0

Editing Expanded Uncertainty Results to CMC Values

Hi All,

 

I like to know if there is a way to manually edit the Expanded Uncertainty results in the MET/TEAM/CAL database so that the report generated will show a different value? 

In my lab, we would need to make sure the reported uncertainty of a calibration is not lower than the CMC value on our scope of accreditation.

Lum

21 comments

Date Votes
0
Avatar
James Smith

MetTrak -> Asset screen -> Calibration Tab -> "Edit this calibration record" The pencil -> Show Full Results -> Edit radio button -> then make the changes you want. This will only work if you have proper permissions to do so.

0
Avatar
David Hicks

Hi Lum, The above comment by James Smith will only work if you are using MET/TRACK. If you are using MET/TEAM, it will not work. MET/CAL results in MET/TEAM are not editable.

0
Avatar
James Smith

Just so I am clear on this, you can not edit the test reports at all in MetTeam? So if for instance one has to adjust the uncertainties or remove multiple false readings, it is not at all possible? 

0
Avatar
David Hicks

Maybe someone from Fluke would like to chime in on this, but this is what we have found out since moving to MET/TEAM. You are able to delete entire rows of data from the MET/CAL results, as you would for multiple false readings. However, you are not able to edit individual result data values, such as if you wanted to adjust the tolerances (for a limited calibration) or adjust the uncertainty values.

0
Avatar
Debbie Olson

Hi,

Editing of results will be available in the next customer release later this year.

 

0
Avatar
Wei Liang Lum

Hi,

Thanks everyone!

I am just wondering if anyone has a workaround using MET/TEAM to satisfy the CMC requirement? That is, other than manually creating an excel report to amend the Expanded Uncertainty stored in the database.

Lum

0
Avatar
James Smith

Modify either the procedure to take into account additional parts, or modify the accuracy file to take into account the calibration uncertainty. The accuracy files provided are just the instruments basic specifications. You can modify the accuracy file per asset or in general. To modify the general file, just adjust the values accordingly.

The second method to do it by asset is to create a copy of the file, the adjust the beginning so that it contains the asset number. Something like the following. If you have multiple files for the same instrument and interval, it will look for asset number first, then if none is found it will take the basic non-asseted file.

Begin Header

      Instrument = Guildline 9211A

      Interval = 365 Days

      Confidence = 2 Sigma

      Asset = 123456789

End Header

 

Be careful when editing your accuracy file. This is really only advised if your comfortable with uncertainties and how MetCal is going to do its part with the values in the file. Of course this only changes the value of the accuracy files and not the report. To get a report with the correct values, you would have to run it again. ALWAYS make a back up of your file before doing anything.

I do not know enough about MetTeam at the moment as we are delaying the install for various reasons, sort of like this one. So if you can not do what you need and should be able to get done in MetTeam, then I suppose you either adjust the accuracy file and rerun, or export the report and edit with an external program. However, using an extrnal program will have records that do not correspond with one another

0
Avatar
Patrick Butler

I am going to concur with James. One of MET/CAL's powerful features is the uncertainty engine. It is designed such that the uncertainy will be calculated in real-time as you run the procedure. So your procedure is where you should account for the uncertainty components. You can use the VSET and TSET FSCs to add additionaly uncertainty components or modify the uncertainty as needed. Then the uncertainty will be calculated automatically and correctly every time you run.

Review the help file on VSET/TSET but here are a few examples.
 
VSET EXP_UNC = 2.5e-3
–ovverrides MET/CAL's built in calculation
 
VSET U3 = 1.8e-6
–Add an optional 3rd uncertainty component
 
VSET MFILE = uncertainty.csv
–Create an output file of the details on the uncertainty MET/CAL just calculated. Usefult to debug/validate calculation.
 
TSET UUT_RES = 0.001
–Create a custom UUT resolution for a one test
0
Avatar
Srilalitha

Hi All,

A have gone through lot of  posts about including uncertainty components in Procedures.

Can I request for some help on how do we incorporate these parameters in instrument specific files such that everytime I need to update the budget, I need not go to each procedure?

Like if a 5522A comes back after annual calibration , its reported uncertainty is entered "somewhere" only once so that every procedure using that 5522A looks into that palce and modifies calculations accordingly.

Sorry, I have seen the post containing such text but could not quite figure out practically how it can be implemented.

Can anyone send sample code?

Or any such file sample where in  METCAL uncertainty componenets are specified(U3,U4....)

