Metso DNA system troubleshooting

Metso DNA system troubleshooting

In modern mining, papermaking, electricity, and heavy industry, the continuous operation of rotating machinery is directly related to the availability and safety of factories. The Metso DNA machine condition monitoring system, as an integrated online monitoring solution, achieves early warning of mechanical faults such as bearing wear, imbalance, misalignment, and gear meshing problems through fixed installation of vibration sensors, temperature sensors, and process parameter collection. However, any complex automation system will encounter problems such as component aging, signal abnormalities, and configuration failures during long-term operation. This article is aimed at on-site maintenance engineers and condition monitoring experts, systematically reviewing the typical fault phenomena, diagnostic logic, and standard troubleshooting steps of the Metso DNA system (including ACN MR controller, ACN I/O M120, AIF8 analog input unit, RVT/TT-125 dual output sensor, and intelligent alarm processing module) to ensure the high availability of equipment and the reliability of predictive maintenance data.

Overview of System Core Components

Before delving deeper into troubleshooting, it is necessary to review the key hardware and software modules of the Metso DNA machine condition monitoring system:

ACN MR Controller: High performance rail mounted controller, supporting redundant configuration (1:1 redundancy), with 5 100 Mbit/s Ethernet ports, detachable SD card for boot parameter storage, control cycle as low as 5 ms, suitable for centralized or distributed I/O architecture.

ACN I/O M120 series: high-density I/O unit, supports hot swapping, channel isolation voltage of 1500/2200 VAC, provides 8-channel fast dynamic measurement unit (such as AIF8V/AIF8T analog input unit), with rich channel level diagnostic functions.

RVT/TT-125 Sensor: Dual output acceleration and temperature sensor, sensitivity of 100 mV/g, frequency response of 1-7000 Hz (± 10%), temperature measurement range of+2~+120 ° C (10 mV/° C), using 316L stainless steel housing and M8 integrated bolt installation.

Analysis tools and intelligent alarm processing: Spectrum analysis, time-domain waveform, envelope demodulation, Bode plot, trend plot and other tools embedded in the Metso DNA Operate interface, as well as dynamic alarm curve (Notice Curve) function based on speed changes.

Common types of faults and diagnostic principles

Based on on-site experience, the main faults of the Metso DNA system can be classified into the following categories:

Sensor and signal link failures: including no output from acceleration sensor, abnormal bias voltage, temperature reading drift, cable breakage or connector corrosion.

I/O module channel failure: For example, a channel of the AIF8V unit cannot measure the 0-24V signal, the AIF8T unit cannot trigger the RTS-226 speed sensor, and there is crosstalk between channels.

Controller redundancy failure or startup failure: The main controller crashes and the backup controller does not automatically take over; SD card damage causes nodes to be unable to load real-time operating systems or application configurations.

Analysis tool data anomalies: unexplained phantom frequencies in the spectrum, loss of historical data in trend charts, inability to update waterfall charts.

Intelligent alarm false alarm or missed alarm: Due to the large range of device speed changes, the fixed alarm limit is no longer applicable; The notification curve is not running correctly or the parameter class division is unreasonable.

The basic principle of diagnosis is to isolate the sensor to the controller step by step: first check the sensor power supply and raw signals, then verify the hardware status of the I/O channel, confirm the controller communication and database configuration, and finally check the analysis parameters and alarm logic of the application layer.

Detailed troubleshooting steps

4.1 RVT/TT-125 sensor signal loss or excessive noise

Phenomenon: On the Metso DNA operator screen, the vibration amplitude of a certain measuring point remains zero or fluctuates violently, and the temperature value displays -50 ° C or exceeds the range.

Troubleshooting steps:

Check power supply: RVT/TT-125 requires 18-30 VDC power supply, with a typical operating current of 2-10 mA. Use a multimeter to measure the voltage between sensor pin A (signal/power) and pin B (common terminal), which should be the nominal bias voltage of 12 VDC (when the sensor is working normally). If the voltage is 0 or below 12 V, check if the upstream I/O unit (such as AIF8V) provides a 4 mA constant current source - note that each channel of AIF8V has a built-in 4 mA ± 0.1% constant current source for powering the acceleration sensor.

Measure output impedance: Disconnect the sensor from the I/O unit and measure the DC resistance between pins A and B at the sensor end. Normally, it should be around 100 Ω (maximum output impedance). If it is an open or short circuit, it indicates that the sensor is internally damaged.

Vibration signal simulation test: Use a handheld vibration calibrator (such as 100 Hz, 1 g) to directly excite the sensor, while observing the AC signal at pin A with an oscilloscope. The sensitivity should be 100 mV/g. If the deviation exceeds ± 5% (at 25 ° C) or the frequency response is abnormal (such as severe attenuation above 1 kHz), the sensor needs to be replaced.

