WEIGHING INSTRUMENTS AND
CALIBRATION OF BALANCES /
MASS COMPARATORS
Mass Standards
INTRODUCTION
Mass of a given object is determined by comparing it
against a standard of known mass using the balance as a
comparator. The precision of measurement of mass
depend on the performance of the balance. Therefore, it
is imperative that before a balance is chosen for
calibration of a given set of weights, it is evaluated for
its performance. The balance must be checked
periodically for its performance.
INTRODUCTION
The results obtained are depended on the environment.
Therefore, an area, which is free from vibration and
excessive air currents, where changes in temperature and
relative humidity are minimal, is selected. The balance
should be placed on a sturdy stone table. The floor on
which the balance is resting should be rigid and
preferably be the ground floor. In addition, high
resolution weighing instruments must also be re-levelled
using a level indicator each time they are moved to a
different location. The reason : their direction of
measurement must be the same as the direction of the
force of gravity (=direction of the gravitational
acceleration) in order to provide accurate results.
ELECTRONIC WEIGHING INSTRUMENTS
Advanced electronic balances and scales determine the
mass of an object by the effect of the earths gravity on
this object; i.e. they measure the weight W according
to the equation :
W =m.g
However, the readout is converted to the mass unit, kg.
CLASSIFICATION OF WEIGHING INSTRUMENTS
The following terms are used to define laboratory balances
according to their readability
Readability Term Usual max. capacity
0.1 g Ultra-microbalance up to 5g
1 g Microbalance 1. 25 g
10 g Semi-microbalance 30200 g
0.1 mg Macrobalance or 50500 g
Analytical balance
1 mg Precision balance 100 g
Accuracy Desig- Verification Scale No. of VSI, Minimum
class nation intervals (e) n = Max/e capacity
Min Max
I Special 0.001 g e 50 000 --- 100 e
accuracy
II High 0.001 g e 0.05 g 100 100 000 20 e
accuracy 0.1 g e 5 000 100 000 50 e
III Medium 0.1 g e 2 g 100 10 000 20 e
accuracy 5ge 500 10 000 20 e
IIII Ordinary 5ge 100 1 000 10 e
accuracy
VERIFICATION SCALE INTERVAL (e) :
A scale division expressed in legal mass unit (mg, g, kg).
This division is used for classification and verification of
balances and scales and referred to the permissible
tolerance.
The verification scale interval for different type of
instruments is as follows
Type of instrument Verification scale interval
Graduated, without auxiliary e=d
indicating device
Graduated, without auxiliary e is chosen by the manufacturer
indicating device according to requirements
Non-graduated e is chosen by the manufacturer
according to requirements
The verification scale interval e is determined by the
expression: d < e 10d where, e = 10k kg
This requirement dose not apply to an instrument of
class I with d < 1 mg, where e = 1 mg.
MULTI-INTERVAL INSTRUMENT : -
Weighing instruments with readout in partial weighing
ranges; for example: 60g readable to 0.1mg; between 60g
and 120g to 0.2mg; and between 120g and 200g to 0.5 mg.
On multiple range instruments, each range is treated
basically as an instrument with one range.
The number of verification scale intervals ni for each
partial range is equal to : ni = Maxi /ei
MAXIMUM PERMISSIBLE ERRORS : -
The maximum permissible errors for increasing and decreasing loads
are given in the table. These Limits must not be exceeded during
verification of the weighing instrument.
