MK5 125VDC grounds

E

Thread Starter

elmarko

I have a 125VDC ground in the MK5 L64D_N comes in and out giving me the alarm. I have removed the 125vdc power to the DTBA, DTBB in QD1 QD2 and CD cores. I have a 90-35 split which I'm reading from DiagC. Removing the 125vdc does not change the split, but when I power down either R, S, or T cores my voltage evens out to where it should be on the positive 62.5 and negative 62.5. Once the core is powered up the split goes back to 90-35. Have you seen this before?
 
You do seem to have an unusual problem. Not all grounds are related to contact input- or solenoid output circuits, and finding such grounds is more difficult. (As we all know, finding any ground in a Mk IV or Mk V SpeedTronic turbine control system can be very difficult!)

What Diagnostic Alarms (if any) are currently annunciated? Are any Diag. Alarms being annunciated "simultaneously" with the battery ground alarm?

One caution: DIAGC can be very misleading. The contents of DIAGC.DAT MUST EXACTLY MATCH the verions of PROMs installed in the Mk V turbine control panel, and there's only one person in the world who can say if DIAGC.DAT does or doesn't (he works for GE in Salem, VA). Now, the ground- and power supply voltage sensing is fairly "generic," so that should probably be okay. But, the best way to measure your power supply voltage split is to measure it with a voltmeter at the <PD> PTB terminals.

The ground you have is sometimes referred as a "soft" ground, meaning that one side of the battery is not solidly grounded (a voltage split of 0.00 VDC with respect to ground on one side of the battery would be a "hard" ground). Soft grounds can be especially hard to find even on contact input- or solenoid output circuits and are usually caused by moisture--water or oil. Soft grounds caused by moisture can even appear to "switch polarity" during troubleshooting or unit operation--an almost sure sign of a ground caused by moisture.

Has your Mk V been kept clean and "dry?" Humidity and dirt can cause the kind of problem you are experiencing. Many people think that the air conditioners provided in the Control Compartments are just for "human comfort;" in fact, they are there to prevent condensation and reduce humidity. In humid environments, sometimes this requires keeping the compartment temperature setpoint slightly elevated depending on ambient conditions. But, the point is that the air conditioning (including heating) in the enclosure where the Mk V is located is very important, as is keeping the Mk V cards as clean and free of dust as possible. If the cards are dusty or dirty and there is any moisture or condensation, then all the conditions for unusual Mk V problems are present....

Many installations resulted in the air from air conditioners blowing almost directly "on" the Mk V. In dusty environments this causes lots of dust and dirt to be directed into the Mk V enclosure. Many installations had cables which were brought up into the bottom of the Mk V enclosure, and they were not sealed very well from dirt and moisture in cable vaults below the enclosure. Even in a location with good air conditioning/humidity control, moisture from cable vaults below can be "drawn" up into the Mk V if not sealed properly.

Maintaining the seals around exterior Control Compartment doors is also very important, as is making sure that doors are closed except for entry/exit.

The 125 VDC supply goes through the <PD> core to the TCPS power supply cards in each processor core, as well as to the TCEA cards and TCTx card in the <P> core. (There have been cases of grounds (soft and hard) on the trip solenoid outputs connected to the <P> core, though this doesn't sound like your problem.) On the TCPS card, the DC is inverted to AC, then transformed to lower voltages, then converted back into regulated DC power supplies; this usually prevents any grounding of the analog I/O (4-20 mA inputs, servo-valve outputs, etc.) from causing 125 VDC battery grounds.

When troubleshooting grounds in a Mk V panel--epsecially soft grounds--it's important not to disconnect a wire or cable, then reconnect it immediately if the ground doesn't disappear. Especially with grounds caused by moisture, there can be more than one wire or termination in the junction box where the moisture is located, so one can lift every single wire in the panel one at a time, relanding it before continuing on to the next wire, and still not find "the" ground--because there is more than one ground in such cases. While it doesn't seem your ground is in the contact input- or solenoid output circuits, if you have disconnected the power supply cables to the DTBA/DTBB cards one at a time and then reconnected them before powering-down the processors, you may have parallel ground paths causing your problem(s).