Check connectors and cables: RVT/TT-125 uses MIL-C-5015 three core connectors, and common faults are pin corrosion or solder joint breakage. Re tighten the connector and replace the prefabricated cable if necessary. For long-distance transmission, ensure that the cable shielding layer is grounded at a single point on the I/O side.

Temperature output verification: Measure the DC voltage between pins C and B, which should be 10 mV/° C (0 ° C corresponds to 0 V? The actual sensor provides a range of 2~120 ° C, for example, about 200 mV at 20 ° C? Note that the specification sheet states "10 mV/° C" and the measurement range is+2~+120 ° C, so it should be around 250 mV at room temperature of 25 ° C. If the reading deviates significantly, a resistance box can be used to simulate temperature signals (requiring additional pull-up circuits) to determine whether it is a sensor or I/O channel issue.

Solution: If the sensor is confirmed to be damaged, replace it with the same model (RVT/TT-125, code 600-10026), and tighten the M8 bolts with a torque of 6 N · m during installation to avoid damaging the ceramic shear element due to over tightening. After replacement, the range needs to be re normalized in the channel corresponding to AIF8V (0~24 V corresponds to engineering units of g and ° C).

4.2 AIF8V/AIF8T units cannot be read or channel data is abnormal

Phenomenon: In the I/O diagnostic graph of DNA Operate, the AIF8V unit displays a red fault or a fixed channel value of -9999 (bad value). For AIF8T, the measured speed value is always zero.

Troubleshooting steps:

Utilizing hot plug recovery: ACN I/O M120 supports true hot plug. If a single channel fails, you can try unplugging the I/O unit and waiting for 10 seconds before reinserting it. The system will automatically reinitialize and restore communication. Attention: Before unplugging, it is necessary to confirm that the channel is not used for safety interlocking.

Check channel diagnostic information: Open the ACN I/O diagnostic interface in DNA Operate (see "ACN I/O diagnostics picture" on page 20 of the PDF). Each channel provides detailed diagnosis, including overload, disconnection, common mode voltage exceeding, etc. For AIF8V, the common diagnosis is "loss of on-site power supply" or "low load of constant current source".

Measure input impedance: The input impedance of AIF8V is 100 k Ω (voltage measurement mode). If the internal resistance of the on-site sensor or transmitter is too high, it will cause voltage division error. Disconnect the external wiring and use a multimeter to measure the resistance between terminal IN and COM, which should be around 100 k Ω. If it is close to 0, it indicates that the input circuit protection device has broken down; If it is infinitely large, it may be due to an open circuit in the sampling resistor.

AIF8T speed triggering issue: AIF8T is designed for synchronous sensors such as RTS-226, with an input impedance of 249 Ω and a channel specific current limiting power supply (30 mA). Use an oscilloscope to measure the trigger signal (which should be a square wave or sine wave with an amplitude of at least 5 Vpp). If there is no signal, check if the sensor is aligned with the gear or reflective tape; If there is a signal but it cannot be collected, try to adjust the trigger level and hysteresis value in the configuration of AIF8T. Additionally, please note that the measurement interval for AIF8T is 50 μ s (equivalent to a 20 kHz sampling rate), and longer acquisition times may be required for extremely low speeds (<1 Hz).

Check isolation performance: AIF8V and AIF8T channels are not isolated (shared COM), but 1500 VAC is isolated between channels and the system. If there is ground loop interference on site, the sensor shielding layer can be directly connected to the COM terminal, while ensuring that the COM is not connected to other ground on the I/O cabinet side.

Solution: If the diagnosis confirms a hardware failure of the unit, replace AIF8V (D201509) or AIF8T (D201510) with spare parts. After replacement, I/O addresses and range normalization parameters need to be reassigned in the Function Block CAD engineering tool (see "Analog RF and low pass filtering" settings on page 21 of the PDF). Attention: AIF8V supports two ranges of 0~24 VDC and -5~+5 VAC, which need to be selected through programming.

4.3 ACN MR Controller Redundancy Switching Failure or Startup Stuck

Phenomenon: After the main controller fails, the backup controller cannot take over, or when the controller is powered on, the PWR LED flashes but the STATUS LED remains red, and the node is not visible in the Metso DNA network.

Troubleshooting steps:

Check redundancy configuration: ACN MR supports 1:1 redundancy, requiring two controllers to be interconnected through dedicated Gigabit Ethernet ports (1000Base-T) and installed with the same SD card and real-time operating system version. Check the redundancy group settings in DNA Engineering Server and confirm that 'Redundancy pair' is enabled.

