Yasnac MX2 E-Stop and Servo Power Off After Phase Converter Swap

Problem Overview

A Matsuura MC-500V2 vertical machining center equipped with a Yasnac MX2 CNC control was running on a Phasamatic R15 rotary phase converter (RPC) that replaced an earlier 10 HP unit. After the swap, the machine booted, homed the X and Z axes successfully, then began dropping servo power during the Y-axis homing sequence. Each time the operator pushed the SERVO ON button on the operator panel, the drives re-enabled for several seconds before tripping again. When the operator finally opened the power cabinet to measure line voltages, one of the 120 V single-phase input conductors feeding the RPC input block had slipped out of its mechanical lug, leaving the converter running open-phase and back-feeding a single-phase condition onto the machine's three-phase buss.

After the loose lug was reseated and torque-checked, the machine powers up but now latches in two simultaneous alarms:

• EMERGENCY STOP on the Yasnac MX2 status line
• SERVO POWER OFF on the Yasnac MX2 status line

Both alarms persist with the E-stop pushbutton released, the overtravel release switch actuated, and all axis covers closed. The SERVO ON button on the operator panel will not latch the contactor. This article walks through the cascaded faults produced by open-phase operation, the alarm/signal logic inside the Yasnac MX2, and a field-proven recovery procedure for a machine that has been single-phased.

Why Single-Phasing a CNC Cascades Into Multiple Alarms

An RPC such as the Phasamatic R15 generates its idler leg from the two single-phase input legs. When one of those input legs opens:

• The idler leg collapses or runs at reduced voltage, leaving the load side effectively single-phase.
• Three-phase induction loads (spindle motor, servos, fans, transformers) see only one phase. Current in the remaining two phases doubles to maintain torque.
• DC bus capacitors in the servo amplifiers discharge asymmetrically, and the regenerative path that normally returns energy during deceleration now sources from the input rectifier under fault conditions.
• Within 100-500 ms, drive overcurrent, DC bus undervoltage, and open-phase detection all fire.

On a Yasnac MX2, the visible result is the alarm pair above, but the underlying faults can be:

The key diagnostic principle: each sub-assembly has its own protective device. Until you walk the chain from the wall to the contactor to the amplifier input, you cannot determine which device actually tripped.

Yasnac MX2 E-Stop and Servo-Ready Logic

The Yasnac MX2 (Yaskawa J50 / MRC controller family, typical on 1988-1994 Matsuura VMCs) uses a hardwired series E-stop loop and a separate Servo-Ready loop. Both are required for the SERVO ON button to latch the main contactor.

E-stop loop series string (typical Matsuura MC-500V2 wiring):

1. Main E-stop pushbutton (red mushroom, NC contact)
2. Secondary E-stop on operator pendant or side of guard
3. Door interlocks (guard switches, NC)
4. Each axis overtravel limit (X+, X-, Y+, Y-, Z+, Z-) — NC, opens on overtravel
5. Spindle drive READY contact (closes only when spindle drive is healthy)
6. Servo amplifier READY contacts (one per axis, wired in series, close when no drive alarm)
7. Back to CPS 24 V return

Any single open contact breaks the loop and forces EMERGENCY STOP on the status display. The MX2 does not differentiate which link is open without further diagnostics — you must measure continuity through the chain.

Servo power latch path:

1. Operator presses SERVO ON
2. CPS receives command, energizes servo main contactor coil
3. Contactor closes, applies 3-phase to servo amplifier inputs
4. Each amplifier DC bus charges, control power comes up
5. Amplifiers close their READY contacts in series with the E-stop loop
6. Yasnac MX2 displays SERVO ON only when the loop is fully closed

If any amplifier is in a latched alarm from the single-phasing event, its READY contact stays open, the loop stays open, and the contactor cannot re-latch until the amplifier alarm is reset.

Initial Safety and Lockout

1. Open the main disconnect at the wall. Apply personal lock and tag.
2. Verify zero AC voltage L1-L2, L2-L3, L3-L1 at the RPC input.
3. Wait five minutes for capacitor bleed, then measure DC bus at each amplifier (typically B and N terminals on CACR drives).
4. Open the cabinet door and visually inspect for discolored conductors, blown fuse cartridges, or popped breaker buttons before any further handling.

Step 1 — Re-Verify the Phase Converter Output

Before assuming the machine is the fault, verify the RPC is actually producing a healthy three-phase output. A Phasamatic R15 should produce idler-leg voltage within about 5-10 % of the generated legs under load.

Measure with the load connected and again with the load disconnected. A healthy RPC idler leg may show normal voltage at no-load but collapse as soon as motor contactors close. If the leg collapses under load, the start capacitors or run capacitors inside the R15 have likely been stressed by the fault and need inspection per the Phasamatic manual.