Class The given maximum permissible error apply to the
following load ranges:
0.5e 1e 1.5e
I 0 m 50 000e 50 000e < m 200 000e 200 000e < m
II 0 m 5 000e 5 000e < m 20 000e 20 000e <m100 000e
III 0m 500e 500e < m 2 000e 2 000e <m 10 000e
IIII 0m 50e 50e < m 200e 200e < m 1 000e
EXPLANATION OF ELECTROMAGNETIC FORCE
COMPENSATION : -
Force on a conductor though which a current passes in a magnetic
field :-
The force on a conductor through which a current passes a magnetic
field is called "Lorentz Force"
F = (I B) . L---------- (1)
B = magnetic flux density I = electric current
L = length of the conductor in the magnetic field
F I
B L I
S N B
S = South pole of the magnet N = North pole of the magnet
Ring-core permanent magnet system : -
7 F r 4
2 5
S N S
6
3 1
1 Active magnetic material 5 Coil winding
2 Pole plate made of soft iron 6 Lines of magnetic flux
3 Magnetic return pass 7 Resulting force due to current
4 coil frame flow
If the electrical conductor is wound to form a cylindrical
coil with the radius 'r' and the number of coil windings
'n', and this coil is located in the air gap of a ring-core
permanent magnet, the conductor length will be:
L = 2. . r. n ----------- (2)
Because the direction of the flow of electric current and
the direction of the magnet field are always perpendicular
to each other, the vector product in equation-1 can be
replaced by the product of the absolute values in the
following equation:
F = I. B. 2. . r. n ----------- (3)
r = coil radius n = number of coil winding
Ring-core permanent magnet system : -
4 5 7 8
3 9
A/D 10
2
000.00 g P
12 11
1 Weighing pan 2 Lower guide 3 Upper guide 4 Force-transmitting lever
5 Coil 6 Permanent magnet 7 Position sensor (null sensor)
8 Servo amplifier 9 Precision resistor 10 Analog/digital converter
11 Signal-processing device (microprocessor) 12 Digital display
STEPS FOR CONVERTING MASS INTO A READOUT : -
Measured variable Possible Conversion Error
Mass
Dependence of gravitational acceleration on the place of installation (S);
additional force from the parallel-guide construction (ZP); additional forces
e.g. electrostatic forces (ZP);
Force (Weight) on the weighing pan
Change in the lever-arm ratio (S); additional forces applied by the joint flexure
of the lever (ZP);
Force on the coil
Change of the magnetic field; e.g. because of external magnetic sources or cause
by aging or the temperature coefficient (S); change in the position of the coil
in the magnetic field (S, ZP);
Electrical current
Change in the precision resistor; e.g. because of aging or the temperature
coefficient (S); thermoelectric field (ZP);
Voltage
Change in the reference voltage in the analog/digital converter; e.g. because of
aging or the temperature coefficient (S); change in amplification or in the offset
voltage of active components in the analog/digital converter (ZP);
Digital readout
[ZP = Effect on the zero point S = Effect on the sensitivity].
EXPLANATION OF TERMS USED IN WEIGHING
TECHNOLOGY : -
1. Automatic weighing instrument : -
An automatic weighing instrument dose not require the
intervention of an operator during the weighing.
2. Non-automatic weighing instrument : -
A non-automatic weighing instrument requires the intervention
of operator during the weighing.
3. Maximum capacity : -
Upper limit of the weighing range.
4. Minimum capacity : -
Lower limit of a weighing range. The value of a load below
which the weighing result may be subject to an excessive
relative error.
5. Warm-up-time : -
This is the time span from the moment an instrument is
switched on to the moment at which the instrument is
capable of providing the guaranteed accuracy.
6. Stabilization time : -
The time between complete placement of an object on the
pan and indication or display of the final result.
7. Span : -
The difference between the indication of a weight at the
instrument's maximum capacity and the indication at zero.
8. Zero-setting device : -
Device for zeroing the digital display when the weighing
pan is unloaded; e.g., zero-setting key.
9. Zero-tracking devices : -
Device for maintaining the zero readout of a weighing
instrument within certain limits automatically.
10. Weighing performance : -
Determination of errors of measurement of a weighing
instrument.
11. Eccentric loading error : -
This error is also referred to as corner load error.
12. Linearity error : -
This is the deviation from the theoretically straight-lined
(linear) slope of two interdependent values.
VERIFICATION STANDARDS : -
1. Weights : -
The standard weights used for the verification of an instrument
shall not have an error greater than 1/3 of the maximum
permissible error of the instrument for the applied load.
2. Substitution of Standard Weights : -
When testing instruments with Max >1 t, instead of standard
weights any other constant load may be used, provided that
standard weights of at least 1 t or 50% of Max, whichever is
greater, are used. Instead of 50% of Max, the portion of
standard weights may be reduce to:
- 35% of Max if the repeatability error is less than 0.3 e
- 20% of Max if the repeatability error is less than 0.2 e
A possible selection table of weights for calibration of weighing machines
Capacity Resolution
100 g 10 g 1 g 100 mg 10 mg 1 mg 0.1 mg <0.1 mg
Up to 50 g M3 M3 M3 M2 F2 E2 E1
Up to 100 g M3 M3 M3 M3 M1 F1 E1 E1
Up to 500 g M3 M3 M3 M2 F2 E2
Up to 1 kg M3 M3 M3 M1 F1 E1
Up to 5 kg M3 M3 M2 F2 E2
Up to 10 kg M3 M3 M1 F1 E1
Up to 50 kg M3 M2 F2 E2
Up to 100 kg M3 M1 F1
Up to 500 kg M2 F2 E2
TESTING PROCEDURES FOR NON-AUTOMATIC
WEIGHING INSTRUMENTS : -
One electronic balance with digital indication:
Maximum capacity : 1000 g
Tare range : 1000 g
Minimum capacity : 200 mg
Readability : 0.01 g
e : 0.01 g
Repeatability : 0.02 g
Linearity : 0.03 g
Drift in balance indication : 2 ppm/C
Weighing pan : (80 80) mm
1. WEIGHING PERFORMANCE TEST : -
Apply test loads from zero up to Max, and similarly
remove the test loads back to zero. The test load shall be
selected including Max. and Min., and values at or near
those at which the maximum permissible error changes.