SD card swap test: All startup parameters of ACN MR (including RTOS, Process Controller software, and applications) are loaded by default from the backup server; In independent operation mode, it is loaded from the local SD card. If the main controller cannot start, you can unplug the SD card of the main controller and insert it into the backup controller to see if it can start normally. If possible, the original controller hardware is faulty (such as processor or memory); If not, the SD card is damaged or the file system is incorrect. Attention: After replacing the SD card, it is necessary to use Metso DNA engineering tool to re burn the image. Copying files alone is not effective.

Power supply voltage check: ACN MR requires 18-36 VDC, with a typical power consumption of 10 W. Measure the power terminals on the MBMT120 or MBMT80 mounting base, and the voltage fluctuation should be less than 5%. In redundant configurations, it is best for two controllers to be powered by independent power modules.

Ethernet communication diagnosis: ACN MR has five Ethernet ports: four 10/100Base-T (for inter node communication and I/O bus), and one 1000Base-T (for redundancy). Use the Ping command to test the controller IP address (obtained from the node list of DNA Operate). If not, check the switch port VLAN settings and cable type (direct/cross). Note: ACN MR does not have a serial console, and the only way to restore network configuration is to use an SD card to reset to factory settings.

Viewing real-time operating system logs: Connecting the monitor to the USB port of ACN MR? Actual ACN MR has no video output. A more professional approach is to enter RTOS maintenance mode through Ethernet using SSH (requiring Metso engineering privileges) and check for startup errors in/var/log/messages. Common errors include 'Failed to mount SD card' or 'License key mismatch'.

Solution:

If the SD card is damaged, use the D201915 ACN MR Node license to remake the SD card, ensuring that it includes the node configuration and process controller capacity license (D200990 per 100 I/O).

If the controller hardware fails, replace the ACN MR unit (40 × 125 × 95 mm, 1.1 GHz Atom processor). When reinstalling, please note that the MBMT120 (for M120 I/O) and MBMT80 (for M80 I/O) bases are different and cannot be mixed.

When the redundancy switch fails, manually force the switch: click "Force Switchover" in the redundancy panel of the DNA Operate controller. If it still fails, check whether the application versions of the two controllers are completely consistent (including the functional block diagram version).

4.4 Spectrum anomalies or inability to identify bearing defect frequencies in analysis tools

Phenomenon: In the Machine Monitoring analysis tool, a large number of unknown peaks appear on the spectrum, or known bearing fault frequencies (such as BPFI, BPFO) are overwhelmed by noise, and cursor movement cannot automatically jump to the calculated defect frequency position.

Troubleshooting steps:

Confirm that the RPM input is correct: In the "Rotational frequency" area of the analysis tool (page 45 of the PDF), check the currently displayed RPM value. If it is manually input, it needs to be consistent with the actual device speed; If it is a trigger signal from AIF8T, check if the trigger source is stable. Pressing the cursor button will automatically mark the positions of 1 ×, 2 × and other harmonics on the spectrum. If the marked positions do not match the expected values, it indicates that the speed reference is incorrect.

Verify bearing database and configuration: Click on "Bearing information" (PDF page 45) to ensure that the current measuring point is associated with the correct bearing model. Metso DNA contains a database of bearing defect frequencies (such as SKF, FAG) internally. If it is not displayed, you need to enter the Machine Structure Editor (PDF page 52) to manually add the bearing type, number of teeth, and transmission ratio. For example, for ball bearings, parameters such as pitch diameter, ball diameter, and contact angle need to be entered, and the system will automatically calculate BPFI, BPFO, BSF, and FTF.

Handling gear mesh frequency: For gearboxes, check if the number of gear teeth is correct in the "Gearmesh frequency" area. If there is sideband modulation (with 1 or 2 RPM intervals), it usually indicates tooth wear or uneven load. The "Sideband" marking mode in the toolbox (PDF page 48) can automatically position the cursor to the main engagement frequency and its sidebands.

Choose the correct signal processing path: In the analysis tool, users can select "time signal ->acceleration spectrum ->velocity spectrum ->envelope signal ->envelope spectrum" (PDF page 47). For early bearing failures, envelope demodulation should be used. Enter the Toolbox, select the "Signal" mode, set the upper and lower frequency limits of the bandpass filter (e.g. high pass 2 kHz, low-pass 10 kHz), and then generate the envelope spectrum. At this point, the fault characteristic frequency will be significantly prominent.

Check Waterfall Trend: If the fault is intermittent, you can call Waterfall to view the spectral evolution of multiple measurements. Select "Waterfall functions" in the "Browsing" area (PDF page 47) and observe the changes in frequency components over time. For example, a slow increase in amplitude of 1 x in hold mode indicates imbalance; And a sudden jump in BPFO amplitude indicates bearing spalling.

Solution:

If the RPM input is unstable, it can be changed to read the speed from PLC/DCS through OPC, or an independent proximity switch can be added to connect to AIF8T.

For bearing frequencies that cannot be automatically recognized, manually calculate and use "Marker functions" to add a fixed frequency cursor, and save the cursor as a user template.

