I'm trying to modify the units of an already defined control constant. The constant is AKPAP_Site. This should be 14.7 psi. The units are specified in Const_Q.SRC as CNT04. When this is added to the pressure transmitter (320 psi) that has units of psi the result id 2047.9 instead of 334.7. It is messing up the mass gas flow calculation. I tried changing it in Const_Q.SRC to psi. Then when it was complied there was an error that says AKPAP_Site is psi, should be CNT04 setting value to zero. Sure enough after download the output of the block was 320 psi. The unit was put back to CNT04 and the output went back to 2047.9. I think there must be another file that has to be changed but I can't figure it out. I also have the problem for an analog input that I want to specify as psi and when compiling it says it needs to be CNT15, which is what the spare ma input was originally defined. Anybody know? Thanks for any help.

Andy

 

Andy,

My question is: Is the Control Constant used in any other calculation? If so, changing the units may adversely affect the other calculation.

So, before we go down this path and possibly cause other problems, let's answer this question.

When you say the spare mA input was originally defined as CNT15, where was that definition? Did you change it? What did you change it to?

 

Andy,

Sect. 8.5 of the Mark V Application Manual, GEH-6195, describes the entire process of "adding" analog inputs, including mA inputs, to a Mark V.

 

Andy,

The very FIRST thing you are going to have to do is <b>MAKE BACK-UPS OF THE UNIT1 FILES!!!!</b>

The second thing you are going to have to do is understand that the Mark V does NOT use floating point math; it uses fixed point math. Later versions of the Mark V Application Manual, GEH-6195, have an appendix section that describes the Mark V's floating point math reasonably well.

What does this mean? It means that you can't just perform mathematical operations on two numbers (which is the maximum number of numbers (???) the Mark V can operate on at a time per rung), without understanding what the result will be. It's not as complicated as it seems, but it's not simple either. CNT04 is a B4 number, just for reference (it's maximum value is 16 which is 2 raised to the power of 4). I'm only mentioning this because if you change the scale code/type and the signal is used in another calculation it may likely cause unintended problems with that, and subsequent, calculations.

The third thing you need to do is use your text editor to open ENGLISH.SCA and search for all the occurrences of PSI. You should find something similar to the following, depending on the PROM revision in use at your site:<pre>
#scale_data 102 16384.000000 0.000000 0 psi PRESH
#scale_data 103 32.000000 0.000000 2 psi PRESL
#scale_data 104 256.000000 0.000000 2 psi PRESM
#scale_data 105 2048.000000 0.000000 1 psi PRESS</pre>
The second-to-the-last column lists the engineering units that will be used for displaying the signal's value, and the last column lists the name of the scale code/type. You sill need to choose an appropriate scale code/type, as "psi" is not an appropriate scale code/type. (The third to the last column lists the number of decimal places for the engineering unit's display value.)

If you look elsewhere in the file you will see that CNT04 has a maximum value of 16.00000, which is 2 raised to the power of 4 (2^5). PRESL, which is the lowest pressure scale, and which would be suitable for 14.7 psi, has a maximum value of 2^5 (32.00000).

Presuming AKPAP_Site is not used for any other calculation, it is "defined" in one of three possible places, one of which MUST NOT be edited. UNITDATA.TPL is a file that is used to help create UNITDATA.DAT, which is the file that is used for the signal definitions for a Mark V panel. UNITDATA.TPL is the file which MUST NOT be edited, because it has to match the contents of the PROM chips on the DCC/SDCC cards.

The other two locations you might find the definition of the Control Constant are FACTORY.ASG or SITE.ASG. In fact, you might find it in both places (though you shouldn't, it's still possible). They are both ASCII text files which can be searched with any text editor.

If you find it in either file, it will be on a line similar to this:<pre>
?VQCC AFPAP_SITE CNT04</pre>
This line tells the Data Dictionary builder to find a spare Control Constant in <Q>, name it AFPAP_SITE, and to use the scale code/type CNT04.

You will need to edit the line to read something similar to the following:<pre>
?VQCC AFPAP_SITE PRESL</pre>

You will then need to edit the entry in CONST_Q.SRC to read:<pre>
AFPAP_SITE 14.7 psi</pre>

At this point, you have changed the scale code/type used for the definition of the signal AND you have changed the engineering units in the Control Constant definition file.

Then, you can run MK5MAKE.BAT to create the new UNITDATA.DAT, and you will need to select Y to recompile the sequencing (since the scale code/type has changed), and then you will need to download USER to all three control processors and re-boot them one at a time for the changes to become effective.

I also subscribe to the theory that every signal that is in the <Q> CDB (control signal database) is a subset of the <C> CDB, so I would also recommend downloading to <C> and re-booting <C>, as well. (Beware, that when you do this, several motor-driven fans and pumps will likely start and run while the <CD> relays are dropped out during the re-boot.)

Lastly, you are going to have to "re-boot" the Mark V operator interface. If it's an <I>, a simple re-boot will do, or you can just type RUN_IDP at any command prompt to do the same thing. If it's a GE Mark V HMI, you're going to have jump through more hoops unless you just re-boot the PC in order to re-start TCI and CIMPLICIY.

That should do it for the Control Constant issue--presuming the Control Constant isn't used for any other calculation! If it is, then changing it from a scale code/type of B4 (2^4) to a scale code/type of B5 (2^5) may cause other problems--but you have been warned.

Lastly, if you find the Control Constant in BOTH files, well, that's shameful, but not your fault and easily overcome. MK5MAKE.BAT was written to use the definitions in SITE.ASG <i>after</i> those in FACTORY.ASG, so just edit the entry in SITE.ASG to match the above (presuming you use the PRESL scale code/type) and then do the MK5MAKE and downloads and re-boots of everything and you will be fine.

If you have multiple Mark V operator interfaces, you are then going to need to copy the files you changed (which should be SITE.ASG and/or FACTORY.ASG, and CONST_Q.SRC) to the other operator interfaces and run MK5MAKE. At that point, all you have to do is re-boot the operator interface, since you've already downloaded to and re-booted the Mark V with the same information. (If you make changes to IO.ASG and IOCFG_x.DAT or IOCFG_x.AP1 for the mA input you are trying to scale properly, you will need to copy those files as well to any other Mark V operator interfaces and then run MK5MAKE and re-boot the operator interfaces.)

 

Andy,

If you don't find the Control Constant in either FACTORY.ASG or SITE.ASG, write back.

There is a "nuclear" solution (just kidding).

 

CSA,

Thank you for all your posts. I appreciate all your time and knowledge that you share. I’ve been busy so haven't had a chance to digest all your post before now. First you were correct that the control constant was used other places and by changing the scale type faulted those calculations. I created a new control constant to use in the mass flow calculation instead.

The milliamp input I was having trouble with is on a machine that is running and we'll be taking down soon. I wanted to add a 4-20 ma input 4-20 ma = 0 – 2.5 psi. I did follow the 6195 instructions and every time I did a MK5MAKE the "should be CNT15" error message kept coming up. I'll get back to it in the next few weeks and let you know if I still have the problem.

Thanks again for your help.

Andy