All

We have 2 x Mark V panels (GT and ST) + 2 nos. HMI (Windows NT 4.0 based running Cimplicity 4.0). The GT HMI is acting as a Master Time Keeper ie. installed with PCI-card connected to a GPS module and supposedly "broadcast" a time signal to everyone in the Arcnet Stagelink network.

At the moment, Mark V panels and both HMIs having different timestamp (3 different times)

Below is the F:\TIMESYNC.DAT output:<pre>
******** GT HMI (TIMESYNC MASTER) ********

TIMESYNC HIGHRES
LOCAL_TIMESET ENABLED
TIME_LOAD LOCAL
I_TIME UTC
MARKV_TIME UTC


******** ST HMI (TIMESYNC SLAVE) ********

TIMESYNC SLAVE
LOCAL_TIMESET ENABLED
TIME_LOAD NETWORK
I_TIME UTC
MARKV_TIME UTC</pre>
Can someone shed some light? According to GEH-6126, there’s no such setting as "TIMESYNC HIGHRES". The syntax is TIMESYNC <controller> MODE <mode> [LEVEL_SHIFT]

 

any help pls...

 

Hello

I suggest to read GEH6370 may be your solution is inside.
what the result of the command :timeutil status ???

bye

 

i never know the existence of GEH-6370.

in my record, i have GEH-5979D, GEH-5980E, GEH-6126 and GEH-6195D.

do u mind sharing it? u can upload it using sendspace.com for example or any other file-sharing website.

Regards.

 

Hello,

Please send me your e-mail adress.
What about the result of timeutil status command

 

Ensure that the time signal output from the satellite clock goes to the correct bnc connector on the PCI clock card.

Open the TCI Applet and change the Timecode accordingly and set it to be on High Resolution Time Master.

Reboot and Open Windows Explorer and go to C:\Program Files\Datum\BC635PCI Demonstration Software\BC635CPP.exe

Open the exe file and you can see the time card status. If all the setting are correct, all indicators should be green. This would take some time so please be patient

If all indicators are not green connect the Time Code Output to the Time Code Input on the time card. The indicators should now be green. This indicates the signal from the satellite clock is defective or the input type definition is incorrect on the TCI setting.

Check this out and tell us what you get.

You can also check StageLink Time Sync Status by opening a CMD Window and enter TimeUtil Status

Good luck!

 

ok nevermind. i managed to get hold of GEH-6370 finally.

ok below is what i get from TIMEUTIL STATUS command.

pls see link below:

http://b.imagehost.org/view/0438/st1_svr_timeutil_status_151210

http://b.imagehost.org/view/0924/gt1_svr_timeutil_status_151210

notice that i have no Stagelink Time Master.
also my Master HMI (ST1_SVR) is not connected to the GPS clock.

 

ok i got more info.

i stop TCI service. then reboot my HMI. still no luck. "TIMEUTIL STATUS" command still gives me GTS not tracking.

i've verified my TCI applet settings. i also disconnect and re-connect the IRIG-B coaxial cable to the GPS panel.

i run Datum software as instructed, the tracking status is red color.

below is my TCI applet and Datum software shot

http://b.imagehost.org/view/0774/gt1_svr_bc635cpp_151210

http://b.imagehost.org/view/0170/gt1_svr_tci_settings_151210

 

What signal are you getting from your GPS clock?

Is it correctly defined at the TCI applet and connected to the correct BNC connector at the HMI time card?

A quick way to find out all these is to connect Time Code Output to the Time Code Input at the HMI time card. Try reboot the HMI and run the BC635CPP.exe program. If all the indicators are green, then most probably the signal definition in the TCI or the connection is wrong. There is also a possibility of the GPS clock output to be defective as well.

Give it a try and send your result. All the best!

 

The card is getting IRIG-B signal from the GPS clock module.

I've tried a simple looping ie. Time Code Output goes to Time Code Input. I then run BC635PP.exe utility but still seeing tracking LED showing red.

any idea sir?? possible that my card is defective or the outgoing connectors perhaps?

Regards.

 

dinster,

I've been following this thread for a while.

When did this problem start? Because, since these are Mark V panels and GE Mark V HMIs, they've been in service for a while. So, it would seem that they were all working for some time, and then something changed and now they're not working. Is this correct?

It seems to have started shortly after a possible daylight saving time change; is this true?

Did someone recently perform a MS-Windows update? Or an update of anti-virus software? Or something like that? Was some new software recently installed on the GE Mark V HMI?

There is a file in the root of F: drive, TIMEZONE.DAT. I believe the executable that creates the file (G:\EXEC\TZ_MAKE.EXE) looks at the MS-Windows Time/Date settings and configuration and then creates a file that is used by the TCI Time Synch utility when TCI is started/re-started.

