Counters on Mark 5

D

Thread Starter

DAMENE

Dear friends,

On our Gas Turbine we realise 2 overhauls. the first overhaul is about the Gas Generator and is done every 50000 hours out of the site (We completely replace by a new one).

The second overhaul is done for LP shaft on site by our team every 50000 hours. (we keep the LP shaft)

On the HMI, (Speedtronic Mark 5 + cimplicity 3.22) we have an animation about the counters (fired hours and start ....)

I need to know how to reset to zero the number of fired hours for the first overhaul, as our engine is a new one.

I have to keep the LP fired hours counter.

Hope my message is clear

Best Regards
 
DAMENE,

The only way I know of to change the values of timers and/or counters is to send your .SRC and .AP1 files to GE along with the desired values and a letter detailing the reasons for the changes. They will return a .AP1 file along with instructions for downloading and rebooting to make the file "take" and "hold" in RAM & EEPROM.

Most sites put a label on the frame/bezel of the HMI with the information required to calculate the next or previous outage intervals.

OR, you could use the simple maths functions of CIMPLICITY to create a value on the CIMPLICITY Timers and Counters display that shows the information you desire using the Mark V values (for example, subtract 50,000 hours from the current value) and display that with some appropriate text.

I'm not even sure GE will modify Mark V accumulator files any more; they only want to sell turbine control upgrades.

Good luck--let us know what you do, and how you did it?
 
I never had to change Mk-5 counters during the years I worked for GE, but I have done so a couple of times since. There are several ways to do it without even being aware the GE tools exist, if you have some basic understanding of the Mk-5 and some time on your hands or good a reason to make the time to think about it. Last time I did this was in the 2005 time frame.

There are a number of blanks you will need to fill in below, if you can't fill in those blanks you don't need to be messing with the counters. Anyone with basic Mk-5 knowledge and time should be able to understand though.

For a type 1 counter: To reset the LSW, use a constant of type cnt15 and use a copy block to write to the LSW of the counter, for example accum_nn_lsw. Adjust the constant to 1 less than the desired value, force the enable point for the copy block and you will see the RAM value for the counter go to the value of the constant. Next, force the trigger point for the counter from 0=>1=>0. You should not see a change in the ram value for the counter lsw. Power off all 4 cores at the same time and then power back up, you should see the counter value is 1 more than the value used for the constant value you used above. I have never had to reset the MSW and that is a little more tricky because I am not aware of a block that can be used to write to its data type, but I would do it by building logic to cycle the counter in a controlled manner to roll it over. Its brute force but should work and would be quicker than even having initial conversations with GE to get them to do it for you.

I found that the above logic must be in a CSP segment executing at at least 8hz, I do not remember why its that way, but back then I recall realizing why that was the case, don't recall at this moment. It was a long day, but got it done.

I have also reset other non-Type 1 counters using the procedure above, first changing them to type 1, reset the value and then change the type back. There are other more elegant methods along the above lines that could be tried to change the value of other counter types, I have not tried any other methods though.
 
The Mark V keeps a copy of the values of timers and counters in RAM, and another one in EEPROM. I believe the EEPROM value gets updated approximately every sixty minutes, but if I recall correctly that period also changed some time during the production period of the Mark V from a longer value to the sixty minutes.

The reason I mention this is that for this method to work, one has to make sure the EEPROM values gets updated BEFORE a processor gets re-booted. The change is being made in RAM, and unless that change makes it into RAM, it's lost.And, I'm not exactly sure how to do that (make sure the EEPROM values of timers and counters is updated from the RAM values). About the only way I could think of would be to actually re-boot the control processor(s) and see if the EEPROM value copied into RAM was the desired value. So, this could take several re-boots if one is not patient. Not to mention the re-boots required to make the sequencing change to add the blocks required.

Further, this needs to happen for all three control processors in a TMR Mark V, and almost simultaneously because the designate processor's value will "win" if not done properly. (I believe the designated processor's values of timers and counters is the one copied into EEPROM for all three control processors just as a control to make sure all three processors are the same.) I would suggest this method is not for the faint of heart, nor is it for someone only casually familiar with the Mark V and not more familiar with the intricacies of RAM and EEPROM and writes and boots and such.
 
Top