MEASUREMENT AND ERROR

ctd.
CHARACTERISTICS OF MEASURING INSTRUMENTS
ANALYSIS OF MEASURED DATA
UNCERTAINTY ANALYSIS
 Uncertainties in Reading Digital Displays

• Gate and clock signals are combined in an

AND gate, case (a) results 3 pulses while
case (b) supplies 4 pulses.
• A digital read-out has an uncertainty of ±1
digit.
2
 Uncertainties in Reading Analog
Displays

• The uncertainty (U) in analog

displays
! !
• 𝑈 = ∗ 𝐹𝑆𝐷 ∗
" #$%&%'%()'
• What is the uncertainty of
reading voltages using the
meter shown?

3
 Example: Uncertainty in analog voltmeter
• In a 100 divisions Voltmeter, voltage reads 6V. If meter has
two ranges (0-10V & 0-100V):
– What is the uncertainty for both ranges:
! !
• 𝑈 = " ∗ 𝐹𝑆𝐷 ∗ #$%&%'%()'
10 V range
– Uncertainty = ± ½ (10/100) = ± 0.05 V
– V = 6 ± 0.05 volt.
100 V range
– Uncertainty = ± ½(100/100) = ± 0.5 V
– V = 6 ± 0.5 volt.

Relative uncertainty: Percentage uncertainty:

– 10 V range – 10 V range
– RU = 0.05V/6V = 1/120 = 0.0083; – PU = (0.05V/6V)x100 = 0.83%
– 100 V range – 100 V range
– RU = 0.5V/6V = 1/12 = 0.083 – PU = (0.5V/6V)x100 = 8.3%
4
 How do we assess the total error?
• To treat a reference standard as a sample.
• The reference standard would be carried
through the entire process to see how
close the results are to the reference
value.
• Make an independent set of measurements,
• Using different instruments
• Different measurement techniques
• Not subject to the same systematic errors.

5
Statistical analysis of the measurement
• Allows an analytical determination of the uncertainty of the
final test result.
• The outcome of a certain measurement method may be
predicted on the basis of sample data without having
detailed information on all the disturbing factors.
• To make statistical methods and interpretations
meaningful, a large number of measurements are usually
required.
• Also, systematic errors should be small compared with
residual or random errors, because statistical treatment of
data cannot remove a fixed bias contained in all the
measurements.
 Arithmetic Mean
• Most probable value of a measured
variable
• best approximation when # of readings of
the same quantity is very large.

Example: A set of independent current measurements was taken by six

observers and recorded as 12.8 mA, 12.2 mA, 12.5 mA, 13.1 mA, 12.9
mA, and 12.4 mA. Calculate the arithmetic mean
 Deviation from the Mean
12.8 mA, 12.2 mA, 12.5 mA, 13.1 mA, 12.9 mA, and 12.4 mA

8
Average Deviation

9
 Standard Deviation of the…
• Population • Sample – part of population

Statistics: the variation in the population should be represented

in the sample

10
Standard Deviation

11
 Histogram
• 100 volt. readings taken at small time
intervals and recorded to nearest 0.1 V
Table 1
20
Voltage # of
Readings reading
99,7 1 Number of Observed 15
Readings
99,8 4
10
99,9 12
100,0 19
5
100,1 10
100,2 3 0
100,3 1 99.7 99.8 99.9 100.0 100.1 100.2 100.3
Reading Values
 Normal Distribution of Errors
• For unbiased experiments
– all observations include small disturbing effects, called random
errors
• Error distribution curve indicates:
– Small errors are more likely to happen than large errors.
– Large errors are less likely to happen.
– There is an equal probability of plus and minus errors

20
Observed Readings

15
Number of

10

0
99.7 99.8 99.9 100.0 100.1 100.2 100.3
13
Reading Values
 Normal (Gaussian) law of distribution of
errors

