blog 3 stupid things that can wreck an otherwise excellent calibration Janice Bennett-Livingston Share Published: Mar 12, 2013 Industrial Manufacturing and Processes Industrial Measurements Life Science This week's blog comes from our calibration and repair department: The top three calibration issues listed below aren’t new problems, but in our experience, they are definitely issues we have to be diligent about. Not looking at the Customer requirements Transcription Errors Inadequate checks of your reference standards Not looking at the customer’s requirements: Unfortunately, this happens. Sometimes, if the device has built-in memory where the information exists, you can take a step back and adjust to the original calibration. But, with most instruments, once you’ve adjusted, you can’t go back. Nightmarish situation, especially if you are also failing at Stupid Thing #3. Transcription errors: Any manual process comes with the risk of human error. It’s easy to put the decimal in the wrong spot. There are two main safeguards: 1. Put checks in place. Look at the results carefully to see if they are realistic, and, 2. Automate processes as much as you can to eliminate data entry. Inadequate reference standard shecks: Obviously, reference standards drift too. The quality of your standards depends on your knowledge of the uncertainties of the device – the type of instrument, how susceptible it is to drift, your own operating environment, etc. Measurement and calibration are simply not meaningful without understanding the measurement uncertainties of your standard. Resources: Here’s a good presentation from the government of Hong Kong’s Innovation and Technology Commission. A nice refresher course: “Evaluation of Measurement Uncertainty.” Also, a good document out of the UK: “The Expression of Uncertainty and Confidence in Measurement” from the United Kingdom Accreditation Service. Do you have your own favorite CE that I have missed? Email us!