You have said that one of the HMIs is the "Master". That HMI should have the time synch card and the time signal input should be connected to that card, and it should be configured as the Time Synch Master. Can you verify this?

If it was working and suddenly stopped working with no operator action (such as a daylight saving time change) then it would seem that something is physically wrong with the "system". Have you checked the cable from the time master to the HMI? Have you checked the multi-headed cable? (I don't know if there is a pin-out diagram for the cable, but there has to be continuity from the input cable being used to one of the pins of the D-sub connector; if not, then something's wrong with the multi-headed cable.

Or, it could be the time synch card.

Again, if it was working and stopped, it's likely either something that was recently done to the system (including someone performing maintenance on the HMI, moving it to clean under/around or even inside the CPU case--perhaps the time synch cards needs to be re-seated in the slot?), or something done to the PC or HMI (such as a time change using the MS-Windows Time/Date control panel apple which may have caused an issue with the TIME_ZONE.DAT file; running TZ_MAKE in the root directory of the F: drive from a command prompt won't hurt anything, presuming the Time/Date settings are correct), or something is physically wrong (cable; connector; card; card slot; motherboard; etc.).

Of course, if manual changes were made to the NTP settings or something else that hasn't been imagined, then the above may not be applicable. But, if it was working and now it's not, it shouldn't be this difficult to troubleshoot and resolve. But, we don't know what happened (if anything) immediately prior to when it was noticed to be not working, and that is one of the most important questions to be asked whenever one is troubleshooting a problem that has appeared on a previously trouble-free system: What has changed? (If anything!)

Hope this helps!

 

CSA

1. it started last month out of a blue moon with no changes done (unless someone made changes without my knowledge). our system has been in operation since 2003.

2. daylight saving time is not applicable to my site.

3. no windows update performed. it's still Windows NT 4.0 as shipped by GE. No antivirus software installed no anything.

4. i have 2 HMIs and 2 Mark V panels. GT-HMI and ST-HMI. GT-HMI has got BC635pci time card from Bancom (aka Symmetricom) and connected to a GPS Satellite module from Arbiter System using RG-58 coax cable.

i haven't got chance to test the RG-58 cable as i don't have spare cable. i also don't have a spare multi-headed cable so i cant test now.

 

At a minimum, a continuity check could be done on the RG-58 coaxial cable (it has a single conductor running through the middle of it, and that conductor should be insulated from the metal BNC connectors at each end).

Also, the multi-headed cable could be removed and a continuity check could be made from the single "head" the RG-58 coaxial cables is terminated on to one of the pins of the connector. There should be continuity to at least one of the connector pins, and it should also be insulated from the connector shell and any other pins.

I believe the time synch card can be set to "free-wheel" and generate a time signal that is broadcast to the other nodes on the StageLink (Mark Vs and HMIs). And I believe that's done through the TCI Control Panel Applet, and that the GE instruction book you have should have some instructions for setting up the card to provide a time signal in the absence of receiving one from the master time source.

I even think the PCs time time signal could be used as the time source for the time synch function, and that the GE instruction book might have some instructions for doing that.

Either would be better than nothing, and it would also prove the time synch function works.

I also wonder about WinNT 4.0.... It's an aging OS and it might be worth checking with GE to see if any MS updates could be applied that might help with the problem.

And, you didn't answer the question about the TIMEZONE.DAT files.

Lastly, have you checked the individual MW-Windows Time/Date Control Panel applets to make sure they are all set for the proper time zone and not for automatic daylight saving time adjustment?

 

Dear CSA

i wish to inform the problem has been resolved. It was the multi-headed cable connecting my PCI timesync card and the GPS clock module. It failed continuity test and we checked it was loosened!!

i'd like to thank you very much !!

 

dinster,

Thanks for the feedback!

It's always a pleasure to be of help.

 

Hi

WE have 4x MarkV panel STG Mark V hmi <I> is master and other are slave. Whenever we got our timeset. After 2 days our time goes back again to 3 minutes less then the current time. Is their any solution. We have DOS based MarkV HMI?

Waiting for your kind help...

 

syed_hassan08,

It's been a VERY long time since I've worked on an <I> Mark V operator interface, so I may be a little rusty on a couple of points.

You didn't tell us if there is a master time source (GPS; IRIG satellite receiver; DCS; etc.; sending a signal to one or more <I>s through a TimeSych card in the <I>(s)). Some sites just used an <I> as the master time source for the Mark Vs, periodically broadcasting a time signal on to the StageLink to adjust/set the Mark V turbine control panel time(s).

