We have a LM6000 PG. We have had issues with the variable geometry for some time. We have worked through most of the issues but we are currently getting stumped by the VBV Voltage Sum Fault.
We get this alarm at random times mostly on a start up or shut down. We have Woodward MicroNet Plus controllers the card part number that drives the LVDTs is 5501-1432
We have worked through the GE trouble guide and ended up replacing the on and off engine cables as requested by GE we have replace the LVDTs and done countless VG Cals. We have swapped the driver card in between the VSV and VBV the problem did not follow the card.
I have taken a DMM and dropped the lines to the LVDTs from the FTM connected into the leads with spring nuts and ran a long jumper into the package connected to the DMM with banana plugs; then attempted to pull push on the connectors wile watching the DMM for deflection. We also had GE bring a manual pump and pump the VBV to 25% 50% 75% and 100% and attempt to pull push on the connections in attempt to recreate the issue yet keep coming up empty handed.
When I read the LVDT with a DMM the values are P1=108.6Ω S1=74.2Ω S2=72.8Ω plus or minus a few 10ths of a ohm across all my LVDTs. I get the same readings also within a few 10ths directly at the LVDT and at the FTM in the MTTB.
Moderator's Note: Not sure what symbol belongs instead of "Ω". Will try to find out.
We also have a issue of not building enough ps3 for sprint to turn on at this point I think it may be to issues with the VBVs letting to much air out of the engine. GE has provided us a schedule of the amount the VBVs should be open and we are within that proper range of approx 25% open at full load but I am having a hard time believing we should be 25% open when our trouble shooting manual and all the text I have found online says that over 33MW or 60% load the VBVs should be closed..
One more thing.. On the HMI the actuator ohms reads approx 40 on A channel and approx 20 on B channel across IGV, VSV, and VBV. I have pulled a new LVDT verified its values with a DMM and placed it on A channel and it will read 40 ohms then moved that same LVDT to B channel and it will read 20 ohms..
Any one have any ideas??
This is a shot in the dark for me, because I have very little experience with LMs, and even less with MicroNET controllers (of the Plus or non-Plus variety).
I'm a little confused about your last couple of paragraphs. You mention taking LVDTs and swapping them around, and the resistance readings for the "actuator" on the HMI not changing for the two channels regardless of the LVDT connected to that channel.
Sounds to me like a wiring issue somewhere. I know you say you have replaced the on- and off-engine cables, but I'm wondering if the unit uses active- or passive IS (Intrinsically Safe) current limiters on the wiring to protect against explosions. I have seen these seemingly innocuous little devices (usually mounted on a DIN rail) cause all manner of similar problems with devices and readings. Sounds like you've replaced the wiring/cabling, but if those devices are installed and they are failing or have failed for the A device but not the B device, then that could explain the issue.
I have also seen high-density DIN rail terminals cause similar problems when they have been over-torqued or were just plain defective (the defective ones, from the factory, are usually found during commissioning, but not always; and well, the over-torqued ones are usually the last things that people think about--and to be found after a lot of head scratching).
When I've tried all of the "low-hanging fruit", I usually find if I go a little "lower" (like terminations, and IS barriers, when used) I sometimes (more often than I would ever have expected) find a very simple solution to the problem. (I once found that some "electrician" had been putting wires in DIN-rail high-density terminals UNDER the capture portion of the screw terminal and "un-screwing" the terminals to try to hold the wires in the terminal. Some wires (thicker gauges of wire) were tight, with only metal on one side of the wire and plastic on the other. Smaller gauges of wire weren't so tight, and in some cases this person had stripped more insulation and folded the bare wire and pushed it in the terminals before trying to "tighten" the terminal by "unscrewing" it. Those terminals--usually on critical I/O terminations over time caused a LOT of problems. We finally had to shut the unit down over a weekend and go around and check and correct about 35% of the terminations which had been done during construction. I don't know what that person was thinking, or where they had learned their trade, or if they were an untrained, unsupervised newbie, but it caused a lot of nuisance trips and alarms! And a couple of days of lost production--and revenue. And, after that, the problems mysteriously just "went away!")
Short of that, I would suggest contacting Woodward, if you haven't already, for assistance. Lately, though, I have heard their Customer Service response times are pretty ... long. And, it sometimes takes multiple contacts to get a response. But, if you keep trying, you can get a reply--usually. Just seems to be taking longer than it has in the past for whatever reason.
I, too, am surprised the VBVs would be open at higher loads.... Best to reconfirm that with GE. Just seems a little odd--but then, you are working on an aeroderivative unit, and we don't know what kind of combustion system is in use and those things can be pretty finicky.
Anyway, just a suggestion to check all the wiring and cabling and terminations--and if the unit uses IS barriers (active or passive) in the circuits, best to just replace them, or maybe just swap them (say A to B, and B to A, and see if the problem follows the barrier). Can't take that much time, and may actually help find the problem.
Please write back to let us know what you find. (Also know that we don't troubleshoot a lot of LMs on control.com, so responses might be few or brief.)
* On this system you would have dual redundant controller cards that current share.
