Yaskawa MRC II Alarm 5390 Servo On Error: Field Troubleshooting

1. Problem Overview: MRC II Alarm 5390 on SK6 Manipulator

Alarm 5390 SERVO ON ERROR on a Yaskawa Motoman MRC II controller indicates that the servo power-on command sequence has failed. The teach pendant and the MTU01 (Main Transformer Unit) relay bank drive the two large main contactors that feed the SVAC/SVDB servo amplifiers; if the controller cannot validate the safety chain, the contactors do not close and the system drops out with 5390 within a few hundred milliseconds of the SERVO ON request.

This alarm is platform-specific to the SK6 (and related SK-series) manipulators controlled by the MRC II (Motoman Robot Controller, generation II) cabinet. It is functionally distinct from later-generation alarms such as NX100 ALARM 0021 (SERVO COMMUNICATION ERROR), which is a serial-link / SVBIF fault on the XRC/NX100 platform and is referenced here only for cross-controller comparison.

Key field observation in the reported case: the SUMX LED on the MTU01 board flashed briefly, relay 4 (OMRON socket) energised, but neither of the two main contactors pulled in. The robot moves correctly when the contactors are manually closed, confirming that the servo amplifiers, MRY01 axis board, MIF01 interface, and the manipulator itself are healthy. The fault therefore lies in the enable chain, not in the drive stage.

2. Affected Hardware Stack (MRC II SK6)

Before chasing alarm 5390, map every board that participates in the servo-on enable path. The MRC II SK6 uses the following hardware:

3. Root Cause: Enable Chain Not Closed

Alarm 5390 is generated when the executive software asserts SERVO ON but the hardware ready feedback (normally the auxiliary contacts of the main contactors, the safety relay, and the DEADMAN loop) does not return a healthy state within the controller's watchdog window (typically 200–500 ms on MRC II firmware revisions for SK6-C00).

The chain, in energise order, is:

1. Operator presses SERVO ON on the teach pendant.
2. MSP board encodes the request and forwards it via the play-back cable to the MCP10.
3. MCP10 consults the MMM02 robot parameters; if the robot type mismatches the manipulator serial, the request is rejected before the contactor stage (typically as 4077 / 4087 input faults rather than 5390 — see Section 7).
4. MCP10 commands MTU01 relay K4 (sometimes K3, depending on wiring revision) to energise.
5. K4 closes the coil circuit of the two main contactors; their AUX NO contacts close and feed back to the MIF01 dedicated inputs.
6. The SRB-NA-R-C15-24V safety relay must also be in the closed/run state. Its X1/X2 output contact is wired in series with the contactor coils on most SK6 cabinets built for the European market.
7. If feedback is missing or open, MCP10 drops the request and posts 5390.

The single most common field cause on a second-hand MRC II is an open safety relay that the previous owner wired to perimeter light curtains, area scanners, or a safety PLC that is no longer present in the cabinet. The contactor coils simply have no path to 24 V DC.

4. Prerequisites for Diagnosis

• Teach pendant with functional keypad and LCD (F1–F5 softkeys present).
• MRC II maintenance-mode password (factory default for SK6-C00 firmware: *** entered from the customer menu — confirm against the maintenance manual; do not publish a default if unknown).
• Digital multimeter capable of reading 24 V DC and continuity.
• Wiring diagram set for SK6 / MRC II cabinet (the SK6-XXXX series drawings are SKU-specific; cross-reference to the controller serial plate).
• Insulated jumper leads for test purposes only; remove before returning the cell to service.

5. Step-by-Step Diagnostic Procedure

1. Verify the robot-type match. Power up in maintenance mode. Press [DISP], then [F5]. On some pendant firmware revisions the F5 key is labelled APL rather than DIAG; press the softkey under the right-most position. Navigate to ROBOT TYPE and confirm it reads SK6-C00. Read the version label on the MMM02 piggy-back EPROM (e.g., JANCD-MMM02-5E REV.D2 DF9200720-C2). A mismatch between software and manipulator is the first thing to rule out.
2. Inspect the MTU01 relay bank. Power down, open the cabinet, and check that all four relay sockets are populated with OMRON-style 24 V DC relays. Some export cabinets have a fifth socket (K5) used for an external enable; a missing relay here will not latch the chain. With power on, listen for K4 pulling in when SERVO ON is requested. If K4 energises but the contactors do not, the fault is downstream.
3. Measure contactor coil voltage. With the controller in Servo On Request state, measure across each main contactor coil. Expected: 24 V DC (or 110 V AC on 110 V cabinets) present when K4 closes, 0 V when K4 releases. If coil voltage is present but the contactor does not close, the contactor is mechanically failed (clean the AUX block with lint-free paper, never emery).
4. Check the DEADMAN / CN8 jumper. CN8 on the MSP board is the deadman loop. With no deadman switch fitted, pins 1–2 (or 2–3, per wiring revision) must be jumpered. An open circuit here will not necessarily post 5390 — it usually posts a separate deadman alarm — but it will still block the chain.
5. Locate the safety relay (SRB-NA-R-C15-24V). The relay is typically a 24 V DC solid-state safety relay in a DIN-rail socket. Identify the X1 / X2 output terminals. These are wired in series with the contactor coil return. Measure continuity between X1 and X2 in the relay's run state. If open, the relay is tripped or de-energised.
6. Determine why the safety relay is open. Trace the loop into the safety relay's input channels. Look for removed light curtains, disconnected E-stop contacts, or a missing safety-PLC ribbon. The relay's internal LED will show the tripped channel on most SRB-NA revisions.
7. Verify the dedicated inputs 4077 and 4087. From the same DISP → F5 (DIAG/APL) menu, scroll to I/O → SPECIFIC INPUT and observe signals 4077 (servo ready) and 4087 (servo on acknowledge). A forced-off state on either pin suppresses the chain. If a pin is forced off, the controller is intentionally blocking the request — back-track to find which interlock is mapped to that bit.
8. Attempt the SD parameter workaround only as a last resort. Some forum references suggest setting S2D0001 = 255 and S2D0002 = 255 in the customer menu. These parameters are write-protected at customer level on production firmware; changing them requires the maintenance-mode password and may void the safety rating of the cabinet. Do not use this to mask a real enable-chain fault.

