Troubleshooting Motoman UP6 XRC After 400V Applied to 220V Input

Problem Definition: Motoman UP6 Overvoltage Incident

The Motoman UP6 is a six-axis articulated robot originally specified for the North American market with a 220 V three-phase supply. In regions where the distribution standard is 400 V three-phase (Europe, North Africa, parts of Asia), a step-down transformer or autotransformer is required to bring the line-to-line voltage into the controller's tolerated range. When that transformer is bypassed and the controller is wired directly to 400 V, every DC bus, rectifier, switching power supply, and low-voltage logic rail in the XRC cabinet is subjected to a roughly 1.82× overvoltage condition. The result is a cascade of silicon and insulation failures that often does not show a single dramatic external symptom — the cabinet can appear dead, the teach pendant may not initialize, and no alarm code may be raised because the supervisory CPU itself has been compromised.

This article consolidates the field-recovery procedure for a Motoman UP6 / XRC cabinet that has been energized from a 400 V source without the step-down transformer in line. The recovery sequence is board-level: the operator must identify which subassemblies have been overstressed, isolate the surviving modules, and substitute known-good spares to bring the controller back to a controlled-power state before any motion enable.

UP6 / XRC Electrical Input Specification

The Motoman UP6, in its 220 V configuration, is documented at 220 V +10 % input tolerance in the manufacturer's service documentation, which gives an absolute upper limit of 242 V line-to-line. A 400 V European three-phase system delivers 400 V ±10 % (360 V–440 V), which is between 1.49× and 1.82× the rated maximum. The service note 147950-1 from Yaskawa Motoman explicitly references the 220 V +10 % input condition and lists the WRCA01 board as the module associated with input-voltage error reporting. The relevant extract is in Yaskawa Motoman service note 147950-1 (PDF).

A correctly sized step-down transformer (typically 220 V three-phase, K-rated for the inrush of the XRC pre-charge) must be installed between the facility feeder and the controller's main breaker. The transformer's secondary is what defines the controller's input. Bypassing it — even momentarily for a quick power check — is the failure mode addressed in the rest of this document.

XRC Controller Architecture (Subassemblies at Risk)

The XRC (eXtended Robot Control) cabinet is built from a small set of replaceable subassemblies. Knowing which one carries which function is what allows the technician to triage an overvoltage cabinet without a full schematic walkthrough.

The UP6 uses four servo amplifiers to drive its six axes. In the standard XRC configuration, axes 1, 2, and 3 are driven from three individual amplifier modules, while axes 4, 5, and 6 share a single combined amplifier. This grouping is significant: when that combined amplifier is destroyed, the controller loses the wrist axes simultaneously, and the visible symptom is no motion on the tool flange regardless of the teach-pendant command.

Root Cause: Cascaded Overvoltage Failure

The overvoltage applied to a 220 V XRC does not produce a single, localized failure. The first elements to see destructive stress are the components designed to clamp or regulate the input — the line filter MOVs and the primary side of the switching power supply. Once those are shorted, the fault propagates to the DC distribution, where the unregulated +DC bus rises to a value proportional to the applied AC. The downstream loads (CPU board, AMP IGBT gate drivers, encoder receivers) are then exposed to voltages well beyond their absolute-maximum ratings.

Field reports on UP6 cabinets subjected to 400 V input consistently show three failure clusters:

1. Front-end power supply destruction. The +24 V and ±12 V rails collapse or are pulled above tolerance. The teach pendant cannot initialize, and the CPU board does not POST.
2. WRCA01 / input monitoring failure. Because the voltage-sense front end is damaged, the cabinet cannot raise a meaningful alarm code. A technician who expects to see error 4107 (overvoltage) or similar will instead see a dead cabinet with no display activity.
3. Servo amplifier and CPU loss. The combined AMP for axes 4, 5, 6 short-circuits its IGBT output stage. The CPU board (XRK01) loses its motion coprocessor or main processor. The main board behind the CPU card (the backplane) is also typically damaged.

