Do the three signals from <R>,<S>&<T> processors to three coils MOOG fuel servo-valve have the same polarity? Because we tried to start our GE Frame 5 MKV HMI gas turbine but it failed to ignite and after reversing the polarity of processor<R> to servo-valve signal, the unit succeded to reach FSNL. What I have mentioned above is the only thing which brought the unit to FSNL after more than twenty failed attembts and after changing approximitly all fuel oil system components(main fuel pump, servo-valve, by-pass valve, clutch, flow-divider and all the ten check valves of combustion chambers).

 

The servo-valves and the servo-valve outputs of GE Speedtronic turbine control systems have to be about the most misunderstood devices and control system outputs in the world.

The servo-valves themselves are polarity-sensitive devices. Positive polarity DECREASES the flow of fuel or air or steam to the turbine/prime mover, while negative polarity INCREASES the flow of fuel or air or steam to the turbine/prime mover. In response to the need to decrease or increase fuel or air or steam, the Speedtronic will output a negative polarity or a positive polarity signal AT THE SPEEDTRONIC TURBINE CONTROL PANEL TERMINAL BOARDS. If the need to decrease or increase fuel or air or steam is large (that is, the error between the reference and the feedback is large), the magnitude of the polarity will be large.

Under steady-state operating conditions when the reference to the device and the feedback from the device (Gas Control Valve or IGVs (Inlet Guide Vanes) or V1 Valve rack, etc.) are equal, the output of the regulator driving the servo-valve output is ZERO mA--because the error between the reference and the feedback is zero. A null-bias current (a negative value) is added to the output to overcome the servo-valve's null bias spring tension which is trying to force the servo-valve to shut off the flow of fuel or air or steam (hence the negative current to keep the servo-valve "open" against the spring tension which is trying to close it).

Because there are two wires attached to each coil of the servo-valve, IT IS CRITICAL THAT THE POLARITY OF THE CURRENT BEING APPLIED TO THE COILS BE VERIFIED ANY TIME A SERVO-VALVE IS REPLACED. (The procedure has been detailed in previous posts to control.com.)

This author, and other contributors to control.com have affirmed, that the colored leads of new-in-the-box and unused Moog servo-valves can be connected to the coils such that the polarity of the current being applied will work in the opposite direction of the coil that is being replaced.

In other words, one can disconnect the red- and white leads of one servo-valve coil and connect the red- and white- leads of another new-in-the-box servo-valve coil EXACTLY THE SAME AS THE PREVIOUS SERVO-VALVE COIL'S red- and white leads, and the internal connections to the coil in the servo-valve will cause the current to be applied OPPOSITE to the way it was applied to the servo-valve which was just removed. FACT.

So, it's not the polarity of the outputs of the Speedtronic that "change" when replacing servo-valves, it's the internal wiring of the servo-valve coils which changes (even though the leads are color-coded identically and should NOT differ) and hence the need to verify the polarity of the current being applied to every coil of every servo-valve being replaced--and corrected if necessary.

markvguy

 

I have read the following:

"Because there are two wires attached to each coil of the servo-valve, IT IS CRITICAL THAT THE POLARITY OF THE CURRENT BEING APPLIED TO THE COILS BE VERIFIED ANY TIME A SERVO-VALVE IS REPLACED. (The procedure has been detailed in previous posts to control.com.)"

Can you help me telling me when those previous posts appeared?

Furthermore, I would like to know the typical reading of coil resistances in MTS 256 servovalves.

Thanks in advance.

 

This author doesn't recognize "MTS 256" servo-valves, but if we're talking about servo-valves used on GE-design heavy-duty gas turbines since at least the early 1980s, generally most of the servo-valve coils this author measured were approximately 1000 ohms (when the servo-valve coil(s) was(were) good).

This actually proves to be very convenient for monitoring servo-valve outputs with a voltmeter. The servo-valve outputs of Speedtronic turbine control panels applied to GE-design heavy-duty gas turbines are +/- 10 mA outputs, so 1 mA equals 1 VDC!

The purpose of verifying the polarity of the servo-valve output current being applied to servo-valve coils is to ensure that all three processors are working together to properly position the device. To accomplish this, the device has to be positioned under the control of each individual processor (the servo-valve output current from a single processor is sufficient to maintain device position at a position nearly equal to the reference).

NOTE: ALL SOURCES OF ENERGY (fuel, steam, etc.) MUST BE PROPERLY ISOLATED UPSTREAM OF THE DEVICES TO BE TESTED, AND VENTED AND/OR PURGED ACCORDING TO SITE PROCEDURES AND/OR LOCAL REGULATIONS.

In preparation for checking the polarity of the servo-valve output currents being applied to the servo-valve's coils, one must determine the terminals in the Speedtronic turbine control panel where the servo-valve coils of the device to be tested are terminated--for all three coils (for a TMR (Triple Modular Redundant) control panel). There should be three pairs of terminals. After determining where these terminals are, locate them in the Speedtronic turbine control panel.

To verify servo-valve polarity, the unit must be shut down. One needs to establish hydraulic pressure, and if necessary, energize a stop valve solenoid or dump valve solenoid to be able to "stroke" the device.