Environmental conditions At start At end
Temperature (C) 23.3 23.2
Relative humidity (%) 50.6 53.4
Pressure (mbar) 979.5 979.4
Time 10:00 AM 10:35 AM
E = I + e - L L
Ec = E - E0 with E0 = error calculated at or near zero (+0.01 g)
Load Indication Add. load Error(E) Corr. error mpe
(L) (I) (g) (L)(g) (g) (Ec) (g) (g)
200 mg 0.20 0.22 0.005 0.006 0.000 +0.019 -0.010 +0.009 0.005
500 mg 0.50 0.50 0.004 0.006 +0.001 -0.001 -0.009 -0.011 0.005
2g 2.02 2.01 0.005 0.007 +0.020 +0.008 +0.010 -0.002 0.005
10 g 10.03 10.02 0.006 0.005 +0.029 +0.020 +0.019 +0.010 0.005
50 g 20.02 20.03 0.003 0.004 +0.022 +0.031 +0.012 +0.021 0.005
100 g 100.03 100.03 0.005 0.004 +0.030 +0.031 +0.020 +0.021 0.010
200 g 200.03 200.04 0.007 0.006 +0.028 +0.039 +0.018 +0.029 0.010
500 g 500.04 500.05 0.005 0.005 +0.040 +0.050 +0.030 +0.040 * 0.015
1000 g 1000.04 1000.04 0.005 0.006 +0.040 +0.040 +0.030 +0.030 0.015
Passed Failed
Remarks : - The test was not found to be satisfactory.
At a certain load, L, the indicated value, I, is noted.
Additional weights are successively added until the
indication of the instrument is increased unambiguously
by one scale interval (I + d). The additional load L
added to the load receptor gives the indication P, prior
to round by using the following formula:
P = I + e - L
The error prior to round is :
E = P L = I + e - L L
The corrected error prior to rounding is :
Ec = E E0 mpe
Where, E0 is the error calculated at zero or at a load
close to zero.
An instrument with a scale interval, e, of 0.01 g is loaded
with 1 kg and thereby indicates 1000.01 g. After adding
successive weights of 0.001 g, the indication changes from
1000.04 g to 1000.05 g at an additional load of 0.005 g.
Inserted in the above formula these observations give :
P = (1000.04 + 0.005 0.005) g = 1000.040 g
Thus the true indication prior to rounding :
E = (1000.040 1000.00) g = +0.040 g
If the changeover point at zero as calculated above
was E0 = +0.01 g, the corrected error is:
Ec = +0.055 g (+0.01 g) = +0.030 g [mpe 0.015 g].
2. TIME-DEPENDENT TEST : -
(for instruments of class II , III or IIII only)
2.1. Zero return Test : -
The deviation in the zero indication before and after a
period of loading with a load close to Max for half
an hour, shall be determined. The reading shall be
taken as soon as the indication has stabilized. For
multi-range instruments, continue to read zero
indication during the following 5 minutes after the
indication has stabilized.
P = I + e - L
Time of Load Indication at Add. load P
reading (L0) zero (I0) (L)
11:00 AM 0 0.00 g 0.005 g 0.000 g
After loading for 0.5 h Load = Max. 1000 g
11:30 AM 0 0.05 g 0.006 g 0.049 g
Change of zero indication |P| = -0.049 g [mpe 0.005
g] [|P| 0.5 e]
Passed Failed
Remarks :- The test was not found to be satisfactory.
2.2. Creep Test : -
During this test the temperature should not vary more than 2 C.