 

0
Avatar
Michael Johnston

You could create a text file with the data you need and pull the components from that file each time you use a component in your procedure. For example,

#Establish the location of the data file
  1.001  MATH         @file = File location of text file
#For each test point, pull the calibration uncertainty from the appropriate
#line of the text file. This example pulls from line 17 of the file
  1.002  MATH         M[1] = READ(@file,17)
#Use TSET to establish the U3 component based on this
  1.003  TSET         U3 = M[1]

 

Then, the next time you send out your 5522A, you can update the text file and it will always use the most current cal uncertainty.

0
Avatar
William Green

Please feel free to correct me on this... I'm hoping Fluke will respond...

See 'Implementing ISO 17025 Measurement Uncertainty Requirements in Software' by Matt Nicholas (no date).

Just Google 'Implementing ISO 17025 Measurement Uncertainty Requirements in Software'. It'll show up.

I noticed ILAC-P14 wasn't listed as a reference. But per my conversations with Fluke, U3 (Uncertainty of the Standard) is not needed because Fluke accuracy files are listed as 99% (K=2.58).

If we add U3 to the calculations, wouldn't we also have to change the ACC file to get the correct uncertainty?

Also, I have always turned on Student T's to get uncertainties more in line with our calculated spreadsheet uncertainties. I've never fully understood or felt conforatable with this. Thanks for the post!

Please feel free to call me an idiot.

 

0
Avatar
William Green

Also, for Michael Johnston, would you also know how to collect U3 from an Excel file?

0
Avatar
Michael Johnston

My initial thought would be to use the LIB FSC to access the Excel file and pull the values that way, similar to the text file. Coding for the text file is just more simple and I haven't had the time to sit down and play with Excel and the LIB FSC as much as I'd like to (yet).

0
Avatar
Michael Johnston

Uncertainty is a very flexible thing, so while I used calibration uncertainty of a 5522A as an example, that isn't necessarily what I'm going to put in U3, or anyone else is going to put in U3. I know that I had heard a couple years back about discussions regarding the inclusion of calibration uncertainty in the Fluke specs, and you're likely right about that, but not everyone sends their standards to Fluke. And this method could be used for any number of standards, not just calibrators.

The ACC file would not need changing, since we're considering an additional contributor. The specification is what it is. If we are going to include U3, regardless of what contributor we consider it to be, it will be combined with the other contributors, including the STD Accuracy from the ACC file.

0
Avatar
William Green

Thanks Michael for your help! Very good information. Clears things up for me a lot.

0
Avatar
Michael Johnston

No problem! I'll admit I'm still figuring out some of the uncertainty stuff within MET/CAL, too, so I try to pass on when I figure things out.

0
Avatar
Srilalitha

Thanks Michael.

This is good information to start with.

So the better way would be ,

#Create a text file specific to each standard

#Re-visit Procedure containing those standards

#Include these steps involving contributors

I shall try with basic UUTs and compare manual results.

Also,

Bill,

Standard uncertainty is a mandatory contributor I think.

Please check A2LA Documents recently released which mandates Std Uncertainty and tolerance as top contributors.

However small they are , most of the auditors have asked  for those parameters in the budget.

That is what drove us to adding optional parameters(which is still in the beginners stage!)

Any corrections and suggestions are very much Welcome.

Thanks again for your replies.

0
Avatar
Michael Johnston

Bill is referring to a specific case for Fluke standards that basically, the calibration uncertainty (when done by Fluke) is included in the 99% specifications, and therefore don't need to be included in your uncertainty analysis at time of cal as a separate contributor. I know I've heard that this is the case, but I don't have a reference I can point to.

There's another discussion on this site that covers this topic as well, which might help.

https://community.flukecal.com/discussion/ilac-p14

Making a speciality ACC file would eliminate some of the coding, since it would pull the full STD uncertainty from the ACC file, but it could be tedious to make the speciality file. If you use an ACC file template in Excel to create it, though, that could allow you to have a secondary sheet to fill in the cal uncertainty and then use formulas on the main template to link them to the secondary data and then export it in the needed format after a cal cycle to update it.

0
Avatar
Srilalitha

Regarding accuracy file creation,

we are trying to upload accuracy file for megohmeter calibrator.(basically, a bank of high resistors that is not remotely controlled ).

First thing that confuses me is ,in help file, it states accuracy files cannot be used with manual standards .

When trying to upload accuracy file for this standard and run a procedure with editor,

"Error trying to convert the line 100E+3   500E+3        N/A   N/A     1   0          1E+03" in accuarcy file.

Please let me know where is the process going wrong.

 

.

 

0
Avatar
Michael Johnston

Perhaps your spacing is off and it's not recognizing the formatting because of it? How did you create the acc file?

0
Avatar
Srilalitha

Sorry for the delayed response.

Michael, thanks for the suggestion.Yes it is the issue with formatting.

I could get that sorted.It recognises now.

I created them using excel .

Please sign in to leave a comment.