Siemens Master Drive 6SE70 F026: CT and IGD Diagnostic Guide

Siemens Master Drive 6SE70 F026: CT and IGD Diagnostic Guide

When a SIMOVERT Master Drive 6SE7023-0FD61-Z throws F026 the instant a run command is issued, the symptom is straightforward but the cause is rarely a single component. Field experience with this fault on the 6SE70 platform shows that the DC link, the motor, and even a freshly swapped IGBT can all be perfectly healthy and the drive will still trip. The fault is in the current-sensing signal chain, and that chain has three independent sections that must all be validated: the IGD (Insulated Gate Driver) modules mounted on the IGBT stack, the small current transformers embedded inside those IGDs, and the sensing / power-supply (PSU) card that supplies the IGDs and conditions their analog output to the CUVC control board.

This article walks the engineer through the full F026 diagnostic tree for the 6SE70, the order in which to test, the specific resistance values that indicate a healthy CT (23.9 Ω and 53.5 Ω are the two reference measurements taken directly from field-proven Master Drive units), and the practical limits of what can be confirmed without the original DriveMonitor or SIMOVIS tool.

1. Fault F026 in the 6SE70 Fault Register

On the SIMOVERT Master Drive (6SE70 series), the CUVC firmware classifies faults into numbered groups. F026 belongs to the current / output stage group and is the firmware's response to an out-of-range reading on the current-feedback path during the IGBT pre-charge or first switching cycle. The drive does not require a fully loaded motor or a mechanical load to trip on F026: the fault is raised at the moment the CUVC samples the first current feedback word, which is during the gating test that precedes torque-producing switching.

Behaviorally, the F026 trip on a 6SE70 has three characteristic shapes:

1. Immediate trip on the first run edge. The drive is given a run, the CUVC commands gating, the IGD either fails to respond or reports a current offset above the firmware window, and the drive is shut down in a few milliseconds. Status word r001 reports fault active, fault buffer r947 / r949 list F026 as the most recent entry, and the CUVC disables the IGBT gate drivers.
2. Trip within 50-200 ms of run. The drive begins pre-magnetization or initial torque build-up, then trips as the CUVC's r029 current readback fails its plausibility check against the r028 torque / r027 frequency demand. This usually points to a degraded IGD or a sensing-card offset rather than a hard open circuit.
3. Trip on every run, no fault on parameterization. The drive accepts parameter downloads, encoder check, and the operator panel's no-fault state, but every physical run command raises F026. This is the classic IGD / sensing-card signature and is the case this article addresses.

The fault is latching. The drive will not auto-reset on F026; the operator must clear it via the panel or the F-key on DriveMonitor. After a hardware repair, a manual reset is mandatory before another run attempt.

2. Identifying the Specific Drive: 6SE7023-0FD61-Z

The order number decodes as follows, which is essential to validate spare-part compatibility and the CUVC firmware branch. The 6SE70 nameplate carries the input kVA / kW and the output A at the rated point; verify the exact rating against the original Siemens catalog page before ordering spares.

3. Confirming the Healthy Subsystems First

F026 is a current-sensing fault, but the most expensive false paths are already eliminated. These must be reconfirmed in writing, not just by memory, because the CUVC's fault buffer is overwritten as soon as another run is attempted.

3.1 DC Link Voltage

Measure the DC link at the test points on the front of the drive (P6000+, P6000- on most 6SE70 frames). For a 400 V class drive the DC link should read:

If the DC link is outside the ready-state band, fix the supply first. Under-voltage causes F008 / F006, and a weak DC link can cause IGD mis-readings that look like F026. A nominal 540-620 V DC reading here is the green light to look at the current-sensing path.

3.2 Motor and Cable

Swap the motor (already done in the case in question) and confirm:

• Insulation resistance phase-to-phase and phase-to-ground at 500 V or 1000 V (depending on motor rating) — typically > 100 MΩ for a healthy motor.
• Cable continuity end-to-end (no open conductors, no shorted pairs).
• Shield grounded only at the drive end. The 6SE70 is sensitive to high-frequency common-mode currents from incorrectly grounded shields; these can corrupt the IGD's reference and look like a current-sensing failure.