The <I> time is kind of a mess.... There is the PCs BIOS time (kept by the battery on the motherboard, and adjusted with the DOS TIME command). There is IDOS time, which the IDOS keeps while it is running on the PC that is used as an <I>. This is a background task of IDOS, and I believe that when IDOS starts it copies the PC's time to the IDOS clock and then keeps its own time (IDOS time) as long as IDOS is running.

And then there's the time on the Mark V turbine control panel.

And all three times can, and usually are, different.

The DOS time is viewed by opening a command line session and typing TIME and pressing ENTER.

The IDOS time can be seen by looking at the time in the upper right corner of the IDOS Main Menu (NOT the Main Display--the Main Menu). This time on the Main Menu usually has a cyan (light blue) background.

The time on animated displays (like the Main Display) and on User-Defined (Demand) Displays is the time in the Mark V that data on the screen is being retrieved from.

I believe what happens (and you can test this) is that when Timeset is run it copies the PC BIOS time to the StageLink and to all Mark Vs on the StageLink. (It either copies the PC BIOS time, or IDOS time, but I believe it's the PC BIOS time.)

If there is one or more time synch cards installed in <I>(s), and they are either in "freewheeling" mode (meaning they are generating a time signal without any input from an external, "master" time source), or they are getting a time signal from a master time source then I don't believe one is supposed to use Timeset--because it doesn't use the time from the TimeSynch card.

I don't remember if <I>s (meaning IDOS clocks) get time from a Master Time source or not if available; but I seem to recall they didn't in early versions of IDOS but did in later versions of IDOS; isn't this fun?).

So, if there is a "master" <I> time "server" then I don't think one should be using TimeSet to send time commands--as the master <I> time server is doing that already, and because I believe the time for TimeSet is not synchronized to the master <I> time server's time that could explain what's happening.

TimeSet was to be used to set the time in the Mark V when TimeSynch was not used. I don't believe it should be used when TimeSynch is used.

And, be careful about the "clocks" on <I>s--remember there are two (PC BIOS time, and IDOS time). And both of these times are quite frequently different (maddening, but true)--and they can both differ from the time on Mark V turbine control panels.!.!.!

If you can be more specific about how TimeSynch is configured at your site, where the master time source comes from, which <I> is being used to send TimeSet commands, and what the PC BIOS time and IDOS time are on the <I> sending TimeSet commands, then we might be able to help. But, I still don't believe that one should be using TimeSet if TimeSynch is in use. (The line with that command should be commented out (by placing an asterisk at the beginning of the line in MENU.DAT; no need to reboot the <I> or re-start IDOS; just type MENU at the command prompt to return to the IDOS Main Menu when finished editing the file).)

Hope this helps; sorry I couldn't be more specific. But you can look at the various times and do some tests (it shouldn't cause any operational problems) to determine which time TimeSet sends to Mark Vs--but remember, that if TimeSynch is enabled and working it will eventually overwrite that time if it is different from the master time source's time.

(Isn't this fun?!?!?!?)

 

syed_hassan08,

I also wanted to mention that just a few days ago I learned of a GE Mark V HMI (running MS-Windows and CIMPLICITY) with a time synch card had developed problems with time synching over several months. Ultimately, it was determined that the multi-headed cable attached to the time synch card and to which the time source signal was connected had developed an intermittent connection--most likely because of the weight of the time source cable which was unsupported. (A large coil of coaxial cable was left hanging, supported only by the single port of the multi-headed cable.)

It was possible to change the time source signal to a different period and reconnect the cable to a different port of the multi-headed connector--after supporting the weight of the coaxial cable better.

And this was similar to one of the other problem resolutions noted previously in this thread.

Don't know if this will help, as we didn't talk about when this problem started and if time synch'ing had been working properly for some period. There's also the TIMEUTIL command utility which can be used (as noted in previous responses to this thread) to determine if the time synch card is getting a signal from the master time source (if used).

 

Refer to same as Syed Hassan mentioned, we investegated the problem a little bit and found that the problem is with DCS signal toward MK-V. But I didn't understand why their is an internal delay of 27msec in upgrading the time of MK-V and the surprizing thing is that this accomulate with the passage of time. I mean last time someone told me that they have updated the DCS time in Dec-2012. And you see its almost 15 months ago. If we calculate the total time is almost 4 Min, and the delay is also 4 min. We are using the Time sync card in STG MK-V <I> and from here time is sent to other <I>'s. The ref time signal from DCS is terminated at 1PPH port of Time Sync card in MK-V <I>.

So, now we are thinking to use some Master Clock for all DCS and 4 MK-V system. for time synchronization. Have anyone experience with this type of problem's?

 

We have solve the problem of MARKV time sync by adjusting the time through the DCS by giving digital time ouput to the Markv Time sync card after every hour in the form of pulse. Now our time is running ok.