AERO61021 Chap.4 Calibration Standards and Resolution
Calibration Standards and Resolution
Calibration Standards and Resolution explains how an instrument reading becomes a trustworthy physical value. Experiments almost always record a proxy, typically a voltage, so the first task is to establish the function that maps that reading onto force, pressure, position or temperature.
The chapter treats calibration as a measured relationship rather than a single number: you vary the input in controlled steps, log the output and fit the law that links them, which may be linear, curved or frequency dependent.
It then follows the pedigree of that law, from your sensor to a laboratory secondary standard, to a national body and finally to the seven SI base units, and explains why fewer links mean lower uncertainty. The practical record keeping that makes a calibration auditable is covered alongside the habit of storing unconverted voltages.
The second half separates resolution from accuracy, converts full-scale percentage figures into absolute steps using the measurement span, and introduces spatial and temporal resolution, including the trade a high-speed camera makes between image size and frame rate. Worked numbers throughout use invented devices so that each step can be checked by hand.
What this chapter covers
- 01
Calibration as a fitted relationship between reading and quantity
- 02
Linear and non-linear calibration laws for potentiometers
- 03
Traceable calibration chains and secondary laboratory standards
- 04
The seven SI base units and national measurement bodies
- 05
Calibration records and storing raw voltages
- 06
Resolution versus accuracy and percentage of full-scale output
- 07
Spatial and temporal resolution in point and camera measurements
Worked example · free
Convert a potentiometer voltage with the right calibration law
- 1Derive the linear calibration coefficient from the span: 40 mm divided by 5 V gives 8 mm per volt.
- 2Apply it to the reading: 8 mm/V multiplied by 1.6 V gives a position of 12.8 mm.
- 2Evaluate the logarithmic law: 1 + 1.8 × 1.6 = 3.88, log10(3.88) = 0.5888, and 40 × 0.5888 gives 23.6 mm.
- 1Compare: the linear coefficient would understate the logarithmic device by about 10.8 mm, more than a quarter of its 40 mm travel.
Key terms
- Calibration curve
- A calibration curve is the measured function that converts an instrument reading, usually a voltage, into the physical quantity required.
- Traceable calibration
- A calibration linked to a national or international standard through an unbroken sequence of comparisons, each with its uncertainty stated.
- Secondary standard
- A secondary standard is a laboratory reference device of known higher accuracy against which working instruments are compared.
- Full-scale output
- Full-scale output is the measurement span of a device, used as the base when resolution or error is quoted as a percentage.
Calibration Standards and Resolution FAQ
Is calibration just multiplying by one number?
Only when the device responds linearly and the measurement is steady. Many sensors follow curved laws that need a quadratic, cubic or other fitted function, and time-varying measurements can need a full transfer function because the response changes with frequency. The safe description is a relationship between reading and quantity, established by stepping the input and recording the output across the whole range.
Why should I save voltages instead of calibrated values?
If a coefficient later turns out to be wrong, every converted file inherits the error and reversing it is very hard. Keeping the unconverted readings, together with a record of when, where, by whom, against which standard and under what conditions the calibration was made, lets you reapply a corrected law in minutes and defend each number.
How do I turn a percentage resolution into a real value?
Multiply the quoted percentage of full-scale output by the span of the device, meaning the difference between its upper and lower limits. A 500 to 5000 N load cell therefore uses 4500 N as its base. Then confirm the acquisition hardware can distinguish a voltage step that small, because noise or converter limits may set a coarser practical floor.
Assessment move
Treat this chapter as two linked skills: proving where a number came from, and knowing how fine that number can be. Start by sketching the calibration chain for an instrument you have used, naming the level at each link and noting that every comparison carries its own uncertainty.
Practise explaining in two sentences why a shorter chain supports a more precise calibration and why a laboratory still recalibrates its secondary standard on a schedule. Next, take any sensor data sheet and identify the calibration law it implies: a single sensitivity for a linear device, or a curve that needs fitting.
Work through conversions with deliberately mismatched laws, as in the potentiometer comparison, so the size of a wrong-law error becomes intuitive rather than abstract. Build a one-line calibration log template with date, location, operator, reference standard and conditions, and use it on every laboratory session in the module, including the group project.
For resolution, drill the span rule until it is automatic: subtract the lower limit from the upper limit before applying any percentage, and express the answer in sensible units such as pascals, micrometres or newtons. Pair each answer with a check on whether the acquisition card could resolve the matching voltage step, since a fine sensor connected to a coarse converter is no better than the converter.
Finally, rehearse the difference between spatial and temporal resolution using one camera example and one point-sensor example, stating in each case what limits the measurement and what you would trade to improve it. Explaining these trades aloud to a study partner is a fast way to find gaps, because the wording of resolution and accuracy is where most confusion hides.
A useful closing exercise is to audit one dataset you have already recorded: can you name its calibration source, the date it was set and the resolution of every channel? Any gap you find is a gap a reader of your report would find too, so fix the record before relying on the numbers.