In its most simple form, a calibration is a comparison between two measurements. The context and associated terms around calibration can get a little confusing. Before we get to those however, let’s back up. What is a comparison between two measurements? Think of it like this: You have a ruler that is 13 inches long, and you have another ruler that is 12 inches long. Each ruler is making a different measurement. You can compare one ruler to another to see which one is correct. That, in essence, is a calibration. Calibration labs (where calibrations occur) that calibrate temperature monitors (like Dickson’s!) put a unit under test and compare it with a device that can accurately tell you exactly how hot or cold a particular point is. To be more specific, it measures accurately all temperatures: 70F, 0C, 137F, etc. The key to calibrating an instrument is knowing that one device (a ruler, thermometer, gas gage) is more accurate than another device. The device that is more accurate, and can tell us exactly what 70F, 0C, 137F, etc. are, is called a standard. The device that is less accurate (our 13 inch ruler), is properly titled the unit under test. That is calibration. A standard and a unit under test being compared with each other.
You should calibrate because we assume you want to know your device is accurate. Back to our ruler metaphor. Rulers aren’t built to be 13 inches too often. But, as part of product quality, when you purchase a ruler, it would be nice to ensure it was indeed 12 inches. That is why you should calibrate. A temperature sensor is built, and you want to ensure its accuracy. When your temperature sensor is put under test, adjustments are made (like sawing off an inch of a ruler) if inaccuracies are found. Changes are made to the device to accurately meet the standard. These changes are called an adjustment.
So who says what is accurate? Who knows what 70F feels like? Who says a yard is a yard? It may seem arbitrarily argumentative, but it’s a compelling question. The answer is NIST. Who is NIST? NIST stands for the National Institute of Standards and Technology. This government agency creates and maintains standards of measurement for length, mass, time, etc. In regards to temperature, standards are sent in by their manufacturers, and those standards are told their accuracy. They then become a NIST standard, of which some variation you will find in Calibration labs.
Sadly, yes, recalibration is a word. Let’s head back to our ruler metaphor one last time. Overtime, your calibrated 12 inch ruler may experience some wear and tear. Let’s say a piece of the ruler breaks off 8 months after you have it. You of course would just go buy a new ruler, instead of trying to fix that one. But let’s pretend that ruler was a really, really nice ruler, and the cost of making it 12 inches again was cheaper than buying a new one. What do you do? You send it back to a company certified by NIST, and then they compare it with a NIST ruler, which is exactly, or nearly exactly, 12 inches. If there is a difference, they add a little bit or subtract a little bit from your ruler. This happens to devices that measure temperature as well. Over time, due to natural wear and tear, temperature sensors lose their pinpoint accuracy. Specifically, environmental factors such as dust, dirt, and humidity can negatively affect a sensor’s accuracy. This is called drift. These devices probably won’t read 95F in a refrigerator that feels cool, but they may be off by a few degrees. For many applications, that matters. So, if you have a temperature monitor, you send it to a Calibration lab (like Dickson!) and we recalibrate, and then adjust your device for any differences with our standards.