The Truth, The Whole Truth, and Nothing But The Truth.
These days it seems uncovering the truth is getting harder and harder, day by day. Fake news? How do I know if it’s fake or not? False social media accounts. How would I tell the difference?
Truth and the concept of truth are becoming devalued. To tell the truth. Truthfully. If truth be told. Meaningless words and phrases.
However, there is one matter on which I do feel qualified to comment and that is technical specifications. Sometimes, to put it kindly, the writers might be better employed on works of fiction.
Here’s my problem with technical specifications.
In any sensor-based NDT test methods, including leak testing the accuracy of sensors will inevitably have bearing on the test results. I think we can all agree on that. But sensors alone are never the whole story. It’s the accuracy of the entire leak testing SYSTEM that needs to be considered.
Too good to be true is almost always true!
There are actually some telling ways to identify sensor accuracy reports that seem like they are “too good to be true” because they ARE “too good to be true”.
Any NIST traceable sensor would always be calibrated by an independent calibration device that has four or five times the accuracy of the sensor being calibrated. Ask if you can see the published and third-party independently verified calibration standards. Better yet, ask if you can watch such a calibration at work in the instrument manufacturer’s facility. If such a visit is off-limits—walk away!
A second and most important question to ask directly and bluntly about any published specification is if that accuracy is ONLY for the sensor or if it is for the entire system that the sensor works within. Think about it. What’s the point of having the most precise sensor if the manufacturer has not paid appropriate attention to A/D conversion, resolution, and software validation?
Now, take your “best in class” sensor that’s inside your instrument and integrate it into an automatic system. What happens when the system designer has allowed loose tolerances or sloppy part handling and sealing into the system? That’s where the REAL discussions on your application should begin. Believe me, these discussions can sometimes be uncomfortable and reality is not always welcome. But, as they say, truth hurts.
I recently received a call asking if we could achieve measurement accuracy a full order of magnitude higher than that used by the lion’s share of medical device manufacturers whom USON has helped with leak testing solutions for over half a century. If I was going to be flip I’d say—“Sure, we can get you that accuracy if you want each test cycle to last about a week.”
However, I don’t want to be flip.
I do want to say that realism matters.
Instrument specifications that make claims such as accuracies taken out of real world contexts are nothing more than works of fiction for managers who want comfort instead of objective sourcing of best-in-class NDT methods that align with the laws of physics.
I rest my case.