Please write back and let us know what you find. Sometimes when troubleshooting soft grounds in a Mk V turbine control panel it is necessary to let the ground "progress" to a point where it can be located when troubleshooting.

The purpose of having an ungrounded battery supply for the Mk V is to permit the unit to continue to run when A hard ground of one leg of the supply does occur until such time as the ground can be located and resolved. If a hard ground were to exist on one leg of the supply and another hard ground were to develop on the other supply leg, then it's like applying a dead short across the battery! So, it's best to aggressively troubleshoot and resolve hard grounds. Soft grounds aren't as critical, but they should be continually investigated especially if the voltage split continues to deteriorate (one leg gets closer to 0.00 VDC and the other gets closer to full battery voltage).
 
The soft ground has not changed it is still a 90/35 split. I tried removing the power once again and leaving the supply to each DTBA, DTBB, DTBC off and there was no change (I removed all power to all terminal boards). With the power disconnected to the QD1, QD2 and CDs DTBA, DTBB, DTBC I then powerd down <R> and there was no change, it was the same for <S> and <T>. With no power on the inputs or outputs and with a core powered down the split remained bad at 90/35. If you plug all of the power back on to the DTBA, DTBB, DTBC with either <R> <S> or <T> powerd off the split goes to 63/62. I guess my next step is to change out the TCPS power supply cards in each core and see what happens. P.S.: I also powered down <P> core and nothing changed.
 
The only card(s) you haven't mentioned are the DTBD cards on the digital I/O cores. Is there no I/O connected to these cards? No solenoid outputs?

Have you checked the filter capacitor on the ground bus bar? This has been found to be "overheated" and causing a soft ground on more than one turbine control panel. (NOTE: It's NOT recommended to permanently remove the filter capacitor.)

Regardless, you seem to have done about everything possible. A soft ground is not a real problem, even should a hard ground develop on the other leg. It IS something to watch and stay on top of. As mentioned earlier, with soft grounds sometimes all one can do is wait for it to get bad enough to be able to find.

Occasionally, the turbine control panel 125 VDC battery (we are talking about a battery here, and not <DACA>s, right?) is used to power some "unusual" load. High-voltage switchear (breakers, protective relays, fire detection systems, hazardous gas detection systems). A ground in that circuit would be annunciated on the SpeedTronic, but not physiclly in- or on I/O connected to the Mk V.

On GE-built units, the 125 VDC battery power for the Mk V usually supplies BOTH the turbine control- AND the generator control panels. (The wires from the breaker are usually landed in the GCP, and jumpers connect the TCP to the GCP.) So, a ground could be developing in some circuit associted with the GCP, but would be annunciated on the Mk V, and interacting with the power supply(s).

The ground voltage level setpoint is in the I/O Configurator. If the ground alarm is an annoyance and you don't want to lock it out and check it periodically, it would be okay to lower the setpoint a volt-and-a-half or two for a period of time while waiting for the ground to get worse, and easier to find. The setpoint is usually on the
last screen of the TCQA card.

The <Q> core power supply monitoring is shown on Fig. D-55 of the Signal Flow Diagrams in Appendix D of the version GEH-6195D of the Mk V App. Manual.

One more thing to check would be the "integrity" of the cable connectors on the I/O cards. GE has been recommending using conductive grease on the ribbon cable connections. If possible, try unseating and reseating the ribbon cables, and using conductive grease if they seem corroded or excessively dry.

markvguy
 
There's one other thing to check which just came to mind. Not too long after the start of production of the Mark V it was necessary to add a "conditioning power filter" (CPF>) to the panels to meet code requirements in certain parts of the world (at least that was the rationale given at the time for the addition of the new component).

The <CPF> was usually mounted at the bottom of the panel, under the <QD1> and <CD> cores against the back wall of the panel (really NICE and ACCESSIBLE location-- but that's what happens with AFTERthoughts...). These filters have been damaged by lightining strike-induced voltage spikes and problems with battery chargers, and on one occasion resulted in a very difficult-to-find ground.

One wonders whether some kind of "interaction" with a failing/failed component of the <CPF> and the TCPS card(s) might cause some kind of soft ground under certain conditions.

markvguy
 
Top