6. Confirmed Field Resolution

The reported SK6 was recovered by jumpering the X1 and X2 output terminals of the SRB-NA-R-C15-24V safety relay while the controller was in maintenance mode. After the jumper was installed, the main contactors pulled in normally, the SUMX LED latched, and 5390 cleared. The cabinet was then taken back to maintenance mode and the safety loop was rebuilt with the originally intended external safety devices, after which the jumper was removed.

Procedure summary:

1. Power down, lock out / tag out the cabinet mains disconnect.
2. Open the safety relay's DIN-rail socket or remove the relay entirely.
3. Install an insulated test jumper between X1 and X2.
4. Restore power, enter maintenance mode, press SERVO ON.
5. Confirm both main contactors close (audible click) and the SUMX LED stays lit.
6. Cycle the controller and verify no 5390 returns.
7. Before commissioning: rebuild the safety loop to the original design intent and remove the jumper.

7. Cross-Reference: 4077 / 4087 Dedicated Inputs

On the MRC II SK6 firmware (SK6-C00), the MIF01 board decodes several dedicated inputs that the executive consults before allowing servo power. The two most relevant to 5390 are:

Access the live state via DISP → F5 (DIAG) → I/O → SPECIFIC. If 4087 reads OFF while K4 is energised, the open circuit is between the contactor AUX and the MIF01 input — usually the safety relay's X1/X2 contact.

8. Servo Power-On Sequence Timing (MRC II, SK6-C00)

The approximate watchdog timing on a healthy SK6-C00 cabinet, measured from SERVO ON press to contactor close, is:

If 5390 is posted before 200 ms, the contactor never closed (Section 5, steps 2–6 apply). If 5390 is posted after 250 ms, suspect the amplifier link or an SVAC/SVDB internal fault, not the enable chain.

9. Verification After Repair

1. Cycle power three times; SERVO ON must succeed on every attempt.
2. Move each axis at 5% speed, then 100% speed; monitor for any 4077 drop-outs on the MIF01 dedicated input screen.
3. Trigger every E-stop in the cell; confirm contactors drop and 5390 (or its E-stop equivalent) posts cleanly.
4. Reset; confirm SERVO ON recovers only after all E-stops are released.
5. Restore the safety devices and remove any diagnostic jumper; re-test the safety chain end-to-end.
6. Update the controller's machine history log with the MMM02 version, the SRB part number, and the date.

10. Troubleshooting Matrix

11. Field-Proven Caveats

• Second-hand cabinets frequently ship with orphaned safety wiring. The SRB-NA-R-C15-24V (or equivalent safety relay) was wired to devices that the new owner did not receive. Always check for unterminated conductors in the safety loop before suspecting the controller.
• MMM02 piggy-back version must match the manipulator. A controller from an SK6 with SK6-C00 firmware cannot be moved onto an SK16 or SK45 without re-flashing the MMM02 and re-teaching the payload. Mismatches typically surface as dedicated-input faults, not 5390, but they can mask as 5390 if the executive is checking payload-derived parameters before closing the contactors.
• The S2D0001 / S2D0002 = 255 workaround is not a fix. These parameters appear to mask dedicated-input checks. On production firmware they are password-protected for a reason. Use only as a last resort to confirm a hypothesis, and document the change in the controller's service log.
• For related controllers (NX100, DX200), alarm codes differ. An NX100 will post ALARM 0021 for a similar root cause, and the diagnostic flow uses the CX-series SERVOPACK alarms rather than the MRC II's dedicated-input screen. Do not cross-apply parameter numbers between generations.
• A buzzing sound from the contactor bay during 5390 usually indicates that the contactor coil is being pulsed by the MTU01 watchdog while the AUX feedback remains open. Investigate the enable chain before replacing the contactor.

12. Frequently Asked Questions