Pre-Diagnostic Checklist

1. Confirm the input feeder. With the controller de-energized and locked out, measure the line-to-line voltage at the cabinet disconnect. Verify it matches what the controller nameplate is rated for. A 220 V XRC connected to 400 V is the failure mode — do not re-energize until the transformer is in circuit.
2. Inspect the fuse holders. Glass and ceramic fuses in the 220 V XRC are sized for ~220 V stress. A fuse that measures 0 Ω is not necessarily healthy if the entire branch it protects has vaporized the trace ahead of it. Visual inspection of the PCB is required.
3. Measure the DC bus. With main power OFF and the controller capacitors confirmed discharged, set a Cat III 600 V meter to DC and measure across the DC-bus test points. A reading above ~5 V indicates the bus is not discharged; wait for the controller's internal bleeder to finish. If the bus reads more than 0 V after 10 minutes, the bleeder resistor has been damaged.
4. Photograph every board. Before removal, photograph the slot positions, cable dress, and any conformal coating discoloration. Heat-stressed boards show brown or yellow halos around hot components.

Diagnostic Procedure: Step-by-Step

Step 1 — Power Supply Verification

Remove the PSU module and bench-test it from a variac. Bring the input from 0 V up slowly while monitoring the +5 V, +12 V, –12 V, and +24 V outputs on a four-channel scope.

• If any rail rises above its nominal before the input reaches 220 V, the PWM controller IC is damaged.
• If the unit draws excessive input current (>2× nameplate) at 100 V input, the input bridge or PFC stage is shorted — replace the PSU as a unit; component-level repair is not cost-effective.
• If the PSU regulates correctly from 0–242 V input, mark it as good and return it to the cabinet.

Step 2 — WRCA01 Input Board

The WRCA01 board is the input-voltage monitor referenced in Yaskawa Motoman service note 147950-1 (PDF). On a healthy unit, this board:

• Scales the three-phase input through a resistor divider.
• Compares the scaled value against a precision reference.
• Drives the alarm latch if the input is outside the +10 %/–10 % window.

Failure modes on 400 V applied:

• The input resistor divider burns open, dropping the scaled voltage to zero. The comparator reads "undervoltage" and never latches an overvoltage alarm — this is why no alarm appears on the pendant.
• The comparator IC itself suffers oxide breakdown and the output drifts or sticks.

Replace the WRCA01 as a unit. Field substitution with a known-good board is the only practical validation path.

Step 3 — XIU01 I/O Unit

The XIU01 hosts the E-stop relay, door interlock loop, and the field I/O opto-isolators. Overvoltage typically destroys the +24 V regulator on this board and shorts the input side of the opto-isolators.

Quick test:

1. With the board removed, measure resistance between the +24 V rail test pad and chassis ground. A reading below 100 kΩ indicates a shorted input.
2. Apply 24 V from an isolated bench supply to the +24 V rail and check that the E-stop relay picks up cleanly. If the relay buzzes or fails to pick, the relay contacts have welded from the inrush.

Replace the XIU01 if the rail is shorted or the E-stop relay fails the bench test.

Step 4 — XRK01 CPU / Main Board

The XRK01 is the controller's CPU. Symptoms of failure on a 400 V event:

• Teach pendant backlight illuminates, but the screen shows garbage or no text.
• Pendant remains blank and the controller's heartbeat LED on the CPU board is dark.
• POST error code on the seven-segment diagnostic display (if equipped) is absent — the CPU never reached POST.

Field triage:

1. Remove all plug-in daughter cards (servo comms, encoder receiver, etc.) and re-test with the minimum CPU configuration.
2. Measure +5 V at the CPU board's logic-rail test point. If the rail is out of spec, the issue is upstream (PSU or backplane), not the CPU itself.
3. If the +5 V is correct and the CPU is silent, replace the CPU board and its main backplane card together — both are typically stressed by the same overvoltage event.

Step 5 — Servo Amplifiers (AMP1–AMP4)

Each amplifier module has:

• A three-phase input rectifier bridge.
• A DC-link capacitor bank.
• An IGBT output stage (six-pack or three-half-bridges depending on axis count).
• Gate drivers and current-sense amplifiers.

400 V applied to a 220 V amplifier drives the DC bus from a nominal ~310 V to ~565 V. The DC-link capacitors are rated for 400 V continuous; they will fail short or vent. The IGBT modules' V_CES rating is typically 600 V, so individual switches may survive a brief overvoltage event, but the gate-driver supply rails will not — and the amplifier will not be able to switch cleanly even if the silicon is intact.

