Hello! I'm happy to help you with any questions about our service offerings.
Yaskawa Converter CIMR-MR5N2015 Hauptbild
13.10.2025 by Viktor Siebert
Repair of a Yaskawa CIMR-MR5N2015 Converter

Entry and Initial Assessment.

Sometimes a device arrives that, at first glance, doesn’t look like it could still function. But as soon as you open it, you see how much improvisation and necessity are part of industrial processes. The Yaskawa converter CIMR-MR5N2015 that came to us from Norway was exactly that kind of case.

The customer gave us a short but very precise error description:
“When I run up the spindle slowly, it works. If I speed it up fast, it fails.”
A sentence that says a lot when you read it the right way. In practice, this means: the current flow in the power stage is unstable at low rpm still within tolerances, at fast acceleration it exceeds the limits of the IGBTs. That suggests a thermal or structural damage in the power stage.

Visual Inspection and First Findings

Already when unpacking, we noticed the typical smell: slightly sweet, burnt an unmistakable sign that something inside had overheated. When we opened it, the suspicion was confirmed. The transistor module, the heart of the power stage, was partially burned. The control board above it showed clear discolorations and carbonization, especially in areas where the heat had accumulated. Despite these serious damages, the device according to the customer, had been running for months. He had no choice, a critical job, ongoing production, no spare device available. A scenario we know all too well in industry.

Risks from Thermal Damage

What many don’t realize: continuing operation under these conditions is an enormous risk. The converter could have short-circuited at any moment not quietly, but with a loud bang, smoke, flames, and a large arc fault. The adjacent PCBs and cable assemblies would also have been affected. Fortunately, that didn’t happen. But the signs were evident: the solder mask on the control board was blistered in several spots, the copper traces slightly charred, and the IGBT body underneath showed black inclusions.

Disassembly and Root-Cause Analysis

We decided to fully disassemble the unit. Every screw, every component was individually inspected, cleaned, and documented. The burnt transistor module was removed, its terminals had practically fused. Such damage often arises when there are long-term contact resistances, e.g. due to loose screws, material fatigue or fine cracks in solder joints. The electric current generates heat, which in turn worsens the contact further, a creeping vicious circle.

The control board above it was still functionally intact on the surface, but the discolorations allowed no peace of mind. We conducted a thermal stress analysis: Several zones had reached temperatures above 100 °C values that drastically shorten the lifespan of PCBs and components. The risk of later failure was simply too high. So we decided to replace that board completely, even though it initially “still worked.” That is a point where experience and responsibility show: just because a component still functions doesn’t mean it’s safe.

Cleaning and Reassembly

In parallel, the unit was thoroughly cleaned. All soot, dust, and old thermal compound residues were removed. The heat sinks were planed and refilled with fresh thermal interface material to ensure ideal heat conduction. The DC-link capacitors were tested, insulation values measured, and the gate driver circuits of the new transistor module were re-calibrated.

This part of the work may seem inconspicuous, but it is decisive: only through meticulous cleaning and restoration of thermal interfaces can a unit run reliably later. Especially for converters like the CIMR-MR5N2015, which operate under high continuous load, such details determine the lifetime.

Test Run and Measurement Results

For the final functional test, the converter went to our Yaskawa test bench designed for such high-performance devices. Here voltage, current, temperature, and signal quality are measured under real load conditions over several hours. On the first commissioning, everything ran stably. After about 30 minutes, the converter was exposed to load peaks, its behavior remained consistently clean. No voltage drops, no current distortion, no thermal anomalies. The new construction performed exactly within the spec range.

During the final measurement, it became clear how important proactive overhaul is. The new transistor module operated at a significantly lower operating temperature, which not only extends its lifespan but also enhances the overall system stability. That confirmed that the decision was correct: not just replacing the failed module, but also substituting the thermally stressed board.

The Human Behind the Machine

What makes this case special is not only the technical aspect, but also the human side. The customer had improvised for weeks because he lacked a quick solution. He knew the risk was there and sought help only when it was nearly too late. This is everyday reality in many operations: production pressure, missing spares, shortage of specialists all factors that lead to running devices you know are dangerous. That’s precisely why specialized repair shops like ours exist. We understand not only the electronics, but also the situation behind them.

Conclusion and Recommendations

After successful repair, cleaning, and testing, the converter was reassembled and prepared for return shipment. We provided the customer with a summary report and maintenance recommendations: periodic checks of fans, cleaning of cooling paths, and a thermal inspection every 12 months to prevent such thermal damages from ever occurring.

This case vividly shows how important the combination of experience, technical methodology, and responsibility is. A transistor module can be replaced trust and operational safety must be earned. For us, this was not an ordinary repair, but a lesson in how narrow the line between function and disaster sometimes is. In the end, there is a working device, a satisfied customer and the assurance that we kept a piece of industry running.

Further information such as price and delivery time for:


Yaskawa CIMR-MR5N2015 Converter

Yaskawa Inverter CIMR-M5N2015

More details about our Yaskawa repair expertise can be found here: Yaskawa Reparatur bei Industrypart

📞 Please feel free to contact us if you have any questions regarding your Omron drive technology. Our experienced team is always available to assist you.

