Simodrive 650 F-81 Fault: Diagnosing DC Link Overvoltage

Problem Summary: Intermittent F-81 on Simodrive 650 Spindle Drive

The SIMODRIVE 650 series spindle drive (Siemens part number 6SC6508-4AA02-Z) used on a Gildemeister CTX400 lathe retrofit is reporting fault code F-81 (DC Link Overvoltage) in a non-deterministic manner. The fault appears at any point during the operating cycle and is sometimes cleared by the front-panel P (reset/clear) key, while in other cases only a mains power-cycle is sufficient to recover. A diagnostic reading at the DC link capacitor forming test point shows the bus voltage peaking at approximately 170 V DC, well below the 300 V DC target for the P06 = 163.0 setpoint display. This symptom combination — random F-81 trip and under-referenced DC link forming voltage — points to a degraded DC link, a fault in the regeneration/braking path, or a defective control board in the U1/G2 signal chain.

Affected System Identification

The host CNC in the Gildemeister CTX400 retrofit is typically an 810T/M or 840C SINUMERIK. The SINUMERIK 840C / SIMODRIVE 611-D Diagnostics Manual provides related stop-F and overvoltage error-coding conventions that share conceptual roots with the 650-series fault list, although the 650 uses its own firmware-level F-code numbering (the 611-D uses 840C-side F-numbers that are not interchangeable).

F-81 Fault Code Definition and Trigger Conditions

F-81 is the SIMODRIVE 650 firmware's identifier for DC link overvoltage, raised when the firmware detects that the rectified DC bus voltage has exceeded a firmware-dependent threshold derived from the mains input. In a 3 × 400 V AC supply the nominal DC link sits near 540–565 V DC (peak line-to-line × 1.414 minus rectifier drop). F-81 is therefore not normally a mains-supply problem but a regeneration problem: the drive's intermediate circuit is absorbing more energy from the spindle than the line-side rectifier can dissipate or sink.

Three root-cause families drive F-81 in the field:

1. Regeneration / brake chopper failure. The internal brake transistor (typically mounted on the G2 board or on the heatsink assembly referenced by it) fails to fire, so the DC link has nowhere to dump the regenerated energy during spindle deceleration.
2. External braking resistor open-circuit. The external ballast resistor (if fitted) is open or the wiring to its terminals is loose, so the brake chopper has no load to switch into.
3. DC link capacitor degradation. The aluminum electrolytic capacitors in the DC link have lost capacitance and/or developed high ESR. This causes the bus voltage to overshoot on every regeneration event and can also confuse the voltage-sense divider feeding the U1 board.

The reported behavior — random F-81 with intermittent reset and a DC link measurement of only 170 V at the forming test point — is the classic signature of item 3 compounded with item 1 or 2.

Diagnostic Parameter Map

Read the following parameters from the front panel of the 6SC6508 and record the displayed values before clearing the fault:

Interpreting the 170 V DC Forming-Voltage Reading

The "DC link capacitor formed" procedure described in the SIMODRIVE 650 manual is a controlled, current-limited charging routine. With the drive in service mode and the rectifier enabled, the technician connects a DC voltmeter to test point P600 / M and reads the bus voltage. For a healthy DC link with 3 × 400 V AC input the reading should reach approximately 300 V DC once the bulk capacitors are reformed; lower line voltages scale this proportionally.

A clamped reading of 170 V DC means the DC link is being held below the expected value. The most likely explanations, in order of probability for a 25+ year-old drive, are:

1. Capacitor leakage / hard-shorted cells. A shorted electrolytic cell will pull the bus down to a level set by the leakage path and source impedance of the soft-charge resistor. The P600 voltage will plateau at a value well below nominal.
2. Blown soft-charge resistor / input fuse. The NTC thermistor and current-limiting resistor used during the pre-charge cycle are open, so the rectifier is feeding the capacitors only through a partial path.
3. Rectifier bridge partial failure. One or two diodes open in the input bridge drop the rectified voltage and cause a high-impedance bus.
4. Load on the DC link from a shorted IGBT module. A shorted brake transistor or inverter IGBT will hold the bus down and is often accompanied by additional fault codes (F-86, F-87, F-95 depending on firmware revision).

Step-by-Step DC Link Capacitor Forming Procedure

The forming procedure is intended to re-form the aluminum oxide dielectric layer on DC link electrolytic capacitors that have been de-energized for long periods. Do not skip the current-limited steps; a full mains voltage applied to a depleted capacitor bank can rupture the cans or vent hot electrolyte.

1. Isolate the drive from mains. Lock out and tag the disconnect. Wait at least 5 minutes for the DC link to self-discharge through the internal bleeder.
2. Verify zero DC link voltage at P600/M with a calibrated DMM set to 1000 V DC. If any voltage remains, the bleeder is open — repair before proceeding.
3. Apply a current-limited DC source of approximately 30–50 mA to the DC link through a 10 kΩ / 5 W resistor connected to P600 (+) and M (−). A bench supply set to 50 V DC works for the first stage.
4. Hold the current-limited voltage for 30 minutes per 50 V step. Increase the supply in 50 V increments up to the rated DC link voltage (≈ 540 V DC for 400 V AC mains).
5. After reaching the target voltage, monitor the leakage current. A healthy capacitor bank draws < 5 mA at rated voltage after 60 minutes. Anything significantly higher indicates a leaky cell that must be replaced.
6. Re-apply mains power with the drive in service mode and re-read P600/M. A healthy bank will reach 300 V DC (or 540 V DC depending on the procedure variant — confirm against the 650 manual revision) within seconds of the soft-charge relay closing.

