Error Analysis Presentation 2019
Error Analysis Presentation 2019
Laboratory
Dr. Kannan Ramaswamy
BITS Pilani Department of Physics
Hyderabad Campus
Error Analysis-Topics
Random errors
Random fluctuations
due to thermal motions
True value
Random + systematic
2±1A
Read-off error Digital Instruments
12.88 ± 0.005 V
16 17
Q Q , i 1,2...N
N
1 2
d i
N 1
Trial, i 1 2 3 4 5
ti (s) 3.9 3.5 3.7 3.4 3.5
1
Best estimate of T: t
N
t
i
i 3.6s
t t 0.2s
1
Best estimate of error in T: t
2
i
N i
T = 3.6 ± 0.2 s
Standard Deviation Normal Distribution
Frequency
ti
x y
z x y
2 2
z x y
2 2 2
x
z cxy; c
y z x y
z x
z cx a a
z x
z
z sin x x cot x
z
Z A B C Z A B C
2 2 2 2
Z A B C
2 2 2 2
Z ABC or AB / C etc
Z A B C
Z A B C
2 2 2 2
ZAB Cn m l
n m l
Z A B C
A
Z ln A Z
A
Z
Z exp A A
Z
Example of error calculation
A r 2
Where r=(50.5)m
A=78.5398m2
r 0.5
0.1
r 5
A r
2 2
2 (2 0.1) 2
0.04
A r
A
0.2 hence A 0.2 A A 0.2 (78.5398m 2 ) 16m 2
A
=3.1735s
2 l 2 g
2 2
1 0.1 1 0.2 4
5.04 10
2 2.5 2 9.8
/=0.022 hence =0.022x3.1735=0.070
=(3.170.07)s
Least square fit method
S yi mxi c
2
S S
0 and 0
m c
m
xi x yi
c y mx
xi x
2
y mx
Minimizing S yi mxi
2
S
0
m
m
xy i i
x
2
i
200
150
100
Y
50
0
-50 0 50 100 150 200
-50
X
BITS Pilani, Hyderabad Campus
Least square fit: Non-linear
functions
How will you do least square fit for non-linear functions like
y ax ? b
(300, 20000)
100
0.90
60
0.80
0.70
0.60
0.50
0 0.40
1
0.30
2 4 6 10 20 40 100 200 1000
0.20
0.07
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 1
Semi-log is useful for relations of the type
C C o exp( E / kT )
Taking log on both sides of this equation and
changing the base to 10
E
log C log C o
2.303kT
Vernier scale:
0 5 10
1 5 10 15 20