP50, the safe-stop option is not fitted. A017 is therefore raised by the firmware when the gate-firing diagnostic on the power board detects an anomaly. The CUCP runs a continuous check of the gate-driver supply and the response of the IGBT module to firing commands. The signal path is short - CUCP to gate driver to IGBT module and back through the desaturation / collector-emitter sense - and there is very little hardware to interrogate, but each stage is also non-repairable in the field.

Typical sub-causes, in order of likelihood on a Compact-Plus unit:

1. Failed Compact-Plus transistor pack (single-module construction: the IGBT, freewheel diodes, thermistor, and gate-driver PCB are integrated into one assembly). The most common field failure.
2. Gate-driver supply out of range on the power board (auxiliary 24 V or isolated +15/-15 V rails dropping under load).
3. Open or high-resistance gate-emitter connection inside the IGBT module, presenting as the CUCP seeing "no response" to a test pulse.
4. Shorted IGBT detected at power-up by the pre-charge / impedance test, which the firmware conservatively maps to A017 on Compact-Plus hardware.
5. Broken solder joint or flex-cable between the CUCP and the power board, intermittently breaking the firing path.

The X4 Jumper on the Power Electronics Board

The MASTERDRIVES documentation describes a connector X4 on the power electronics board at the location where the safe-off plate is installed when the option is present. On Compact-Plus units the X4 footprint is exposed and accepts three different plug configurations, but only one of those configurations correctly nullifies the safe-stop input. If the wrong configuration is fitted - for example, a cable harness that was reused from a larger frame - the CUCP will read a permanently open X4 input and raise A017 on every power-up.

Procedure for X4 verification:

1. Power down and lock out.
2. Locate X4 on the power electronics board (it sits behind the front-panel area where the safe-off plate would normally mount on a larger frame).
3. Pull the X4 connector. Inspect each pin for oxidation, bent contacts, and proper crimp retention.
4. Re-mate the connector. The factory configuration on a Compact-Plus unit is a closed loop / shorting plug across the two pins that would otherwise carry the safe-stop signal. The loop resistance must be below 1 ohm.
5. Re-apply power and observe whether A017 persists.

Operating-State Machine and Alarm Path

Diagnostic Procedure - Step by Step

Prerequisites

• PMU or DriveMonitor / STARTER connected for parameter read/write.
• Insulation tester (megger) rated 1000 V DC for motor and cable checks.
• Multimeter with low-ohm range for X4 and X9 continuity.
• Service manual for the specific frame size; firmware version noted from parameter r069 (CUCP firmware version) and r069.1 for the power-section firmware.
• Lock-out / tag-out applied to the line supply; the DC bus must be discharged - confirm with the PMU or by waiting the prescribed 5 minutes after power-off.

Procedure

1. Capture the alarm word. Read r947 (alarm word) and r949 (fault word) before any reset. Confirm bit patterns consistent with A017.
2. Read r833 (heat-sink / module temperature). Compare against the ambient reading taken at the air intake. A reading that is wildly off ambient (negative, or pinned at the upper limit) is an indicator that the IGBT module's integrated thermistor is open or shorted. Note that r833 is the principal in-cabinet thermal reference for the CUCP and is the parameter used by the firmware to validate the firing circuit; a corrupt r833 can cause a false A017.
3. Verify the cooling fan. Confirm the fan is running, free of debris, and the air path is not blocked. A failed fan will eventually overheat the module, but a stuck fan can also drag the internal temperature sensor reading low, masking an actual fault.
4. Check X4 on the power board. Per the section above, confirm the shorting plug is correctly fitted and the loop resistance is low.
5. Check X9/5-6 (only on drives with safe-stop option fitted). Continuity less than 1 ohm when the safety loop is healthy.
6. Megger the motor and feeder cable. Disconnect the motor leads from the drive (U2, V2, W2). Megger phase-to-phase and phase-to-ground. The reading must be greater than 1 megohm at 1000 V. A low reading points to an interwinding or insulation fault, which can also present as a gate-firing diagnostic failure on power-up.
7. Visually inspect the IGBT module. Look for discoloured epoxy, cracked housings, or burn marks around the gate connections. On Compact-Plus the module is a single assembly; if any visual damage is present, replacement is the only field option.
8. Check the connector between the CUCP and the power board. Reseat, inspect for bent pins, and check the ribbon / flex cable for damage.
9. Acknowledge and re-initialise. Set P398 = 0 to acknowledge, then run a power-on reset. If the alarm returns, the diagnosis is hardware - the IGBT module / power card must be replaced.

Temperature Anomaly on r833 - Cross-Check

Field reports on the Compact-Plus frame show that a fault on the IGBT module thermistor or on the analogue sampling chain on the power board can present simultaneously as:

• A017 latched immediately after self-test, AND
• r833 reading a value inconsistent with the measured cabinet and intake air temperature (e.g., pinned to -20 °C, pinned to the upper limit, or jumping by 30 °C between scans).

This pair of symptoms is a strong indicator that the IGBT module is the source of the fault, since r833 is read from the module's integrated NTC. A separate cooling-system problem will usually show a plausible r833 that tracks slowly with load, and will not correlate with A017 latching at start-up.

Spare Parts and Repair Considerations

Verification - Confirming the Repair

1. Re-apply control power only. Confirm the PMU advances to the same pre-state as before, and that A017 is not raised until the DC bus is pre-charged.
2. Pre-charge the DC bus (close the line contactor). Confirm the bus voltage rises to nominal within 2 seconds and that the pre-charge current decays to zero.
3. Verify state 8 is reached with no active alarm. A017 must not be present in r947.
4. Issue the ON command. The drive should transition to state 9 within 50 ms. If the load is the motor, confirm the motor follows the speed setpoint with no surge or stall.
5. Monitor r833 over 30 minutes of loaded operation. The reading should stabilise within 10 °C of the value observed before the repair and should track the load current in r019.

Troubleshooting Matrix

Related Parameters

FAQ