Position the device to be tested at mid-stroke using AutoCalibrate's Manual Position feature, or whatever method of manual positioning ("stroking") is used on the version of Speedtronic panel you have.

Then disconnect one of <T> processor's servo wires from the Speedtronic turbine control panel. Due to the loss of the null bias current from <T>, the device position will decrease slightly, but remain at approximately the mid-stroke position.

With the <T> processor's servo wire still disconnected, disconnect one of <S> processor's servo wires from the Speedtronic turbine control panel--leaving the device under the control of <R> processor's servo-valve output only. If the servo current from <R> is properly connected to the servo-valve (that is, the "polarity" of the applied current--meaning the way the coils are being connected to the Speedtronic turbine control panel), the device's position will decrease a little more (again due to the loss of <S> processor's null bias current), but should remain at approximately mid-stroke.

If the device stays at approximately mid-stroke under the control of <R> processor only, then continue with the testing below.

However, if the polarity of the servo current from <R> is not being applied properly to the servo-valve, the device will move very quickly to the minimum flow position (in the case of the Liquid Fuel Bypass Valve, that would be full open; for all other devices, that would be full closed/minimum mechanical position).

If the polarity of the servo-current from <R> is not correctly connected, reverse <R>'s servo-valve wires at the Speedtronic turbine control panel, and the device will return to approximately mid-stroke.

To continue with the testing, reconnect the <S> processor servo-valve wire at the Speedtronic turbine control panel and disconnect one of <R> processor's servo-valve wires. If the servo current from <S> is properly connected to the servo-valve, the device will remain at approximately mid stroke; and testing can continue.

If the polarity of the servo-current from <S> is not correctly connected, reverse <S>'s servo-valve wires at the Speedtronic turbine control panel, and the device will return to approximately mid-stroke.

To continue with the testing, reconnect the disconnected <T> processor servo-valve wire at the Speedtronic turbine control panel and disconnect one of <S> processor's servo-valve wires. If the polarity of the servo current from <T> processor only is correct, the device will remain at approximately mid stroke; and the procedure can continue.

If the polarity of the servo-current from <T> is not correctly connected, reverse <T>'s servo-valve wires at the Speedtronic turbine control panel, and the device will return to approximately mid-stroke.

Once the device's position can be maintained at approximately mid stroke under the control of each single processor, reconnect the servo wires of <R> and <S> and the procedure can be considered to be complete. With the restoration of the null bias currents from the remaining two processors, the device's position feedback should be nearly equal to the reference.

NOW, having said all of the above--there are two devices which are a little more difficult to "stroke" manually: the SRV (Stop-Ratio Valve) and any LFBV (Liquid Fuel Bypass Valve).

The regulator of the SRV is primarily a pressure control loop, so unless it is "reconfigured" to make it a position control loop, if a reference is applied to the SRV with no pressure upstream (and the gas fuel supply MUST be isolated upstream of the SRV and vented off _BEFORE_ the SRV is stroked) it will go wide open. The SRV will still slam shut if the polarity of servo-current being applied from any single processor is incorrect, but it can difficult to understand why if it has a manual reference of 50% it goes full open....

The regulator of the LFBV is primarily a flow-rate control loop, so unless it is "reconfigured" to make it a position control loop, if a reference is applied to the LFBV with no liquid fuel flowing through the liquid fuel flow divider it will go to the full closed position (full flow). The LFBV will still move to the full open position (zero flow) if the polarity of servo-current being applied from any single processor is incorrect, but it can difficult to understand why if it has a manual reference of 50% it goes full open....

Also, LFBVs without LVDTs (Linear Variable Differential Transformers), such as the ones used on Frame 5- and Frame 6 GE-design heavy-duty gas turbines, can be especially problematic since one can't see the valve stem or see position feedback. In this case, the valve piping has to be removed to be able to observe the valve plug's movement.

This procedure will work for steam turbine servo-operated devices as well.

AND, it should be noted that servo-valves used with SIMPLEX Speedtronic turbine control panels are usually two-coil servo-valves, and the polarity of servo-valve output currents applied to each of the coils needs to be verified just as with a TMR turbine control panel. It is more difficult because it may be necessary to disconnect the servo-valve wiring at the JB (Junction BOX) closest to the servo-valve--difficult, but not impossible.

The Control Specifications provided with most Mark IV, and many Mk V, turbine control panels contains comments saying something to the effect that to verify servo-valve current polarity that the device must be manually stroked and if the movement is "jerky" or not smooth that the polarity is not correct. This is simply not a valid method of verifying servo-valve currents/polarity.

markvguy

 

Dear MarkVGuy,

Your notes regarding Mark V and Mark VI GE control systems are extremely useful and interesting.

From below explaination I understand that even if one coil is working good, the turbine will continue to run. Is this right?

Regards...
Venkat

 

Not to speak for markvguy, who's no longer contributing, but one coil of a three-coil servo-valve on a TMR Speedtronic can exert enough torque to control the flow of hydraulic oil to an actuator to make the actual position very close to the required position.

The loss of the effect of the null bias current from the other coils will cause there to be a slight error between actual and reference positions.

Some Speedtronic systems have diagnostics monitoring servo outputs and will initiate alarms when no current is detected.