Regular automatic spectrum baseline comparison for critical equipment operation: Select the "Spectrum RMS" trend in the "Trend" window (PDF page 55), set the full spectrum to be automatically saved once a week, and trigger an alert when the deviation between the new spectrum and the baseline exceeds 30%.

4.5 Improper configuration of Intelligent Alarm Handling (IAH) leading to false alarms

Phenomenon: When the device operates at different speeds, the fixed alarm limit often triggers warnings (such as higher normal vibration values at low speeds and lower values at high speeds), or the alarm does not work at all after the speed changes.

Background: Metso DNA's Intelligent Alarm Handling uses a "Notice Curve" instead of a fixed alarm limit. This curve divides the speed range into multiple intervals, with independent alarm thresholds set for each interval to adapt to variable speed machinery.

Troubleshooting steps:

Check if IAH is enabled: In the DNA Operate measurement window, check if there is an "I" symbol on the right side of the feature value bar (PDF page 65). If not, intelligent alarms need to be enabled for this feature value (such as acceleration RMS, envelope peak).

Validate Parameter Group and Recipe: Enter the "Intelligent Alarm Handling" button (PDF page 60). Firstly, confirm the existence of at least one parameter group, such as different groups for low-speed rollers (0-200 RPM) and high-speed motors (2000-6000 RPM). Each group needs to specify the feature value type (such as PEAK-HF, RMS-LF), speed reference trend (rotation frequency or machine speed), notification area (minimum maximum value of speed), and class number (PDF pages 62-63). A common mistake is setting too few classes (such as 5), which results in a rough alarm curve that cannot track the actual vibration trend.

Run notification curves: In the "Running notification curves" window (PDF page 64), select the target device (Process Area ->Machinery ->Eigenvalues), choose the parameter group, and set the time range for calculation (usually at least 2 weeks of historical data when the device is in good condition). After clicking "Run", the system automatically calculates the average value or a certain percentile (default 95%) of the characteristic values in each speed category as the alarm baseline. If no curve is generated after running, check if there are enough valid data points in the historical trend (at least 10 points are required for each class).

View and edit notification curves: Click the button under the "I" symbol next to the feature value to open the curve view (PDF page 66). The curve is in a stepped shape, with each step corresponding to an alarm threshold for a speed class. If the threshold of certain classes is abnormally high or low, you can directly drag the handle to adjust it. Note that the edited curve must be saved in order to take effect.

Delete notification curves to restore fixed alarms: If IAH causes confusion, the notification curves for that feature value can be deleted ("Delete notification curves" on page 64 of the PDF), and the system will revert back to using traditional grouped alarm limits (set in the Scaling Tool).

Best practices:

Run the notification curve only when the equipment is in good condition (check the vibration trend first and remove the fault data points before maintenance).

The amplitude range in the parameter group is set wide enough (e.g. 0-20 mm/s) to avoid compressing the ratio due to individual high points.

The recommended number of classes is 10-15, which can smoothly track speed changes without overfitting.

Regularly (quarterly) rerun the notification curve, as the equipment foundation vibration level may drift slowly due to wear or maintenance.

Preventive maintenance and system optimization suggestions

In addition to troubleshooting when faults occur, proactive maintenance measures can significantly extend the trouble free operation time of the Metso DNA system:

Regular calibration of sensors: The sensitivity drift of RVT/TT-125 is usually less than 1% per year, but in high vibration or corrosive environments (such as G3 industrial atmosphere), it is recommended to use a portable vibration calibrator for on-site calibration every 12 months. For the temperature channel, use a dry well furnace to verify at two points, 50 ° C and 100 ° C.

Cleaning and fastening of I/O modules: Open the cabinet every six months and use low-pressure compressed air to remove dust from the ACN I/O M120 module, with a focus on checking for signs of oxidation on the front-end connectors. For modules with optional conformal coatings (suitable for G3 environments), avoid using solvents for wiping.

Controller redundancy test: Perform a manual redundancy switch once every quarter (right-click on the controller node in DNA Operate ->"Switchover"), and observe whether the backup controller completes takeover within 10 seconds without losing process data. Simultaneously check if the SD card configurations of both controllers are synchronized (using the 'Synchronize SD card' function).

Trend database maintenance: Metso DNA's historical database defaults to retaining 3 years of vibration data. When the disk space approaches 90%, the system will stop storing new data. An automatic archiving task should be set up to dump raw signal files that are over 12 months old to network storage, retaining only the trend of feature values.

Remote diagnostic channel testing: If the factory has configured remote connectivity (such as through VPN to Metso expert center), test the remote desktop and file transfer function once a month. Ensure that experts can quickly intervene and analyze in emergency situations (such as diagnosing sub synchronous resonance or oil film turbulence).