Lab #1: Error: ME 422 Fall 2018
Lab #1: Error: ME 422 Fall 2018
THEORY
All measurements are an approximation of the true value of a certain variable. There is no such
thing as a perfect measurement. Accuracy shows how closely a true value is approximated by a
particular measurement.
0 1 2
The above line has a length that falls between 2” and 2 ½”. The approximate length can be said
to be 2 ¼”. If the desired accuracy is to the nearest ½” or ¼”, then this measurement is accurate. If the
desired accuracy is to the nearest 1/16”, then you have too much inaccuracy with this ruler. A ruler must
be found with more divisions on it.
When looking at a group of values, it is necessary to see how much the values deviate between
each other. Precision shows how reproducible the values are. Precision does not evaluate incorrectness,
it just looks at how close the values are together. Precision will depend on the amount of readings taken
and on the deviation between readings.
Precision may also be applied to tolerances between different quantities of the same type. Each
time you use a precision resistor, it will be within certain a specified tolerance.
PREFIXES
Prefix Scientific Notation Symbol
giga 1 × 10 G
mega 1 × 10 M
kilo 1 × 10 k
milli 1 × 10 m
micro 1 × 10 μ
nano 1 × 10 n
pico 1 × 10 p
ERROR
A perfect measurement cannot be made. There is always some difference between the
measured value and the true value. The difference is error. In general, there are three types of error,
gross error (people error), systematic error (equipment error), and random error (non-predictable
error). Statistically, there are ways to anticipate and correct for the error in the measuring process.
The first quantity to be discussed is the arithmetic mean. Consider that n measurements are
made of a variable m.
⋯
=
⋯
= = (1 standard deviation)
From statistics, when many measurements are obtained of one variable, the values will be
normally distributed about the mean. Approximately 68% of the values will be between the mean
and ± one standard deviation. Furthermore, approximately 95% of the values will be between the mean
and ± 2 standard deviations.
2
STANDARDS AND CALIBRATION CURVES
All measuring devices must refer to some standard. The unit of length now used in practically all
scientific fields is the meter. A standard meter is based on the number of radiated wavelengths of
Krypton, and this standard with its measuring devices are kept in France. This non-derived (basic)
standard is known as an international standard. All countries keep their own standards, which are
occasionally compared to the international standards. These are the Primary standards. A calibration
shop (within a company) will keep its own standards, which are occasionally compared to the Primary
standards. These are the shop standards, or the working standards.
As mentioned previously, systematic error is the error which is due to the equipment that is
used in the measuring process. One method of correcting for the systematic error of an instrument is by
calibrating against a standard. For this, a quantity is measured both against the standard and by the
instrument to be calibrated. No calibration is perfect so Calibration Curves are used to show how the
instrument acts over a range. It will allow the experimenter to determine the trustworthy (accurate)
range of the instrument.
= ∗
For two resistors in parallel, the current divider formula can be used:
= ∗
QUESTIONS
1) Can you have accuracy without precision? Explain.
Yes, you can have accuracy without precision. Accuracy is simply the closeness to the actual
value where precision is uniformity in the actual value. An accurate measurement may be taken,
but if it is not precise as well, it is fairly worthless due to the changing values.
2) Can you have precision without accuracy? Explain.
Yes, you can be precise, but not accurate at all. A gun out of sight is an example of this. The gun
might have excellent precision and a cluster of bullets, but little accuracy to the actual intended
target. A machine may give the same measurement each time, but if it is not calibrated correctly,
it will not be accurate.
3) If a voltmeter reads 10.0 volts, and the specified accuracy is ±5%, into what voltage range might
the true value fall?
9.5-10.5v
3
Match the following:
4) Find the mean, average deviation, and standard deviation for the measurements given below:
n m
1 9.8
3 10.0
4 9.7
6 10.0
7 9.9
9 9.9
10 10.1
4
5) A voltmeter was calibrated and the following table of data was obtained.
Use the information to:
(a) Complete the table,
(b) Plot a calibration curve showing the deviation vs. the voltage measured by the instrument,
(c) Determine the accuracy of the voltmeter when the meter reads 11.00V
Measurements Measurements
Deviation from true
made by the made by the
voltage*
standard instrument
(V)
(V) (V)
0.00 0.00 0.0
1.00 1.05 0.05
2.00 2.05 0.05
3.00 3.08 0.08
4.00 4.09 0.09
5.00 5.10 0.10
6.00 6.11 0.11
7.00 7.11 0.11
8.00 8.13 0.13
9.00 9.15 0.15
10.00 10.15 0.15
11.00 11.16 0.16
12.00 12.18 0.18
13.00 13.18 0.18
14.00 14.20 0.20
15.00 15.20 0.20