When a device is calibrated, you need to know that the reference equipment is also correct within certain tolerances. This sets up a hierarchy of calibration, where you can trace your measurements up to a higher standard. The reference equipment will itself be calibrated behind the scenes, and the documentation allows you to follow that hierarchy typically to a national standard. In practice, all calibrations will have some form of traceability.
Most measurement instrumentation will be built to a national or international standard – for example, BS EN 61672-1 defines the characteristics of sound level meters. A ‘calibration standard’ will define how that instrument is calibrated, so in this case, BS EN 61672-3 documents the procedures for calibration. Very often, calibration standards don’t exist for some types of instruments, so laboratories will perform a series of tests, using their own procedures, to check conformance with the instrument standard.
The standards are not proscriptive on this, but as a rule of thumb, instruments should be re-calibrated every two years, and calibrators annually. For example, the recommendation in BS 4142:2014 follows this approach for industrial noise measurement. However, if your measurements are really critical, you may wish to calibrate more frequently.
This depends on what is being calibrated. A four-channel instrument will take longer than a single channel one. A complete sound level meter will take longer than a microphone. Also, a lot of time is spent in the laboratory setting up for a specific instrument type. You can minimise the turnaround time by planning with the laboratory in advance – for example, if several sound level meters of a particular type are to be done, the tests are set up just once, saving time. Also, let the laboratory know the timescales – there’s no point sending in a meter if their input shelf is already full.
Modern instrumentation is very accurate and reliable. But it’s not perfect and can go wrong, sometimes in ways which are not completely obvious. This is exacerbated by the fact that precision instrumentation is used in hostile environments, such as building sites, where rain, dust, handling and corrosive substances can affect your measurements. For example, dropping an accelerometer on the ground may break a crystal – the device will still have an output, but it may not be as linear as it was when delivered. A calibration will allow you to check this and have confidence going forward. Think of it as a kind of MOT test for your instrument.
When you are measuring something accurately, you need to know that the results are really what they say they are. Therefore, it’s necessary to compare your measurement capability to some reference. This might be a simple hand-held calibrator, or reference to an absolute standard in a National laboratory. The choice depends on the purpose for which your measurements are to be used.