Board-Level Root Cause Analysis: U1 and G2

With DC link forming voltage at 170 V but the drive still attempting to run, the firmware flags F-81 because the under-voltage condition is interpreted by the watchdog as an overvoltage event during the next regeneration cycle. The 650 firmware monitors bus voltage through a divider on the U1 board; if the divider itself is drifted or if the G2 board is reporting a false "brake fired" feedback, the firmware can latch F-81 spuriously.

U1 Board (6SC6500-0UC01) Checks

• Inspect the U1 board for capacitor leakage around the DC link voltage-sense divider. Electrolytic SMD caps in the 47–100 µF range near the bus-voltage input pin are the most common failure point.
• Measure the resistance between the bus-sense pin and ground. It should match the divider ratio to within ±5%; large deviation indicates drift.
• Confirm the firmware EPROM is seated and that the firmware checksum is correct (P99 if available on this firmware revision should match the value on the EPROM label).

G2 Board (6SC6100-0GE01) Checks

• Inspect the gate-drive optocouplers and the brake-transistor gate resistor network. The G2 board is the interface between the low-voltage U1 control and the high-voltage IGBT module; optocoupler LED degradation causes intermittent gate firing.
• Check the brake IGBT directly with a curve tracer or by substitution. A shorted collector-emitter on the brake transistor is the single most common hardware cause of F-81.
• Inspect the heatsink-mounted current-sense resistor for the brake path. Open sense resistor causes the brake to be inhibited by the desaturate protection, but the firmware then sees no regen path and raises F-81.

Replacement and Sourcing Strategy

With the 650 series out of production for over two decades, sourcing is the principal challenge. The recommended sequence is:

1. Refurbish in place. Replace the DC link capacitor bank as a matched set and rebuild the G2 board's optocouplers and the U1 board's sense-divider electrolytics. This is the lowest-cost option and is appropriate when the IGBT modules themselves test good.
2. Order tested used spares. Reputable industrial surplus suppliers stock tested 6SC6500-0UC01 and 6SC6100-0GE01 boards. Order the U1 board first — it is the more frequent failure point for F-81 when the DC link itself is healthy. If F-81 persists, order the G2 board.
3. Migration to SIMODRIVE 611A spindle module. When neither repair path is viable, the supported migration is to a SIMODRIVE 611A modular system using a 6SN1123-1AA0_ spindle module. Confirm motor compatibility: the 1PH6161-4CF00-Z is a 1PH6-series AC spindle; some 1PH6161 variants map directly to 611A spindle modules but the exact -Z option code must be cross-referenced against the 611A motor list before committing to a swap.

Braking Resistor and Regeneration Path Verification

Before declaring the boards defective, verify the external regen path. The 650-series spindle drive typically connects to an external ballast resistor through terminals on the power section. With the drive isolated:

1. Disconnect the ballast resistor leads and measure its resistance. Compare to the value stamped on the resistor body (commonly 10 Ω, 500 W or 20 Ω, 1 kW depending on the 650 variant).
2. Inspect the wiring from the resistor to the drive for oxidation or broken strands — a high-resistance joint mimics a brake failure.
3. If no external resistor is fitted, confirm the drive is configured for the internal braking option (parameter dependent on firmware rev 14 — refer to the 650 manual's parameter list, P105 group).

Verification and Re-Commissioning

After repair, perform the following verification sequence before returning the machine to production:

1. Power the drive in service mode. Confirm P95 = 0, P96 = 0, P98 = 0 (no active faults).
2. Read P06 and P600/M with a calibrated DMM. The P600 reading should be within ±5% of the value indicated by the P06 parameter scaling.
3. Run an unloaded spindle ramp from 0 to maximum RPM and back, monitoring the DC link voltage on P600/M. It must stay below the firmware's F-81 trip threshold (typically 720 V DC for 400 V AC input) throughout the deceleration phase.
4. Apply a full-load cutting cycle representative of the lathe's typical work envelope. Monitor for F-81 recurrence over a minimum of 8 hours of intermittent duty.
5. Document the final parameter set and the values of P95/P96/P98 at the end of the verification run.

Long-Term Reliability and Migration Planning

Even with a successful repair, the 650 series is operating well past its designed service life. The supporting components — line filter capacitors, fan bearings, display EL panels, and the front-panel membrane keypad — are all time-limited. For a production-critical machine, plan a staged migration to a SIMODRIVE 611A spindle module or to a current-generation SINAMICS S120 spindle drive within a 24-month window. Until then, keep at least one each of the U1 (6SC6500-0UC01) and G2 (6SC6100-0GE01) boards on the shelf as hot spares.