Time of Load Indication Add. load P P
reading (L) (I) (g) (L) (g) (g) (g)
11:50 AM 0 min 1 kg 1000.01 0.008 1000.007
11:55 AM 5 min 1000.01 0.007 1000.008 -0.001
12:05 AM 15 min 1000.01 0.006 1000.009 -0.002
12:20 AM 30 min 1000.01 0.004 1000.011 -0.004
(*)
1h
2h
3h
4h
(*) If |P| 0.5e during the first 30 min and if the
variation of |P| between 15 min and 30 min 0.2e,
then the test is terminated. If not, the test will continue
for the next 3.5 hours. Check that during the total
4 hours : |P| mpe.
|P| 0.005 g during the first 30 min and
|P| 0.002 g between 15 min and 30 min
Passed Failed
Remarks :-
The test was terminated because the test was found
to be satisfactory during the first 30 minutes.
3. ECCENTRICITY TEST : -
ECCENTRIC LOADING
The indication for different positions of a load shall
meet the maximum permissible errors.
i. Unless otherwise specified hereafter, a load
corresponds to 1/3 of the sum of the maximum
capacity and the corresponding maximum additive
tare effect shall be applied.
ii. On an instrument with a load receptor having 'n'
points of support, with n>4, the fraction 1/(n-1) of
the sum of the maximum capacity and the maximum
additive tare effect shall be applied to each point
of support.
iii. On an instrument with a load receptor subject to
minimal off-centre loading (e.g. tank, hopper, .....) a
test load corresponding to 1/10 of the sum of the
maximum capacity and the maximum additive tare
effect shall be applied to each point of support.
iv. On an instrument used for weighing rolling loads (e.g.
vehicle scale, rail suspension instrument) a rolling test
load corresponding to the usual rolling load, the
heaviest and the most concentrated on which may
be weighed, but not exceeding 0.8 times the sum of
the maximum capacity and additive tare effect shall be
applied at different points on the load receptor.
ECCENTRICITY TEST : -
Large weights should be used in preference to several
small weights. Smaller weights shall be placed on the top
of larger weights, but unnecessary stacking should be
avoid within the segment to be tested. The load shall be
applied centrally in the segment if a single weight is used,
but applied uniformly over the segment, if several small
weights are used.
5 2 5 2
1 1
4 3 4 3
E = I + e - L L
Ec = E - E0 with E0 = error calculated at or near zero
Load Loca- Indication Add. Load Error Corr. error mpe
(L) tion (I) (g) (L) (g) (E)(g) (Ec) (g) (g)
600 g 1 600.02 0.003 +0.022 +0.012 0.015
2 600.03 0.005 +0.030 +0.020
3 600.01 0.007 +0.008 -0.002
4 600.04 0.004 +0.041 +0.041
5 600.02 0.006 +0.019 +0.009
1 600.02 0.005 +0.020 +0.010
5 600.05 0.004 +0.051 +0.041
4 600.03 0.006 +0.029 +0.019
3 600.02 0.008 +0.017 +0.007
2 600.04 0.003 +0.042 +0.032
1 600.00 0.005 +0.000 -0.010
Passed Failed
Remarks :- The test was not found to be satisfactory.
4. DESCRIMINATION TEST : -
The following tests shall be performed with three different
loads, Min., Max. and Max.
A load plus sufficient additional weights (say 10 times
1/ d) shall be placed on the load receptor. The additional
10
weights shall then be removed successively until the
indication, I, is decreased unambiguously by one actual
scale interval, (I - d). One of the additional weights shall
be placed and a load equal to 1.4d shall then be gently
placed on the load receptor and give a result increased
by one actual scale interval above the initial indication,
(I + d).
Instrument with d = 10 mg (I + d)
add. 1/10 d add. 1.4 d = 14 mg
1000.03 g 1000.04 g 1000.05 g
(I d) = 1000.03 g I = 1000.04 g
Load Ind. Removed Add Extra load Ind. (I2 - I1)
(L) (I1) load(L) 1/10 d = 1.4 d (I2)
(0.2 + 0.01) g 0.20 g 0.008 g 0.001 g 0.014 g 0.21 g 0.01 g
(500 + 0.01) g 500.03 g 0.006 g 0.001 g 0.014 g 500.04 g 0.01 g
(1000 + 0.01) g 1000.04 g 0.009 g 0.001 g 0.014 g 1000.05 g 0.01
Check if (I2 - I1) = d
Passed Failed
Remarks :- The test was found to be satisfactory.
5. REPEATABILITY TEST [Half Load]: -
No. of Indication (I) Add. load (L) P
obs. (g) (g) (g)
1 500.01 0.004 500.011
2 500.02 0.004 500.021 *
3 500.01 0.006 500.009
4 500.00 0.007 499.998
5 499.99 0.005 499.990 *
6 500.01 0.005 500.010
7 500.02 0.004 500.021
8 500.01 0.006 500.009
9 500.00 0.003 500.002
10 500.01 0.005 500.010
[Pmax Pmin] (1 - 10) 0.031 g mpe 0.015 g
Passed Failed
Remarks :- The test was not found to be satisfactory.
