I'm not really sure if these questions are applicable to Labs that use the military K, NAVAIR, etc. procedures as when I was in the Navy, we always had a cal procedure for anything we recieved for. But I would certainly like to hear opinions, experience and views of those that can only use military cal procedures.
My intention is not to bash OEMs, I worked for the world leader in test equipment and loved that job and company.
When we receive a new item in for calibration we look at the OEM performance test procedure and use the same testpoints they do when we develop our test points. Some of my concerns:
The OEM often does non-sensical rounding of tolerances, sometimes up, sometimes down (although this is not as prevelant as it once was)
What do yo do? Use the OEM performance test rounding limits or use the ASTM E29 or application calculated rounding?
The OEM often does not test parameters that are specified. These are usually ACV and ACI ranges.
What do you do? Use the OEM procedure knowing you are not testing every parameter of the UUT? Use the OEM procedure in addition to your additional testpoints to test every specified parameter?
The OEM suggests standards that will not meet 4:1 test ratio requirements in their performance tests. (This seems to be a ridiculous queston but it is real world)
What do you do? Use it anyway and not report less than 4:1 test ratios?
Use it anyway and report less than 4:1 test ratios?
Ignore their suggestion and use a standard that meets 4:1?
As an OEM of equipment if there isnt another "Accredited/Fed Spec" procedure follow the OEM procedures, at a minimum. This doesnt mean you couldnt test more functions of the equipment, although if you create this procedure you have to validate it. If you do have questions/concerns by all means you should call tech support or get engineering on the phone from the OEM. ALWAYS test, at a minimum, what the OEM tests in order to eliminate any responsibility should something happen with the equipment during use and pending legal action occurs. Unfortunately most OEM procedures are written by engineers who have no idea what specs are required for a Z540 or 17025 cal requirements.
Quote from: WestCoastCal on 02-20-2008 -- 11:46:16
I'm not really sure if these questions are applicable to Labs that use the military K, NAVAIR, etc. procedures as when I was in the Navy, we always had a cal procedure for anything we recieved for. But I would certainly like to hear opinions, experience and views of those that can only use military cal procedures.
My intention is not to bash OEMs, I worked for the world leader in test equipment and loved that job and company.
When we receive a new item in for calibration we look at the OEM performance test procedure and use the same testpoints they do when we develop our test points. Some of my concerns:
The OEM often does non-sensical rounding of tolerances, sometimes up, sometimes down (although this is not as prevelant as it once was)
What do yo do? Use the OEM performance test rounding limits or use the ASTM E29 or application calculated rounding?
The OEM often does not test parameters that are specified. These are usually ACV and ACI ranges.
What do you do? Use the OEM procedure knowing you are not testing every parameter of the UUT? Use the OEM procedure in addition to your additional testpoints to test every specified parameter?
The OEM suggests standards that will not meet 4:1 test ratio requirements in their performance tests. (This seems to be a ridiculous queston but it is real world)
What do you do? Use it anyway and not report less than 4:1 test ratios?
Use it anyway and report less than 4:1 test ratios?
Ignore their suggestion and use a standard that meets 4:1?
Uhhmm, do you mean a 4:1 T.U.R if you do then you are required by ISO 17025 to state the Total Uncert Ratio. So yes you can do it legally as long as you list the T.U.R. , You don't go by your standards MFR Specs either, you should be using your expanded measurment uncert. Maybe I am not real clear too what you are asking.
I use far more OEM than other procedures. Some thoughts
1. I trust the OEM will have a better idea of what's going on inside the instrument such as where to check for spurs ar other things that might get out as a result of mixing or what not.
2. "Typical" is something that I am seeing more often in specs, Agilent defines it as a characteristice that 80% of the population of the model will achieve. Sometimes the parameter is tested, sometimes not.
3. ACV/ACI, I recall Fluke stated fro the 8842A trhat if it met DCI and ACV that ACI would also be good, I took them at thier word.
I like to think that the OEM procedure is a result of input from the design engineers in consultation with the metrologists at the company but I alo like to think the easter Bunny hid candy around the house when I was a kid.
I have seen it both ways though, OEM procedure being more in depth than the TB and vice-versa. OEM procedures are more likely to have typos IMO.