Device Data

Model: Yaskawa CIMR-MR5N2015
Input Voltage: AC 3-Phase 200–220 V 50 Hz / 200–230 V 60 Hz
Input Current: 39.3 A (50 Hz cont.) / 53.6 A (60 Hz 30 min)
Output Voltage: DC 270–325 V
Output Current: 52.4 A cont. / 71.4 A 30 min
Weight: 17 kg
Manufacturer: Yaskawa Electric Corporation (Japan)

Operating Environment & Compatible Devices

The Yaskawa CIMR-MR5N2015 is part of the VARISPEED-626M5 / 656MR5 system, which consists of an Inverter (CIMR-M5) and a Converter (CIMR-MR5).
This system was specifically developed for machine tool spindle drives, designed to deliver high torque density, stable speed control, and regenerative power feedback to the grid.

Typical applications include:

  • CNC machining centers with integrated spindle and axis drives
  • High-speed spindles for precision metal cutting and hard material processing
  • Systems with braking and regenerative operation for improved energy efficiency

Compatible inverter and motor series:

  • Spindle Inverters: Yaskawa CIMR-M5A, CIMR-M5N, CIMR-M5D-….
  • Spindle Motors: Yaskawa UAASKA-… and UAASKB-… series (Standard and Winding-Selection types, 200 V / 400 V classes)
  • Encoders: UTMSI-10AAGAZA, UTMSI-10AAGBZA
  • Control environments: Stand-alone Yaskawa systems or customized CNC controllers using analog or digital speed reference signals (e.g. ±10 V or 12-bit digital input)

The overall system forms a closed Yaskawa spindle drive architecture, where the converter and inverter are connected via a DC link.
Regenerative energy from the spindle is fed back through the MR5 converter into the power supply, ensuring high energy efficiency and thermal stability during continuous operation.

Converter Alarm Table (Fault Codes and Corrective Actions)

Fault No.Alarm NameDescriptionPossible CauseCorrective Action
01OvercurrentOutput current exceeded detection levelShorted load, incorrect wiring, overvoltage on inputCheck load, wiring, and input voltage
04Main Circuit Fuse BlownFuse in main power circuit has openedDamaged transistor or ground shortReplace damaged components and fuse
05OverloadOutput current exceeded overload limitLoad too high, improper capacity selectionReduce load or verify converter sizing
11Output OvervoltageDC bus voltage exceeded limit (approx. 400 V for 200 V class)Input surge, regenerative voltage from inverterCheck input voltage and motor deceleration
12Main Circuit UndervoltageDC voltage below minimum thresholdLow supply voltageCheck supply voltage and connections
14Control Circuit UndervoltageControl power below thresholdLow or unstable control voltageVerify control power source
15Power Supply Frequency FaultSupply frequency outside ±5 % of 50/60 HzPower instability or phase issueCheck power waveform and line conditions
16Initial Charging FaultDC link capacitors not charged in timeCapacitor aging or rectifier issueReplace converter or repair charge circuit
17Power Supply Open PhaseMissing input phase detectedFaulty wiring or fuseCheck phase connection and input lines
23Built-in MC Operation FaultMagnetic contactor failed to operateCoil or driver failureReplace magnetic contactor or unit
43Heatsink Overheat 1Temperature exceeded upper warning limitBlocked air path or fan slowdownClean cooling ducts, check fan
44Heatsink Overheat 2Temperature exceeded for more than 1 minCooling failureVerify ventilation and ambient temperature
45Heatsink Thermistor DisconnectionSensor disconnected or faultyDamaged thermistor or wiringReplace thermistor or repair wiring
46Control PCB Overtemperature 1Control board > 80 °CInadequate airflowImprove cooling and environment
47Control PCB Overtemperature 2Control board > 85 °CProlonged heat exposureEnsure ventilation, clean interior
50CPU / A-D Converter ErrorInternal CPU or ADC faultControl board failureReplace control PCB
52EEPROM / ROM ErrorMemory read/write faultAging or corrupted memoryReplace control PCB
53WDT Time ExceededWatchdog timer errorSoftware or hardware lock-upRestart, if repeated replace control PCB

Components (Main Internal Parts)

SectionDescription / Part NameBoard Code / ModelFunctionNotes
Control BoardMain Control PCBYPHT31149-xDControls converter logic, protection, and communication with inverterMounted on top of power section
Power BoardIGBT Power PCBYPCT31244-1B / ETC626095Drives IGBT transistors, manages power stage switchingThermally sensitive component
Power ModuleTransistor / IGBT ModuleConverts AC input to DC output via rectification and switchingSubject to thermal load
DC Link CapacitorsElectrolytic CapacitorsStabilize DC bus voltage and filter ripplePeriodic inspection recommended
Heatsink AssemblyCooling SystemTransfers heat from power stage to air flowRequires functional fan and clean surface
Connector Interface1CN / 5CN / 51CN / 52CNSignal and power interface to inverter and control circuitsEnsure firm seating and clean contacts
Fan UnitCooling FanEnsures forced air cooling for IGBT and control sectionReplace every 2 years or 6000 h

We are happy to assist you with any questions and inquiries.

Feel free to contact us anytime through our social media channels.

Get in touch with us

Mandatory fields are marked with (*)