System Overview and Maintenance Background
GE SPEEDTRONIC ™ The Mark VI turbine control system is an integrated control platform developed by GE based on over 30 years of turbine control experience, widely used for control, protection, and monitoring of gas turbines and steam turbines. Mark VI adopts a distributed processor architecture and supports two configurations: Simplex and TMR (Triple Modular Redundancy), with high availability and fault tolerance.
In actual operation, the control system may trigger alarms or shutdowns due to aging I/O cards, communication abnormalities, power fluctuations, sensor failures, and other issues. This article is based on the typical structure of the Mark VI system, summarizing common fault phenomena, diagnostic processes, and maintenance strategies to help on-site engineers respond quickly and reduce unplanned downtime.
System core architecture and redundancy mechanism
2.1 Control module and VME backplane
The core of Mark VI is the Control Module, which uses standard VME racks (13 or 21 slots). The main processor card (occupying 2 slots) executes application software and supports frame rates of 10ms, 20ms, and 40ms. The I/O card is equipped with a TMS320C32 DSP processor for data filtering and floating-point conversion. The data interacts with the VCMI communication card through the VME bus.
Fault prompt: If there is a periodic data refresh delay in the system, priority should be given to checking whether the VME backplane connection is loose and confirming whether the main processor card load is too high.
2.2 TMR Redundant Architecture
The TMR system consists of three independent control modules, power supply, and IONet network, using SIFT (Software Implemented Fault Tolerance) software voting mechanism. The input signal is voted in the software, and the output uses a 3-coil servo valve or 2/3 relay logic. TMR supports online replacement of single point of failure, significantly improving system availability.
Maintenance strategy: When replacing faulty cards in the TMR system, it is necessary to first confirm that the module where the faulty card is located has been "isolated" to avoid triggering cross tripping by mistake.
Troubleshooting of I/O interface and universal card components
3.1 Discrete and Analog I/O
VCRC card: 48 point DI/24 point DO, compatible with TBC1, DTC1 and other terminal boards. The typical fault is that the input point has no state change, and the common reasons are external dry contact oxidation or optocoupler damage.
VAIC card: 20 channels of AI/4 channels of AO. If the reading of a certain channel jumps, it is recommended to check whether the grounding is good and use a signal generator to verify the linearity of the channel.
VTCC/VRTD card: thermocouple and RTD inputs. If the temperature display is abnormal, the sensor disconnection or grounding problem should be ruled out first, and then the jumper settings on the terminal board TBTC/TRTD should be checked for correctness.
3.2 Vibration Monitoring (VVIB Card)
The VVIB card supports seismic probes, Proximitors, Velomitors, and accelerometer inputs. There are a total of 16 vibration channels, 8 axial displacement channels, and 2 Keyphasor channels. If a certain vibration value continues to be high, check whether the probe gap voltage is within the nominal range (-24Vdc power supply limited current protection).
On site experience: When the speed exceeds 3000rpm, the sampling rate of the VVIB processor needs to be adjusted, which may cause the number of available channels to decrease from 16 to 8. After replacing the probe, recalibration is required.

Special card malfunction and handling
4.1 VTUR card: speed, synchronization, and trip interface
The VTUR card provides 4 speed inputs (2-14kHz), 2 PT synchronous inputs, and a trip relay interface. Common faults:
Loss of speed signal: Check the gap between the magnetic sensor (recommended 0.5-1.0mm) and the grounding of the cable shield.
Synchronization failure: Confirm if the PT input voltage on the TTUR terminal board is normal, and check the action logic of K25P and K25A relays.
4.2 VSVO card: servo valve control
VSVO card provides 4 servo channels, supporting LVDT/LVDR feedback and pulse input (flow meter). Three coil servo valves are commonly used in TMR systems.
Fault symptoms: Valve shaking, slow response, or inability to operate.
Diagnostic steps:
Check if the Status LED on the front panel of the VSVO card is yellow (hardware alarm).
Check if the LVDT feedback values are consistent in the software maintenance tool (it is recommended to use at least 2 LVDTs).
Measure whether the resistance of the servo valve coil matches the nominal value (see Table 7: e.g. coil # 1 1000 Ω± 10mA).
4.3 VGEN Card: Generator Interface with PLU/EVA
VGEN card is used for three-phase PT/CT monitoring, power load imbalance (PLU), and early valve action (EVA). The common alarms are "PT disconnection" or "CT saturation".
Suggested solution: Check the fuse and wiring on the TGEN terminal board to confirm that the PT secondary voltage is normal. If it is an EX2000 system, abnormal network communication may also cause VGEN data to be invalid.
4.4 VPYR Card: Infrared Pyrometer
VPYR card monitors 2-channel LAND pyrometer for measuring blade temperature distribution. If the temperature suddenly drops to 0 or reaches full range:
Check if the ± 24Vdc power supply on the TPYR terminal board is functioning properly.