3.3 IGBT Module (Already Replaced)

If the IGBT was swapped, the engineer must verify the replacement was of the correct part number (not just "same case") and that the IGD was transferred or replaced with it. A common error is to replace the IGBT module but reinstall the original IGD, which is the actual point of failure. The IGD is mounted on the IGBT with three screws and a heatsink clip; the ribbon cable from the IGD to the sensing card is keyed and short.

After a confirmed-correct IGBT swap, the gate-emitter voltage test should be done before any further run attempt:

1. With the drive locked out, DC link discharged, and the IGD ribbon disconnected, measure V_GE on the IGBT. A healthy IGBT shows a stable gate threshold (typically 5-7 V at zero collector current, the IGBT's natural V_GE(th)) and no short from gate to emitter or gate to collector.
2. If V_GE reads shorted or zero, the new IGBT is also damaged — typically from a missing anti-static procedure during installation or a previous shoot-through. Replace the IGBT again and recheck the IGD ribbon cable for damage.

3.4 CUVC Board (Already Cross-Verified in a Second Drive)

The CUVC has been verified as functional in a second drive. This is a powerful negative test — it eliminates the entire control section (firmware, processor, parameter memory) from the fault path. The remaining candidates are: the IGD, the CT inside the IGD, the ribbon / wiring between IGD and sensing card, the sensing / PSU card itself, and the IGBT stack (already replaced).

4. Root-Cause Matrix for F026

The diagnostic tree is shown below. Each row is a confirmed path to F026, ranked by field frequency on the 6SE70 platform.

5. IGD (Insulated Gate Driver) Failure Analysis

The IGD is the single component most often misidentified on the 6SE70. From the outside it looks like a small plastic brick bolted to the IGBT module, with a short ribbon cable carrying gate drive, current-sense, and ±15 V supply. Internally, the IGD integrates:

• Gate-drive circuitry for the IGBT, with V_CE(sat) desaturation detection (this is the UCE / F025 path — the well-known UCE fault).
• A small toroidal current transformer (CT) that senses the AC current in the IGBT emitter lead, with a turns ratio that produces a few hundred millivolts per ampere on the secondary.
• A linear optocoupler or isolated amplifier that transmits the CT secondary signal across the high-voltage isolation barrier to the sensing / PSU card at low-voltage potential.
• A local DC-DC converter that generates the isolated supply rails for the gate drive from the +24 V back-plane power.

When the IGD fails, it can fail in any of the four sub-sections. The most common failure modes, in order, are:

1. CT secondary open or shorted. This is the failure that produces the 23.9 Ω / 53.5 Ω reference-resistance checks. The CT's primary is the IGBT emitter lead (essentially a single turn, so 0 Ω on the primary side is normal), and the secondary is a multi-turn winding with a defined DC resistance. Reading 0 Ω or open circuit on the secondary pin pair is the failure signature.
2. Optocoupler LED degradation. The optocoupler that transmits the CT signal to the low-voltage side ages; the transfer ratio drops, the offset drifts, and the CUVC eventually sees a current that does not match the demand. This is a soft fault — the drive may run for a few seconds, or only trip on a particular phase.
3. DC-DC converter collapse. The +24 V-to-isolated-rail converter inside the IGD fails and the IGBT gate is not driven, or the CT is not biased. The drive may still raise F025 (UCE) first if this collapse happens suddenly, or F026 if the gate drive is partially working but the CT reference is unstable.
4. Desaturation detector false trigger. Caused by a noisy V_CE sense circuit inside the IGD, often correlated with gate-emitter resistance drift. This raises F025, but on some CUVC firmware versions it can be misclassified as F026 if the current-sense path is the first out-of-window signal.

The IGD is field-replaceable as a unit. It is not repaired at component level by most service shops. Siemens part numbers for the IGD depend on the frame size (the "23" in the MLFB) and on the power section variant. The replacement must match both the frame and the variant — verify the part number on the existing IGD's label against the Siemens spares catalog before installing.