Triage on the combined axes 4/5/6 amplifier:

1. With the amplifier removed, measure resistance phase-to-phase on the AC input. A reading below 50 kΩ indicates a shorted input rectifier.
2. Measure resistance from each output terminal (U, V, W) to DC+. A shorted IGBT will read under 1 kΩ in both polarities.
3. Check the DC-link capacitor housings for venting (the scored top is bulging or pushed out).

Replace any amplifier that fails any of the above. The combined axes 4/5/6 amplifier is a single replacement part — there is no field-repair path on the IGBT modules themselves.

Replacement Strategy and Spare Board Sourcing

The minimum set of boards required to bring an overvoltaged UP6 back to a controlled-power state is:

Order the WRCA01, XIU01, and XRK01 from a Yaskawa Motoman authorized distributor. The amplifier modules are also distributor-sourced. Used boards pulled from a known-good retired cabinet are acceptable provided they are bench-tested before installation.

Reassembly and First Power-Up

1. Reinstall the boards in the order: PSU, WRCA01, XIU01, XRK01, AMP1–AMP4. Seat each connector firmly; the high-current AMP connectors require a firm click.
2. Verify the input transformer is in circuit and the secondary measures 220 V line-to-line at the cabinet disconnect. Do not skip this step.
3. Power on the controller with the pendant disconnected. Listen for the contactor pull-in, the PSU fan spin-up, and the CPU card's heartbeat LED.
4. Reconnect the teach pendant. The pendant should display the XRC controller ID screen within 5–10 seconds of power-on.
5. Navigate to System → Alarm History and clear all alarms. The controller may raise a 4107-class overvoltage alarm the first time it sees the input if the WRCA01 has a valid comparison reference; this is normal after a board replacement.
6. Enable servo power from the pendant. Each axis should hold its current position with no fault. If any axis drops with a 4104 (overcurrent) or 4310 (encoder) alarm, return to the amplifier diagnostic step.

Verification Matrix

Troubleshooting Matrix: Symptom to Suspect Module

Prevention: Step-Down Transformer Sizing and Verification

For a UP6 / XRC, the step-down transformer from 400 V to 220 V three-phase must be sized for the controller's full-load input current, not just the average. The pre-charge inrush of the DC bus is several times the steady-state current for the first 100–200 ms, and an undersized transformer will saturate and deliver a voltage far below 220 V on inrush, causing undervoltage trips on its own.

Rough sizing rule (verify against the nameplate FLA of your specific cabinet):

• Single-phase apparent power: kVA = V × I / 1000
• Three-phase apparent power: kVA = √3 × V_LL × I_line / 1000

For a typical UP6 XRC, the input FLA is in the order of tens of amps at 220 V three-phase. As a single worked example, if the input is three-phase line current at 30 A line-to-line:

• kVA = 1.732 × 220 × 30 / 1000 ≈ 11.4 kVA

Apply the same formula with the actual controller FLA from the cabinet nameplate. The transformer should be sized at 1.25× to 1.5× of the calculated kVA to handle inrush without saturation. Verify the specific FLA against the cabinet's data plate — the values above are illustrative and must be replaced with the actual rating before ordering hardware.

Field Notes and Engineering Caveats

• No alarm is normal for this failure. Operators who expect a specific overvoltage alarm will be confused by a silent cabinet. The absence of an alarm is the diagnostic finding, not a healthy state.
• The combined axes 4/5/6 amplifier is the most common single point of failure. This is the only amplifier in the cabinet that drives more than one motor, and its destruction disables the wrist entirely.
• The backplane card behind the CPU must be inspected. Heat discoloration on the backplane is a sign that the CPU socket itself has been stressed; replace both together.
• Do not reflow the WRCA01 input divider in the field. The divider is matched to the rest of the input monitor; replacing the board is faster and more reliable than component-level repair.
• Re-validate every AMP, even the survivors. An amplifier that survived 400 V applied for several seconds has had its gate-driver electrolytic capacitors stressed past their rated voltage. Bench-test before trusting it in service.

Related Documentation

• Yaskawa Motoman service note 147950-1 (PDF) — input voltage specification and WRCA01 alarm association