Step 2 — Inspect Branch Protection Devices

Walk the cabinet top-to-bottom with the panel cover off and the disconnect off. Document every breaker and fuse in the path:

1. Main branch breaker feeding the cabinet. Should not have tripped if the upstream RPC disconnect opened first, but check anyway.
2. CPS input breaker / fuse. The Composite Power Supply typically has its own dedicated 5-10 A breaker. A single-phasing event often pops this first because the CPS input transformer sees the most asymmetric saturation.
3. Per-axis servo amplifier input breakers/fuses. Yaskawa CACR drives on the MX2 are usually fed by 10-20 A breakers (X, Y, Z, plus a fourth for the spindle if separate). Each has its own protection — check each independently.
4. Spindle drive input protection. Often a 30-50 A breaker on a 7.5-15 kW spindle. Single-phasing a loaded spindle motor will trip this instantly.
5. Coolant, chip conveyor, lubrication pump breakers. Sometimes stacked on a multi-pole breaker on the right side of the cabinet.

Replace any blown fuse with the exact type and rating — do not up-rate. A blown fuse is evidence, not the fault.

Step 3 — Walk the E-Stop Loop with a Meter

With power still off, use a continuity tester or ohmmeter to walk the E-stop loop end-to-end. The loop is normally closed when healthy and will read open until the fault link is found.

1. Locate the E-stop terminal strip inside the cabinet (typically labeled ES1, ES2 or XE1, XE2 on Matsuura wiring diagrams).
2. Place one meter lead on the 24 V return side of the CPS, the other on the input side of the E-stop chain.
3. Expected healthy reading: 0-2 Ω with the E-stop button released and all doors closed.
4. If open, isolate segments by lifting wires one at a time to bisect the loop.
5. Common failure points after a phase-loss event:
   • A servo amplifier READY contact that has welded or latched open
   • A spindle drive READY contact open due to a phase-loss alarm (Yaskawa CACR spindle drives throw alarm F1 or F2 on phase loss)
   • A limit switch that was actually hit during the fault — remember the X and Z axes homed, but the Y-axis was mid-sequence when power dropped; it may have moved slightly past the home dog into the overtravel switch
   \li>The E-stop pushbutton itself, which often takes mechanical damage during panic presses

Step 4 — Read Amplifier Alarms on Each Axis

Each CACR servo amplifier has a 7-segment LED or a row of LEDs that displays the active alarm code. With the cabinet powered (after the E-stop loop is closed) and the main contactor in, read the LEDs on X, Y, and Z drives.

Alarm reset procedure on CACR drives:

1. Cycle control power (24 V to the drive) by turning off then on the drive's control-power breaker.
2. Or, momentarily short the alarm-clear terminal to common — refer to the specific drive's manual for terminal numbers (commonly CN1 pin or dedicated ALM-RST terminal on Yaskawa SGDB series).
3. If the alarm re-appears immediately, the fault is still present — do not keep cycling.

For Matsuura MC-500V2 documentation, refer to the Yasnac MX2 Operator's Manual and Maintenance Manual for the specific alarm-clear sequence used on that control. Matsuura also publishes machine-specific electrical diagrams labeled MC-500V2 in their service documentation set.

Step 5 — Check the Composite Power Supply (CPS)

The CPS unit on Yasnac MX2 takes the three-phase input (typically 200 V or 220 V class, sometimes stepped down by an autotransformer) and produces the DC rails used by the control:

• +5 V DC — logic, CPU, memory
• ±12 V DC — analog servo reference, position feedback
• +24 V DC — I/O, relays, contactor coils, E-stop loop, servo brake release

Open the CPS (usually a card-cage module at the bottom or side of the cabinet). Look for:

• Blown input fuse on the AC side
• Popped crowbar or MOV across any output rail
• Discolored PCB or transformer smell
• DC OK LED extinguished (most CPS modules have one green LED per rail)

Measure each rail under no load and under load (with all drives connected). Spec limits per Yaskawa MX2 service manual are typically ±5 % on +5 V, ±10 % on ±12 V, and +20/-10 % on +24 V. Ripple must stay below 100 mV pk-pk on +5 V or the CNC will throw a memory parity alarm.

A common failure mode after single-phasing is one of the internal rectifiers shorts. The CPS appears to power up but the 24 V rail sags under load, dropping the E-stop relay below its dropout voltage (typically 19-21 V on 24 V relays). The E-stop loop appears to close on a meter but cannot sustain the relay current in operation.

Step 6 — Inspect the Main Servo Contactor

The main servo contactor (typically a 3-pole, 30-50 A contactor with a 24 V or 110 V coil) is the single point that energizes all three servo amplifiers simultaneously. If the contactor does not seat fully on a fresh SERVO ON command:

1. Measure coil voltage at the contactor coil terminals during the SERVO ON button press. Expected: nominal coil voltage ±10 %.
2. If coil voltage is present but the contactor does not seat: replace the contactor. Coil may be marginal or the armature may be mechanically worn.
3. If coil voltage is absent: trace back to the CPS 24 V rail, the SERVO ON pushbutton, and any interposing relays.
4. With power off, manually depress the contactor armature. It should snap in cleanly with a definite click. If it feels gritty or hesitates, the contactor has degraded — replace.
5. Inspect the contactor tips. Pitted or welded tips from the single-phasing surge will give high resistance when closed, producing the "contact closes but drives do not power up" symptom.