Confirm if the Keyphasor signal is stable (for speed synchronization sampling).
Troubleshooting of Communication System Malfunctions
5.1 IONet and VCMI Card
IONet is a polling based Ethernet (ADL protocol) based on 10Base2, which is more deterministic than standard Ethernet. VCMI card serves as the master station and communicates with the slave station module.
Common faults: Some I/O card data is not updated, CRC check error.
Troubleshooting steps:
Check if the coaxial cable connector is oxidized or loose.
Use IONet monitoring tool to check packet loss rate.
Compare the data consistency of three IONets in the TMR system.
5.2 UDH Network and HMI Communication
UDH is based on the Ethernet Global Data protocol and supports direct communication with the main processor card. If HMI data refresh stagnates:
Check the Ethernet indicator light on the main processor card.
Confirm that the switch port is not faulty and that the IP address falls within the Class C private address range (192.168.x.x).
In the TMR system, two UDH drivers are connected to one switch and the other to a backup switch to avoid single point of failure.
5.3 Modbus Communication
The main processor card of Simplex system provides RS232 Modbus Slave port (RTU/ASCII). If DCS reads abnormal data:
Check if parameters such as baud rate and parity match.
Confirm that the Modbus address mapping table is not conflicting.
Modbus in TMR system only supports monitoring and does not support control commands.
Power system and grounding diagnosis
Mark VI is typically powered by a 125Vdc battery system and also supports 120/240Vac isolated rectification for power supply.
Power alarm: If the "Ground Detected" alarm occurs, it indicates that the insulation of the 125Vdc floating ground system to the ground has decreased, and the feeder and terminal board should be checked in sections.
Undervoltage alarm: Check if the AC input is below the set value and confirm if the diode high select circuit is working properly.
Trip circuit power supply: The TRPG terminal board provides the negative terminal for the trip coil. If the trip circuit cannot be reset, check if the 3.2A fuse on TRLY is blown.
Important reminder: The ignition transformer of the gas turbine requires a separate 120/240VAC power supply, and the short-circuit protection should not exceed 20A.

Software maintenance tools and online diagnostics
7.1 Control System Toolbox
This tool runs on the Windows NT/95 platform and supports:
Online Logic Modification (Password Level 5)
Signal forcing and trend analysis (frame rate)
I/O configuration and limit settings (high/high, high/low, low/low)
Attention: Before downloading the application software online, it should be confirmed that the program has not caused any servo valve misoperation. Suggest downloading to the backup module for verification in the TMR system first.
7.2 Alarm and Event Management
All process alarms and diagnostic alarms are accompanied by frame rate timestamps (10/20/40ms), with SOE event accuracy of 1ms. If the alarm queue is abnormal:
Check the UDH communication between HMI and main processor.
Confirm whether the alarm limit setting is reasonable (such as high/high limit lower than the hardware limit value).
7.3 Hardware Diagnosis and LED Indication
Each card front panel provides:
RUN (green): Blinks during normal operation.
FAIL (red): Hardware malfunction.
Status (yellow): Software limit alarm or configuration error.
On site operation: After replacing the card, the hardware diagnostic alarm needs to be reset through the toolbox, otherwise the STATUS light may continue to flash yellow.
Environment and Preventive Maintenance
8.1 Environmental Requirements
Temperature: Operating at 0 ℃~45 ℃, it can withstand 50 ℃ for a short period of time (for air conditioning maintenance).
Humidity: 5%~95% non condensing.
Pollution: Complies with EN50178 gas corrosion requirements and has an IP20 dust protection rating.
Suggestion: Clean the cabinet filter once a year and conduct a visual inspection of the circuit board, especially paying attention to whether the electrolytic capacitors are bulging.
8.2 Seismic Resistance and Installation
Mark VI meets the seismic requirements of UBC Zone 4. If there has been an earthquake or impact on site, the fixing bolts and terminal board connectors of the VME rack should be checked for looseness.
Typical Fault Case Quick Reference Table
Troubleshooting steps for possible causes of fault phenomena
A certain analog input is always 0. The sensor is disconnected and the terminal board jumper is incorrect. Measure the input voltage/current, check the TB jumper, and replace the VAIC channel
Servo valve jitter LVDT feedback inconsistency, abnormal coil resistance check VSVO feedback value, measure coil resistance, check grounding
IONet communication interruption VCMI card fault, coaxial cable short circuit inspection VCMI status light, segmented measurement of cable continuity, replacement of terminal resistance
HMI data display delay>1 second UDH network congestion, high main processor load check network traffic, reduce unnecessary EGD broadcast data
Frequent false tripping of the system, circuit interference, and power supply ground potential drift inspection of TRPG/TREG terminals, measurement of 125Vdc ground voltage, and addition of filtering capacitors