6. Current Transformer Resistance Verification

The two reference values 23.9 Ω and 53.5 Ω are the DC resistances of the CT windings inside the IGD, measured at the IGD ribbon connector with the ribbon disconnected from the sensing card. The exact mapping of which pin pair corresponds to which winding is documented on the IGD data sheet, but for the 6SE70 service-test purpose the engineer needs to know only that two distinct resistance values must be measured, both non-zero, and both within a few percent of the reference.

6.1 Test Procedure

1. Lock out, tag out, wait five minutes, verify DC link < 60 V DC.
2. Open the drive, identify the IGDs (one per output phase, mounted directly on the IGBT modules).
3. Disconnect the IGD ribbon cable from the sensing / PSU card end. Do not disconnect at the IGD end if it is not a pluggable connector — leaving it on the IGD does not affect the resistance measurement.
4. Set a digital multimeter to resistance (Ω), low range (200 Ω or autoranging). Use the meter in low-voltage resistance mode (typically < 1 V test signal) to avoid driving the IGD's internal protection.
5. Identify the CT secondary pin pair on the ribbon connector. On the 6SE70 IGD this is a defined pair; the location is in the wiring diagram on the inside of the drive door. Probe the pair and record the reading.
6. Compare to 23.9 Ω ± 5% and 53.5 Ω ± 5%. A reading of 0 Ω indicates a shorted winding (catastrophic failure, replace IGD). A reading of open circuit (OL) indicates an open winding (catastrophic failure, replace IGD). A reading outside the ±5% window but non-zero usually indicates a partially shorted winding or a corroded connector — clean and re-test; if the value does not return, replace the IGD.
7. Repeat for all three phases. A failure on one phase is enough to produce F026; replace the IGD on the failed phase first and re-test before considering all three.

6.2 Phase-by-Phase Tracking

For multi-phase tracking, label the IGDs U, V, W (matching the IGBT stack labeling) and record the resistance of each winding on each phase. A table is the most useful format for cross-comparison.

7. Sensing / PSU Card Diagnosis

The "PSU card" or "sensing card" referenced in service discussion of the 6SE70 is the auxiliary board that sits between the CUVC and the IGDs. It performs three functions:

1. Generates or distributes the +24 V / ±15 V power rails that feed the IGD DC-DC converters and the analog signal path.
2. Conditions the IGD current-sense output (a few hundred millivolts at zero current, a few volts at full load) into a clean analog signal in the CUVC's input range (typically 0-10 V or ±10 V depending on CUVC variant).
3. Provides the IGBT gate-driver fault feedback (the "OK" line) back to the CUVC. A missing or stuck-low OK line raises the UCE fault (F025). A noisy OK line, or one that drops only on the first current pulse, can be classified as a current-sensing failure on some firmware versions, producing F026.

7.1 Power-Rail Verification (Drive Powered, No Run)

With the drive in ready state (line on, run off, no fault active), measure at the IGD ribbon connector:

Any of these rails being missing or out of tolerance is sufficient to raise F026 on the first run command. The CUVC's first check on receiving a run is "is the current feedback at zero?" — if the rail is collapsed or the offset is hundreds of millivolts, the CUVC's check fails and it raises F026 before even trying to switch the IGBTs.

7.2 Replacement Note

Replacement sensing cards must match the frame size and the CUVC firmware version. The 6SE70 has at least three sensing-card variants, distinguished by the output stage (analog ±10 V vs. digital word) and the isolation voltage class. Cross-referencing the original card's part number is mandatory; the card's connector pinout differs across variants, and a wrong card will fit the connector but produce either no feedback or a scaled-wrong feedback, both of which raise F026 immediately.

8. CUVC Board Validation

The user has already cross-verified the CUVC in a second drive and found it functional. Two notes from field practice:

1. The CUVC's current-feedback input stage uses small-signal op-amps that can drift over years. A CUVC that works perfectly in a small test drive may still be marginal in a larger drive if the input stage has aged and the absolute voltage the CUVC expects at full load is slightly different. The cross-test is strong but not absolute; if all other paths are eliminated, the CUVC remains a candidate.
2. The CUVC has on-board EPROM / flash for firmware. The 6SE70 has several firmware versions; some early versions had a tighter F026 window than later versions. Re-flashing to the latest firmware is sometimes a workaround, but it requires the Siemens DriveMonitor tool and a serial / USS connection. Do not attempt a firmware flash on a drive that is still on the production line without a controlled backup of the parameters.