* When one LVDT, one card, or one side torque motor fails the other card takes over.
* The Woodward servo (HSU) on the gearbox has 3 torque motor coils for each VG loop.
* For simplex cards two coils are wired in parallel.
* When the redundant card is added it uses the 3rd coil.
* Each coil is around 40 Ohms so that is why you have one at 20 and the other at 40.
The sum of each of the secondary LVDT coils stays constant through the stroke. The linearization voltage to % scaling uses (A-B)/(A+B). The voltage sum on the bottom is constant and the delta part of this equation is what change with the stroke.
The alarm is a result of the sum portion not being constant within the programmed tolerance. I suspect a problem related to one of the secondary coils.
The LVDT Ohms readings you have mentioned are correct.
Since you mention it is more random and happens more on start up or shutdown, maybe it is a noise issue with something that happens in the sequence.
One check is to lift the ground point on the shield and measure it back to make sure you have no other grounds.
The other thing to check is to lift the MicroNet cable from the FTM
and measure all the signals to ground. Make sure there is not a grounded +/- LVDT or drive signal.
As for the VBV scheduling question I have a couple of ideas.
Check to see if you have a % difference between the A and B LVDTs.
At the difference alarm the loop picks the LVDT that will open the VBV (safer position). If you capture a high speed datalog it may show more details as well. Adding all the voltages and faults from the VBV actuator channels to the datalog configuration may show something.
The other thing I may suggest is looking back at some historical data. Take a look at the VBV position at similar load (N25/HP compressor speed)and ambient temperature conditions.
One more thought is to check the PS3 instrumentation sense line for loose fittings or water in the line ... possibly if accidentally not isolated during a water wash.
>When I read the LVDT with a DMM the values are P1=108.6Î©
>S1=74.2Î© S2=72.8Î© plus or minus a few 10ths of a ohm
>across all my LVDTs. I get the same readings also within a
>few 10ths directly at the LVDT and at the FTM in the MTTB.
>Moderator's Note: Not sure what symbol belongs
>instead of "Î©". Will try to find out.
From the context and historical experience with character misinterpretation of the extended ASCII character set, the Î© characters are very likely the Ohm symbol.
The Ohm symbol is frequently misinterpreted as a W character, as noted a almost a decade ago:
The first step would be to get high speed (10ms) data with all the LVDT voltages, feed backs, position demands, torque motor currents when the event occurs. Since it apparently tends to manifest itself when the engine shuts down I would suggest making sure you have it running the next time you shut down. Looking at the response of the voltages during the shutdown may give you a clue as to the nature of the problem.
The first thing to check in that trend (log file) would be the magnitude of the voltage sums. The voltage sum for the VBV LVDTs should be in the neighborhood of 5.6 to 5.7 volts. The upper range limit should be 6.0 V and the lower range limit should be 5.3 V. Basically, the actual voltage should be roughly centered between the limits. So check that the actual voltage is centered between the limits and relatively constant. If it is on the low side it may indicate that you are loosing voltage somewhere along the way. One possibility would be that the cable lengths are longer than typical. Those limits are adjustable so if it just a centering problem the limits can be changed to re-center the voltage. However, I would not recommend doing that until you were certain where an why the voltages are less than expected.
Also, during the shutdown it is not uncommon for the voltage sum to drift up but only very slightly. Basically, the electrical load on the control system is higher when the engine is running so after shutdown the excitation voltage for the LVDT may increase very slightly so the secondary voltages may also increase very slightly. Voltage regulation in the control should minimize this so typically the drift is less than 0.1 V. There is logic in the control to detect when the sum of secondaries drifts too fast so a larger than expected drift in a short amount of time could be causing the fault. If this is the case it would point to some kind of electrical problem in the control causing it not to regulate voltage like it should
The other possibility that sometimes occurs during shutdown is that the cable runs for the LVDTs are too close to some high power cables that supply power to equipment that is getting turned off at the moment of shutdown. It is possible that the EMI from high power cables is affecting the LVDT voltages. This problem should manifest itself as short duration voltage spikes on the secondary voltages. If you see these spikes at shutdown then check to make sure cables carrying the LVDT voltages are not laying right next to high power cables.
Also, look for abrupt changes, typically abrupt changes downward, in one or more of the secondary voltages. Short spikes downward tend to indicate an open circuit and are an indication of connector/cable issues. The most common causing being a connector that was not properly tightened or a cable that is not properly supported. A cable that is not properly supported will have a tendency to pull on the connector causing movement of the connector. The end result for both conditions is a fretting on the pins of the connector. The fretting will eventually cause oxidation of the surface of the pins resulting in poor electrical connections/open circuits. You would think that an open circuit would result in 0.0 V but on the Woodward control an open circuit on the VBV LVDTs often results in reading about 1.5 volts. The reason why is buried in the Woodward circuit design. The open circuit issues tend to happen more often when the engine is running due to the greater vibration caused by the engine. You can visually examine the pins for fretting but you may need to use a magnifying glass.
Finally, I would not rule out an electrical issue with the coils themselves as has been already mentioned above.
I hope this helps,