The DCC I/O card objid #4, reset alarm keeps coming in spuriously. When it comes in we get numerous voter mismatches for the affected core. What could be the problem and corrective action? We also get a lone l86cstall on this same core during a startup and shutdown. How do I fix this problem?

 

If the Diag. Alarm message is understood correctly, it sounds like there's a problem with the TCDA card for the affected processor (core). OBJ ID #4 or target 4 or Socket #4 (three ways of referring to the same printed circuit card--isn't the Mark V fun?!?!!) is the TCDA card for the affected processor. Are most of the voter mismatch alarms which come in when the OBJ ID #4 alarm is annunciated related to contact inputs and relay outputs? Are you able to watch the LEDs on the TCDA card for the affected core when the reset alarm occurs?

Have you tried checking all the connectors to be sure they are all tightly seated (ribbon cables and power cables--power cables have the individual wires)? It would be best to shut the unit down, power down the affected processor, and the unplug each cable/connector (carefully, and one at a time) and check the male pins (usually on the printed circuit card) for corrosion/build-up), and then re-seat the connector firmly. If you find any corrosion/build-up you should clean it (a light spray of contact cleaner and possibly a cotton swab gently inserted between/over the pins--you might also lightly spray a small amount of contact cleaner on the female sockets of the connector--LIGHTLY spray being the operative term). GE has also begun shipping small tubes of conductive grease to apply to the connectors when replacing cards; check in the boxes your spare cards are stored in for the TIL (Technical Information Letter) and tubes of conductive grease. At least one recent contributor to control.com has found this to be the cause of a nuisance problem (see http://www1.control.com/1026231538/index_html#1026231586).

If that doesn't work, you can try replacing the TCDA card, or swapping it with another TCDA card to see if the problem follows the card--then you'll know relatively for certain if it's the card!. (If you change the TCDA card, you don't have to download any information; the TCDA gets its info from the EEPROM on the DCC/SDCC card.)

As for the second issue, if memory serves correctly the Compressor Stall Detection is done by the IOMA "card"--which is really just a chip on the DCC/SDCC card (isn't this fun?!?!??). This author has seen this happen on one processor when an incorrect I/O Configuration is downloaded to one processor and not the other two processors. Is it possible to download a known, good I/O Configuration to the affected processor, re-boot the processor and see if this problem goes away?

As a last resort, you could replace the DCC/SDCC card of the affected processor--just be aware, when you do so, you will need to set the Voter ID of the new DCC/SDCC card using the LCC Display/keypad on initial power-up and re-boot the processor to make the change effective and then download ALL to the processor, and re-boot again to make the download complete.

If you would, please send the Diagnostic Alarm Drop Number for the first Diagnostic Alarm you cited in your post.

Let us know how the troubleshooting progresses!

markvguy

 

Thanks for getting back so quickly.

The drop for the DCC i/o card alarm is 0047.

In response to your reply, I examined all the pin connectors on the affected TCDA, no corrosion. I put the GE grease on the connectors to see if that will help. If the problem comes in again I will switch TCDA cards with another core and see if it moves.

As for the cstall alarm, I don't understand how a core could have a different IO configuration. When I change any logic I always check the mark v log for errors, then I always download (eeprom down tn core user) to all three processors, then reset the processors with no errors. In response to your reply I downloaded user to all three cores and rebooted, will see if it comes in again.

A little more background, the mark v ran great for years, no problems. Then we had a bad ground and during the troubleshooting we replaced the power supply board in the pd core. It seems like after that we had to replace 2 tcda cards (not in the above problem core) and one card in the p core. Could we have a power supply problem? Diagc shows no problems and direct measurements with a fluke show no problems.

Thanks for the help.

 

It sounds as if the procedure you use would make it difficult to get the L86_CSTALL alarm due to an improper download. The CPD (Compressor, Pressure-Discharge) pressure transducer feedback is the input to this alarm comparator. Have you looked at the Prevote Data Display values of CPD before and after this alarm is annunciated to see what the three processors think the CPD value is? Does your unit have one CPD transducer, or three? If the unit has three CPD transducers, and the calibration of one has drifted from the others, this has been known to cause nuisance Diagnostic Alarms. (Same for redundant P2 Pressure transducers out of calibration causing nuisance FPRG_INT Diagnostic Alarms.)

The TCPD card in <PD> is not a power supply, as such. It is a Power Distribution (the PD in TCPD) card. The 125 VDC and 120/220 VAC comes into the TCPD and it's distributed through the various cables/connections to the processors and other TBs in the Mk V panel. It's difficult to see how the board could be the problem. There is a large power diode UNDER the bottom of the card which could have been damaged by a severe enough ground. Did you replace the power diode? Are you sure the screws through the card are connecting the card to the power diode properly? (This has been found to be a problem when the screws weren't replaced.)