9. Step-by-Step Diagnostic Procedure for F026

The complete, ordered procedure is below. It is written so that the most likely cause is checked first, the cheapest fix is tried first, and the drive is not re-energized until a measurement is made.

1. Confirm safety. Lock out, tag out, wait five minutes, measure DC link < 60 V DC.
2. Read the CUVC fault buffer. On the operator panel, navigate to r947 / r949 (or use DriveMonitor parameter P947 / P949) and record the last eight faults. F026 with no preceding fault confirms the current-sense path. F026 preceded by F025 (UCE) suggests the IGD's UCE detector is the source — but the underlying cause is often the same IGD, so continue.
3. Visual inspection. Open the drive, look for burnt ribbon cable, cracked IGD housing, electrolyte leaks from the sensing card, or discolored connectors. Any of these is a fast path to replacement.
4. Re-seat the IGD ribbons. Power down, re-seat each IGD ribbon at both ends, restore power, attempt run. If the fault clears intermittently, the ribbon / connector is the source. Replace the ribbon.
5. Measure CT resistance on all three IGDs. Use the 23.9 Ω / 53.5 Ω reference and the table in section 6. Replace the IGD on any phase that fails.
6. Swap IGDs between phases. If all CT resistances are in spec, swap the IGD from the U phase to the W phase, then attempt run. If F026 is replaced by a different fault or by an F025 on a specific phase, the IGD is the source. Replace the IGD that produced the original F026.
7. Measure sensing-card power rails. Drive powered, no run. Confirm +24 V, +15 V, -15 V, and zero-current offset per section 7.1. Replace the sensing card if any rail is out of spec.
8. Substitute a known-good sensing card. If a spare is available, swap the card and re-test. The CUVC firmware parameters do not need to be changed when only the sensing card is swapped, as the scaling is on the CUVC side and the sensing card is a passive / analog conditioner.
9. Substitute a known-good CUVC. Already done in the case in question. Cross-test eliminates CUVC.
10. Verify the IGBT gate. With the IGD ribbon disconnected, measure V_GE on the IGBT per section 3.3. A failed IGBT gate can pull the IGD's optocoupler out of range, producing a chain failure. Replace the IGBT if V_GE is shorted or zero.
11. Re-energize and run. With all of the above clean, restore power, clear the fault, and attempt a no-load run. Monitor r029 (output current readback) and r028 (torque) for the first 500 ms. Both should ramp smoothly to their commanded values with no transient step.

10. Component Replacement and Cross-Verification

After any component replacement, the following sequence is required before declaring the repair complete:

11. Recurrence Prevention and Commissioning Notes

For drives returning to service after an F026 repair, three field notes reduce the risk of recurrence:

1. Torque-check the IGD mounting screws. A loose IGD has a poor thermal path to the IGBT, drifts in offset over a few thermal cycles, and is a high-probability cause of intermittent F026 that does not show on the bench. The Siemens service manual specifies the torque value for the IGD-to-IGBT mounting; use a torque driver, not a screwdriver.
2. Verify the ribbon cable is not pinched by the cabinet door or the IGBT clamp. Pinch damage shows up as a high-resistance short that drifts with temperature. This is the second most common intermittent cause after loose mounting.
3. Document the firmware version of the CUVC and the part number of every replaced IGD and sensing card in the drive's service log. The 6SE70 platform has long service life (20+ years in some installations) and the next F026 event on the same drive will be diagnosed in a fraction of the time if the previous repair is recorded.

For a 6SE70 that has not yet been commissioned, the Siemens DriveMonitor / SIMOVIS commissioning sheet includes a current-loop test step that exercises each IGD and verifies the sensing-card output against the CUVC's readback. This step is documented in the Master Drive 6SE70 commissioning manual on the Siemens Industry Online Support portal. Run the step on every drive that has had a power-section intervention, not only on drives that have actually failed.

12. Cross-Reference: F026 vs. Adjacent Faults

Several faults on the 6SE70 are related to F026 by signature. The diagnostic shortcuts are:

13. Quick Field Reference Card

FAQ