Key takeaways
A Mitsubishi control that stops with an alarm gives you a family code, a detail number, and usually an axis name, and that combination tells you most of the diagnostic story before you open a single cabinet. This guide is for maintenance technicians, maintenance managers, and plant engineers who need to read those codes correctly on M700/M800 series and MELDAS-era machines, run the right first checks, and reset safely.
Mitsubishi structures alarms by source, not by severity alone. Servo and spindle drive faults report as S-series alarms with a detail number identifying the specific drive-side failure. System-level problems, including absolute position errors, report as Z-series alarms. Communication faults between the NC and the drive units report as Y-series alarms, operation errors report as M01 with a sub-number, and anything defined by the machine builder appears as a PLC or user alarm.
One caution before the details: exact code meanings shift between control generations and drive families (MDS-A through MDS-E and beyond). The examples below are commonly documented patterns, but the maintenance manual for your specific control is the final authority. If a code here does not match your screen, trust the manual.
| Display (typical) | What it usually means | Likely cause | First check |
|---|---|---|---|
| S01, detail 0032 | Power module overcurrent | Short or ground fault in motor or power cable | Insulation test by a qualified electrician, after lockout/tagout |
| S03, detail 0052 | Excessive position error | Mechanical binding, crash, brake not releasing | Jog the axis slowly and watch the load display |
| Z70 | Absolute position data lost or illegal | Battery failure, encoder fault, interrupted initialization | Check battery voltage and encoder cabling, then re-establish zero |
| Z73 | Battery voltage drop warning | Encoder backup battery near end of life | Replace the battery promptly, with control power ON |
| Y02 / Y03 | NC-to-drive communication or drive not detected | Cable, connector, noise, or a dead drive unit | Reseat communication cables, check drive LEDs |
| M01, detail 0006 | Hardware stroke end (overtravel) | Axis sitting on a limit switch | Jog off the limit in the opposite direction |
The S family reports faults detected by the servo drive itself. On most generations the family indicates the reset class: S01 alarms require the control power to be cycled, S02 flags an initial parameter error at startup, and S03 alarms can be cleared with the reset key once the cause is gone. The detail number identifies the actual failure, and it is the part worth writing down before you touch anything.
Commonly documented detail numbers include:
Safety first: anything beyond reading the screen means lockout/tagout. Drive cabinets hold charge in the DC bus capacitors after power off, so wait the time stated in the manual and verify with a meter before touching terminals. Support vertical axes mechanically before releasing a brake, since gravity is stored energy too, and never bypass door switches or interlocks to speed up testing. If you also maintain Fanuc machines, the same servo-ready discipline applies to Fanuc alarm 401 (VRDY OFF), which is that platform's equivalent servo drive readiness fault.
Spindle drive faults use the same S-series display with their own detail numbers, because on Mitsubishi platforms the spindle drive is a sibling of the servo drives. Three patterns cover most real-world events:
Y-series alarms report problems between the NC and the drive network. Y02 typically flags a data transfer error to a drive, and Y03 typically means a configured drive unit was not detected at startup. First checks: drive status LEDs, communication cable seating at both ends, cable routing away from noise sources, and drive control power. A Y03 after maintenance work very often turns out to be one connector left loose.
Z-series alarms are system-level. Some relate to data and memory, but the ones technicians see most are the absolute position group, which deserves its own section.
Mitsubishi axes with absolute encoders keep their position memory alive through a backup battery while the machine is powered off. When that battery fades, you get one of the most common real-world Mitsubishi fault sequences:
The correct response to Z73 is to replace the battery with control power ON, so the encoder stays powered from the control while the battery is out of circuit. Wait for a Z70 and you have added a zero-point initialization to what should have been a five-minute battery swap. After any Z70 or Z71, fix the cause (battery, encoder cable, or encoder), then re-establish the zero point per the maintenance manual and verify position against a known reference before cutting parts. If alarms persist with a fresh battery, test the detector itself; our guide to encoder failure symptoms and testing walks through cable, connector, and signal checks in order.
The prevention lesson is boring and effective: replace encoder batteries on the interval your manual specifies, not when the warning appears, and always before long planned shutdowns.
M01 alarms are not hardware failures; they mean the operator or program asked for something the control will not allow right now. Commonly documented sub-numbers include 0006 (hardware stroke end, axis on an overtravel limit), 0007 (software stroke end, commanded position outside the soft limits), and 0101 (no operation mode selected, often a faulty mode switch or PLC condition). Interlock-related sub-numbers in the same series mean a PLC condition is holding the axis. These clear as soon as the condition clears: jog off the limit, correct the program, or restore the mode selection.
PLC and user alarms are written by the machine builder, not Mitsubishi, so the same number can mean different things on two machines standing side by side. Look them up in the machine tool builder's manual, and check simple physical causes first: lubrication level and pressure switches, air pressure, chip conveyor overload, and door or magazine sensors are the usual suspects.
A single Z70 after a long shutdown is a battery story. A Z70 every month is an engineering story, and you can only tell the difference if the data exists. Log every alarm stop as a downtime event with the full code (family, detail number, axis) as the cause code, then review MTBF and MTTR per asset. Falling MTBF on one axis flags a developing mechanical or encoder problem while it is still cheap; high MTTR flags missing spares, missing documentation, or missing training.
Those alarm minutes also belong in your availability numbers. Repeated short servo and interlock stops quietly erode OEE long before anyone schedules a repair, so counting them accurately is what turns a recurring annoyance into a justified engineering fix.
Reading codes well gets a machine back up; making sure the same code stops recurring requires a system that connects detection to execution. Fabrico is computer-vision-verified OEE plus closed-loop maintenance execution: cameras catch stops and micro-stops that manual logs and sensors miss, and maintenance work orders close the loop from detection to fix. If your Mitsubishi machines generate more alarm history than answers, book a Fabrico demo and see the full detection-to-fix loop on your own downtime data.
Z70 indicates the absolute position data for an axis is lost or invalid, most often because the encoder backup battery failed while the machine was powered off. Fix the cause, then re-establish the zero point per your maintenance manual before running production.
Replace the encoder backup battery with control power ON, check the encoder cable and connectors, then perform the absolute position initialization procedure for your control generation. The alarm will not clear meaningfully until position data is re-established and verified.
Generally no. On most Mitsubishi generations S01 is a power-reset class alarm, so the control power must be cycled after the cause is corrected. S03 class alarms usually clear with the reset key. Confirm the reset class for your specific code in the maintenance manual.
The encoder backup battery is at end of life and can no longer hold the encoder memory through a powered-off period. Replace it promptly with control power ON, and put battery replacement on a fixed preventive maintenance interval so it stops recurring.
In the machine tool builder's documentation, not the Mitsubishi manual. PLC alarms are defined per machine by the builder, so the same number can mean different things on different machines. Check simple physical conditions such as lube level, air pressure, and door sensors first.
Anmeldung
Blog