Area Under the

Probability
Curve

Deviation ±s Fraction of
total area

0.6745 0.500
1.0 0.682
2.0 0.954
3.0 0.996
14
Range of possible values: confidence intervals
• Two resistance values: 2.65 ± 0.04 kW and 2.7 kW.
• Do the two values agree?
• Rule of thumb: if the measurements are within 2 SD, they agree
with each other. Hence, ± 2 SD around the mean value is called the
range of the variable.
range CI
0.682

0.954

0.997

0.999
Range of the
variable
Random distributions are typically
15
Gaussian, centered about the mean
Why take many measurements ?
• The error on the mean improves dramatically!
• It goes as s / sqrt(N) , where N is the number of
measurements

𝑁! > 𝑁" > 𝑁*

16
Calculating Probable Error
𝑃𝑟𝑜𝑏𝑎𝑏𝑙𝑒 𝐸𝑟𝑟𝑜𝑟 = ±0.6745𝜎

= 0.2Ω

𝑃𝑟𝑜𝑏𝑎𝑏𝑙𝑒𝐸𝑟𝑟𝑜𝑟 = ±0.1349

The SD is more convenient and is

used in most applications
In case a distribution doesn’t exist
Probable Error can be used 17
 Determining Random Errors
• Instrument limit of error and least count
– least count is the smallest division that is marked on
the instrument
– The instrument limit of error is the precision to
which a measuring device can be read, and is always
equal to or smaller than the least count.
• Estimating uncertainty
– A volt meter may give you 3 significant digits, but you
observe that the last two digits oscillate during the
measurement. What is the error ?
• Average deviation or standard deviation
– Based on repeated measurements of the same
quantity
18
19
20
Limiting Error & Expected Error
■ Two methods are commonly used for
determining the uncertainty.
■ The first one is called the method of
equal effects and it yields the limiting
(guarantied) error (maximum uncertainty
possible).

21
22
Expected Value of Uncertainty
■ The second method is called the square
root of sum of squares.
■ It is based on the observations stated
before for the random errors.
■ It yields the expected value of the
uncertainty and computed as:

23
24
 Summary of how to propagate the errors ( z
= f (x,y))
Function Relation between Dz, Dx Comment
and Dy
1 z=x+y Addition and subtraction: add
absolute errors
2 z=x-y
3 z = xy Multiplication and division: add
4 z = x/y relative errors. Multiplication by
an exact number (a*x): multiply
absolute error by the number
5 z = xn
6 z = ln x
7 z = ex

Here it is all explained from MathWorld http://mathworld.wolfram.com/ErrorPropagation.html

25
 Active Learning Exercises (1of 2)
1. The potential of an electrical power source is measured 12.47 volts by a
recently calibrated digital voltmeter. Two other voltmeters are used in the
lab to measure the same voltage by six different observers in a short
interval of time and following results (in volts) are recorded:
Observer 1 Observer 2 Observer 3 Observer 4 Observer 5 Observer 6

Meter-1 11.456 V 11.324 V 11.562 V 11.243 V 11.472 V 11.376 V

Meter-2 12.45 V 12.34 V 12.67 V 12.76 V 12.21 V 12.54 V

a) Determine the resolution of each meter in volt. Which one has a better
resolution?
b) Determine the accuracy and precision of each meter. How much is the bias in
each meter? Which one is more precise? Which one is more accurate?
26
Active Learning Exercises (2 of 2)
2. Three resistors are in series. The values are (in kW) 47.23, 2.205, and
180.2, with an uncertainty of one digit in the last place. Find the total
resistor and express the result using significant figures only.
3. A digital thermometer is used to measure the temperature of boiling
water. It is known that water boils at 96.2°C under the given
environmental conditions. The measurement is repeated 5 times and
following readings are obtained: 95.9, 96.2, 96.5, 95.8, 96.1.
Determine the percentile accuracy, the precision and the bias of the
thermometer.
4. The DC current in a resistance R = 10 kW ± 0.5% is measured to be I =
10 mA ± 1%. Find the power dissipated in this resistance with its
uncertainty and limiting error.

March 11, 2019 BME 424E Electronic Measurement 27

and Display Technologies
Thanks!