Step 7 — Verify Servo Brake Release

Yaskawa servo motors with electromagnetic brakes (typically the "-B" suffix on a SGM or SGMP motor) hold the motor shaft locked until 24 V DC is applied to the brake release coil. The brake release circuit on the Yasnac MX2 is fed from the same CPS 24 V rail through a brake-release contactor or solid-state switch.

1. Listen for the brake "click" when SERVO ON is pressed. All axis brakes should release within 200 ms of the contactor closing.
2. If one axis brake does not release, measure the DC voltage at the brake terminals on that motor. Expected: 24 V DC ±10 % with SERVO ON.
3. If 0 V at the brake: trace back through the brake-release relay. The relay coil is fed from CPS 24 V; if the CPS is sagging, the relay may not pick up.
4. If 24 V present at the brake but brake does not release: the brake coil is open or the brake has mechanically seized. Disconnect the motor leads and measure brake coil resistance (typically 50-100 Ω on a 24 V Yaskawa brake). Open = replace brake; Yaskawa brakes are field-replaceable on most SG-series motors.

A motor driven with the brake engaged will pull locked-rotor current on two phases and single-phase itself. This is a common secondary failure mode where a single-phasing event at the wall turns into a second single-phasing event at the motor if the brake release failed silently.

Step 8 — Reset and Power-Up Sequence

Once all branch protection is verified and the E-stop loop is closed through the chain, perform the controlled power-up:

1. Open the main disconnect.
2. Verify all breakers in the cabinet are in the ON position, fuses are intact.
3. Verify E-stop button is released (pulled out / rotated to release).
4. Verify no axis is sitting on an overtravel limit. Jog any axis that is on a limit off the limit using the OT Release sequence described in the MX2 manual.
5. Close the main disconnect. The Yasnac MX2 should boot to its menu within 30-45 seconds.
6. Confirm the status line reads READY or NOT READY without EMERGENCY STOP.
7. Press SERVO ON. The main contactor should close and the status line should switch to SERVO ON.
8. Read each amplifier's 7-segment display — all should read 0 or RD (Ready) within 2 seconds of contactor closure.
9. Jog each axis at low feedrate (F0 = 100 mm/min or equivalent) to confirm motion and brake release.

Verification Tests Before Returning to Production

Prevention — Phase Converter Swap Best Practices

A rotary phase converter swap is one of the most common causes of single-phasing damage to a CNC. The mechanical lug failure mode described in this case (a 120 V leg backfeeding into the converter input block) is also one of the most preventable.

• Torque every lug. Mechanical lugs on RPC input blocks need to be torqued per the manufacturer's spec, typically 35-50 in-lb for a #6 AWG to #2 AWG range. A "finger-tight" verification followed by a real torque-check after one hour of operation is the field-proven method.
• Use a pull-test. After torquing, tug each conductor firmly. A properly torqued lug will not allow any visible movement of the conductor in the barrel.
• Mark the wild leg. On 240 V RPC installations, the generated (wild) leg is typically L3. Mark it with red tape or heat-shrink so it can be visually distinguished during any future work.
• Install a phase monitor relay. A Yaskawa CMP-1 or equivalent phase-loss/phase-sequence relay upstream of the CNC cabinet will trip the cabinet main breaker on phase loss in under 200 ms — well before the servo drives see the undervoltage. This is the single most effective upgrade for any CNC on a rotary phase converter.
• Add a step-start or soft-charge. Some RPCs and many CNC cabinet contactor configurations include an inrush-limiting resistor or soft-charge contactor. Verify this is functioning — a healthy step-start limits inrush to 1.5-2× FLA instead of 6-8× FLA, reducing stress on lugs during every power-up.
• Record baseline voltages. Before any RPC work, record L1-L2, L2-L3, L3-L1 at the cabinet input under three load conditions: idle, spindle running unloaded, and full rapid on all three axes. File this with the machine documentation. Any future deviation is immediately visible.

Yasnac MX2 Quick Alarm Reference

The MX2 displays a two-line status message plus the SERVO ON / E-STOP indication. The amplifier-level alarms below appear on the CACR drive's 7-segment display. Cross-reference the amplifier code to the MX2 status line:

For the full MX2 alarm list and corrective action per alarm code, refer to the Yasnac MX2 Maintenance Manual (Yaskawa publication CMP-Y8002xxx family) and the Matsuura MC-500V2 Electrical Diagram set supplied with the machine. Yaskawa also publishes per-amplifier alarm lists in each CACR drive's User's Manual, which is the authoritative document for drive-level reset procedures.

When to Call Service

The recovery procedure above covers the typical single-phasing aftermath on a healthy Yasnac MX2. Escalate to a Yaskawa / Matsuura service technician if:

• Any amplifier display shows a code that does not clear after two control-power cycles with verified three-phase input.
• DC bus voltage at an amplifier does not reach rated value (typically 280-310 V DC for 200 V class, 560-620 V DC for 400 V class) within 2 seconds of contactor closure.
• The CPS shows internal damage (popped components, transformer smell).
• Any axis motor has seized or shows bearing damage from the locked-rotor / brake-engaged event.
• The spindle drive shows persistent F1/F2 alarms even with healthy input — this often indicates damaged IGBT modules, which require drive-level repair, not just a reset.