5. REPEATABILITY TEST [Full Load]: -
No. of Indication (I) Add. load (L) P
obs. (g) (g) (g)
11 1000.00 0.005 1000.000
12 1000.02 0.005 1000.020
13 1000.03 0.006 1000.029 *
14 1000.01 0.007 1000.008
15 999.99 0.003 999.992
16 1000.00 0.005 1000.000
17 1000.02 0.004 1000.021
18 999.98 0.005 999.980 *
19 1000.00 0.006 999.999
20 1000.00 0.003 1000.002
[Pmax Pmin] (11 - 20) 0.049 g mpe 0.015 g
Passed Failed
Remarks :- The test was not found to be satisfactory.
Example for Uncertainty Calculation
1. Standard uncertainty due to Repeatability (Type A)
Calculated standard deviation from the repeatability test at 1 kg
load, = 0.015 g
Type A uncertainty, uA = / n = 0.015 / 10 = 0.0047 g
2. Standard uncertainty due to Reference Standard (Type B)
Expanded uncertainty of 1 kg load, Us = 0.005 g; k = 2 (taken
from calibration certificate)
then, us = Us / k = 0.005 g/2 = 0.0025 g
3. Standard uncertainty due to drift in the mass values between
calibration (Type B)
If, drift = 0.002 g (from previous data)
then, udrift = 0.002 / 3 g = 0.0012 g
4. Standard uncertainty due to Resolution of the balance
(Type B)
Readability of the balance, d = 0.01 mg
Standard uncertainty due to resolution, ud = (d / 23)
= (0.01 / 23) = 0.0029 g
5. Standard uncertainty due to drift in Sensitivity due to
temperature (Type B)
Given drift in balance indication with temperature as 2 ppm/C,
for 1 kg mass the coefficient relating scale sensitivity drift to
temperature = 2 10-6 1000 g / C.
Assuming temperature variation during calibration is 0.5C
then, usens = 0.5 2 10-6 1000 g / 3 = 0.001 / 3 g = 0.0006 g
6. Standard uncertainty due to eccentric loading (Type B)
d1 where, D = Difference between maximum and
D minimum values
d2 d1 = Estimated distance between the centre
uE = of the weights
23 d2 = Distance from the centre of the load
receptor to one of the corner
40 mm
If, maximum deviation from the centre to any
other corners is -0.022 g d2 40 mm
d1
D
d2 1/
2 (-0.022) g
uE = = = -0.0032 g
23 23
7. Standard uncertainty due to linearity (Type B)
Maximum linearity of the balance = 0.040 g
then, uL = 0.040 / 3 g = 0.0231 g
8. Total Type B standard uncertainty,
uB = (us2 + uDrift2 + ud2 + usens2 + uE2 + uL2)
= [(0.0025)2 + (0.0012)2 + (0.0029)2 + (0.0006)2 +
(-0.0032)2 + (0.0231)2]
= 0.0237 g
9. Combined uncertainty
uC = (uA2 + uB2) = [(0.0047)2 + (0.0237)2] = 0.024 g
10. Expanded uncertainty
uC4(mt)
eff = (n-1)
uW4(m)
(0.024)4
=
(0.0047)4 (0.0025)4 (0.0012)4 (0.0029)4 (0.0006)4 (-0.0032)4 (0.0231)4
+ + + + + +
9
= 6119
For Veff = ; k = 2
Expended uncertainty, U = 2 0.024 g = 0.05 g
Type Source Value Prob. Divi- Sens. Standard DOF
(g) distn. sor Coeff. uncertainty
(g)
A uA Repeatability 0.0047 Normal 1 1 0.0047 9
B us Mass of the 0.005 Normal 2 1 0.0025
reference weight
B uDrift Drift in mass 0.002 Rectangular 3 1 0.0012
B ud Display resolution 0.0071 Rectangular 3 1 0.0029
B usens Sensitivity variation 0.001 Rectangular 3 1 0.0006
B uE Eccentric loading 0.0073 Rectangular 3 1 -0.0032
B uL Linearity 0.025 Rectangular 3 1 0.0231
Total Type B = (us2 + uDrift2 + ud2 + usens2 + uE2 + uL2) 0.0237
Combined standard uncertainty 0.024
Effective degree of freedom, eff 6119
Coverage factor, k 2
Expanded uncertainty, U 0.05