There have also been problems with battery charger output filter capacitors causing problems with the Mk V--not eliminating spikes and ripple.

markvguy

 

Hello,

Thanks for all of the help to date. Since I have reset the processors we have not received the cstall alarm on any of the subsequent start ups, (it was not a calibration problem as I had previously checked that) maybe the mark v was confused. I have also not received any i/o alarms since applying the ge grease to the connectors.But now another problem arose. I will get the 125v dc ground alarm followed by the diagnostic (drop 1360) overvoltage alarm on all 3 processors (very spurious). Could this all be the same problem just getting worse, or new because of the melting snow. You mentioned the power diode, would you know a part number?

Thanks,
Jan

 

Glad to hear that your previous intermittent problems have subsided. Sorry to hear that new gremlins have sprouted up to take their place.

Can you equate the alarms with any activity on the unit--such as at 95% speed during start-up, or generator breaker opening, or generator breaker closing, or Compressor Bleed Valve solenoid (20CB-1) actuation--or is it just more or less random? Do they occur more often at a certain temperature or time of day? Do they occur when large motors (Exhaust Frame Blowers, Cooling Water Pump motors, compartment enclosure vent fans, Aux. L.O. Pump motor) are starting or stopping?

The melting snow could be an issue (particular temperature/time of day), and generally when problems like this (grounds) do occur, the best place to look is in junction boxes and electrical compartments of devices which are mounted outside the Accessory- or Turbine Compartments, or on the roof of the compartments, or outside off-base modules (Cooling Water Module vibration switches, Gas Fuel Module, Turbine Inlet Filter House, IBH (Inlet Bleed Heat) Control Valve limit switch, etc.

Also, look for the extremely poor (and unfortunately all too common) construction practice of conduits penetrating the tops of junction boxes and not sealing properly. In general, make a sweep of all of the devices and junction boxes which are mounted outside the enclosures, opening each one, and looking for moisture. Also, conduit fittings with loose or no covers (LBs, LRs, LLs, etc.) can be a source for moisture entering conduits and getting into JBs and devices.

All but very early Mk V turbine control panels were shipped with modules in the very bottom of the control panel, in the very back of the very bottom of the control panel, a <CPF> (Conditioning Power Filter). The main purpose of the <CPF> was to filter the DC and the AC coming into the Mk V--but it also serves to prevent electrical noise from the panel from being "transmitted" out of the panel on the incoming power wiring.

Lightning strikes and serious grounds can cause problems with the <CPF>. The filter components, when they do fail, have been known to cause problems such as the ones you are describing. Also, failed battery charger output capacitors can cause this same problem. This author doesn't know of any test for the <CPF> components (nor a GE Part Number).

Have you tried putting a scope on the DC input power and monitoring it?

This author has no access to GE Bills of Materials, but, from the side of a power diode in a power distribution module: Powerex, CC410899B 06231P.

markvguy

 

Hello Again,

Thanks for the quick reply. To answer your question, 95% speed most definately. When we are shutdown the printer is quite, on a start up right at 95% speed it starts racking the overvoltage diagnostics and ground alarms. Once the breaker is shut the alarms become very sporadic. During the steady state operation I can find no event occuring that would cause the alarms. 95% is also when our Field flashing occurs, allong with all the other events that happen then. On a shutdown once the breaker opens the alarms stop. That is what makes it confusing and hard to troubleshoot. If it was just a simple ground from water issues it should be there all the time. Very frustrating!

Thanks Again,

Jan

 

You're right, a lot of things happen at 95% speed. On some units, the compressor bleed valves close. And the solenoid for the compressor bleed valve actuator air supply, 20CB-1, is usually a very large solenoid and has bee known to cause similar problems. Older solenoids had a switch inside to reduce the number of turns once the coil had been actuated (to reduce the amount of "hold-in" current required to keep the valve open), and newer ones actually have an electronic module inside to electronically reduce the hold-in current. If these newer modules get damaged (moisture or excessive heat) they can cause nuisance problems.

But, it does sound like you're on to something with the field flashing thing. Does the generator have a field ground detector relay? Sometimes failures of this relay can cause problems with the battery ground alarms, but usually only while field flashing is active, which isn't very long on units which flash the field at 95% speed.

Field flashing can draw a fairly significant amount of current while it's active. The majority of this current will be supplied by the battery charger, and the amount of current supplied by the battery charger while the unit is running is higher than when it's not. If there is a problem with the output and/or filtering circuit of the battery charger, this could be the problem that is "aggravated" when field flashing is active and when the unit is running and pulling slightly higher current to keep the 125 VDC solenoids all energized.

Are there any other solenoid-operated valves (with 125 VDC coils) which operate at 95% speed? Have you checked the generator breaker close coil circuit for grounds by removing the power and meggaring various portions of the circuit (remember there are usually several permissives which may be open at zero speed before and/or after the Mk V PTBA terminal board (presuming the unit uses the Auto Synchronization feature of the Mk V))?

Please let us know what you find! And if you have any other possible suggestions, run them by us and see if someone else has had a similar experience.

markvguy