MI - Errors

ERROR ANALYSIS
Absolute Static Error: Difference between measured value and true value.
:Error/Absolute Static Error ; :Measured Value ; :True Value
Absolute Value of an error does not precisely indicate the accuracy of

measurement. E.g. Error is 2A
Relative Static Error

= =
=
+
Static Correction: Difference between true value and the measured value
=
Errors and uncertainties are inherent in the process of making
any measurement & in the instrument with which
measurements are made.
Study of errors is important To find ways to reduce them

LIMITING ERRORS (Guarantee Errors)
Usually, accuracy is within a certain % of full scale reading Deviations from

this nominal value has to be specified
Limits of deviations from specified value LIMITING ERRORS
Relative Error: Ratio of error to true quantity of a component
=
=
Error = =
Relative Limiting Error ( )

=
Percentage Limiting Error (% )

x 100
RESULTANT ERROR
A measuring instrument may be formed by combining several

components, each having its limiting error
The limiting error of combination may be computed based on how the

various components are connected by considering relative increment
of a function
(i) Result is a sum or difference of various components
(i) Result is a product or quotient of various components
(i) Result is a composite factor of output of various components

Result is a sum or difference of various components
If, , , , . = Outputs of each device
(sum or difference) Q =
= . + . + . +
Resultant Limiting Error = Sum of products formed by

multiplying individual limiting errors by the ratio of each term
Result is a product or quotient of various components
=
=
=
= + + +
Relative Limiting Error is a sum of relative errors of all

components
Result is a composite factor of output of various components

= + + .+
= + + +
Guarantee values are obtained by taking a direct sum of
possible errors adopting algebraic signs for worst possible case
Pessimistic process
KNOWN ERRORS
When error of a certain quantity is already known effect of

this can similarly be considered and computed.
Only difference is signs of relative errors be presented in

the calculations
1) A multimeter having a sensitivity of 2000/V is used for the
measurement of voltage across a circuit having an output resistance of
10k. The open circuit voltage of the circuit is 6V. Find the reading of the
multimeter when it is set to its 10V scale. Find the percentage error.
Ans: Measured Output Voltage EL= 4V
Percentage error = -33%

2) An oscilloscope having an input resistance of 1 M shunted by 50pF
capacitance is connected across a circuit having an effective output resistance
of 10k. If the open circuit voltage has 1V peak for a 100kHz sine wave, what
will be the voltage indicated on the oscilloscope when the frequency is
(i) 100kHz (ii) 1MHz?

3) It is desired to measure the value of current in the 500 resistor as shown
in figure by connecting a 100 ammeter. Find (i) The actual value of current
(ii) Measured value of current and (iii) The percentage error in measurement
and the accuracy.
Thevenins equivalent circuit

4) During a test the current was measured as 26.5A flowing in a resistor of
0.12. Later it was discovered that ammeter reading was low by 1.1% and
marked resistance was high by 0.25%. Determine true power as % of power
originally calculated.
STATISTICAL ANALYSIS OF DATA
Experimental Data: i) Multisample Test
ii) Single- Sample Test
Multisample test: Repeated measurements using different test

conditions.
Single-sample test: single/multiple measurements under

identical conditions.
Statistical analysis is valid only for Multisample Test Data.

Different methods of statistical analysis:
Histogram
Arithmetic Mean
Measure of dispersion from the mean
Range
Deviation
Average Deviation
Standard deviation
Normal or Gaussian Curve of Errors
Precision Index
HISTOGRAM:
Graphical representation of the distribution of numerical data.
It is frequency distribution curve.

ARITHMETIC MEAN:
Most probable value of measured variable.
..
= =
where -> arithmetic mean
.. -> readings or variates or samples
n > number of readings

MEASURE OF DISPERSION FROM THE MEAN:
Extent to which the values are dispersed about the central value.
Its an indication of the degree of precision and regularity of data.

RANGE:
Measure of dispersion.
Difference between greatest and least values of data.
DEVIATION:
Departure of observed reading from the arithmetic mean.
Departure d1= x1 -
d2= x2 - and so on..
( )
Then, =
AVERAGE DEVIATION:
Indiaction of precision of the instruments.
Highly precise instruments low average deviation between

readings.
STANDARD DEVIATION (S.D):
Square root of the sum of the individual deviations squared,

divided by the number of readings.
..
S.D = = =
NORMAL OR GAUSSIAN CURVE OF ERRORS:
Law of Probability The normal occurrence of deviations from

average value of an infinite number of measurements or
observations can be expressed by y = exp (-h2d2).
Its an indication of the degree of precision and regularity of data.
y= exp (-d2/2 2)

MI - Errors

Uploaded by

Copyright:

Available Formats

MI - Errors

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MI - Errors

Uploaded by

Copyright:

Available Formats

ERROR ANALYSIS

:Error/Absolute Static Error ; :Measured Value ; :True Value

Absolute Value of an error does not precisely indicate the accuracy of

Relative Static Error

Study of errors is important To find ways to reduce them

Usually, accuracy is within a certain % of full scale reading Deviations from

Limits of deviations from specified value LIMITING ERRORS

Relative Error: Ratio of error to true quantity of a component

Relative Limiting Error ( )

Percentage Limiting Error (% )

A measuring instrument may be formed by combining several

The limiting error of combination may be computed based on how the

(i) Result is a sum or difference of various components

(i) Result is a product or quotient of various components

(i) Result is a composite factor of output of various components

If, , , , . = Outputs of each device

Resultant Limiting Error = Sum of products formed by

Relative Limiting Error is a sum of relative errors of all

When error of a certain quantity is already known effect of

Only difference is signs of relative errors be presented in

measurement of voltage across a circuit having an output resistance of

Ans: Measured Output Voltage EL= 4V

Percentage error = -33%

capacitance is connected across a circuit having an effective output resistance

will be the voltage indicated on the oscilloscope when the frequency is

(i) 100kHz (ii) 1MHz?

and the accuracy.

Thevenins equivalent circuit

marked resistance was high by 0.25%. Determine true power as % of power

Experimental Data: i) Multisample Test

ii) Single- Sample Test

Multisample test: Repeated measurements using different test

Single-sample test: single/multiple measurements under

Statistical analysis is valid only for Multisample Test Data.

Measure of dispersion from the mean

Normal or Gaussian Curve of Errors

Graphical representation of the distribution of numerical data.

It is frequency distribution curve.

Most probable value of measured variable.

where -> arithmetic mean

.. -> readings or variates or samples

n > number of readings

Its an indication of the degree of precision and regularity of data.

Difference between greatest and least values of data.

Departure of observed reading from the arithmetic mean.

d2= x2 - and so on..

Indiaction of precision of the instruments.

Highly precise instruments low average deviation between

STANDARD DEVIATION (S.D):

Square root of the sum of the individual deviations squared,

Law of Probability The normal occurrence of deviations from

observations can be expressed by y = exp (-h2d2).

Its an indication of the degree of precision and regularity of data.

You might also like