Chapter 4
Accelerometer Testing
4.1 Introduction
The accelerometer designed and realized so far is intended for an
aerospace application. Detailed testing and analysis needs to be
conducted to qualify the product for the end use and establish the
suitability of the product for the intended use. These tests can take many
different forms including chip level probing for electrical characterization,
sensitivity estimation, bias stability, performance evaluation under
thermal and dynamic environments. Details of performance testing
procedure of commercially available accelerometers are presented by
number of researchers [47-49].
MEMS pendulous accelerometers have already demonstrated good
performance in automobile and other commercial applications. The
challenge of using of this technology in aerospace inertial navigation is
about, significantly improving the bias stability, cross-axis sensitivity,
temperature
sensitivity
and
detailed
measured
performance
87
broadly adopted from IEEE STD 337-1972 [50]. This describes the test
procedures for linear, single axis pendulous analog, torque balance
accelerometer.
The test procedure consists of observing the output of the test
device to input acceleration using Earths gravitation field or an external
excitation source.
The objective of the test is to characterize the accelerometer
Bias stability
Linearity
Hysteresis
Temperature sensitivity
Bandwidth.
Shock response
It was presumed that physical sources of the errors described in the
error
model
components
selected
for
micromachined
88
E1 Bias drift
E2 Non linearity
E3 Cross- axis sensitivity
E4 Temperature sensitivity of bias drift
E5 Misalignment of input axis.
T change in device temperature during testing.
4.2 Chip level probing
The primary aim of the chip level probing is to validate the wafer
fabrication process and to certify the device functionality for further
packaging and integration operations.
89
Average nominal
capacitance [Co1]pF
[Co2]pF
3.0102
3.1124
Observations and comments:1. The average nominal capacitance between electrodes and proofmass measured is around 3 pF which is very close to the simulated
value of 2.5 pF. The discrepancy in nominal capacitance may be
due to stray capacitance caused by electrical routing & pads and
also due to fine difference in air gap on both sides.
2. However the variation in nominal capacitances Co1 and Co2 is
small hence can be balanced using internal trim capacitors of the
ASIC during electronic integration.
90
91
92
basis for further detailed performance testing. The test method block
diagram is shown in fig 4.2
accelerometer
sensor
is
mounted
on
an
automatic
93
sensitivity
of
the
accelerometer
is
62
mV/g
at
room
94
steps.
The
output
measured
at
different
accelerations
95
Hysterisis
4500.00
4000.00
Output V (mV)
3500.00
3000.00
2500.00
2000.00
Up
1500.00
Down
1000.00
500.00
Acceleration g
0.00
-30
-20
-10
10
20
30
Acceleration (g)
UP (mV)
Down (mV)
-30
320.00
322.00
-18
1063.00
1064.00
-6
1800.00
1802.00
2181.00
2180.00
2548.00
2550.00
18
3294.00
3297.00
30
4041.00
4041.00
96
Observation and remarks:From the readings presented it can be seen that the maximum deviation
in output is at 18 g. The hysteresis is given as
97
Output(V)
2.1850
2.1800
2.1750
2.1700
0
20
40
60
80
100
120
time(Min)
98
99
100
101
100
100
102
103
Cross-axis sensivitty
sensor output (Volts)
2.1900
2.1850
2.1800
2.1750
2.1700
2.1650
2.1600
0
10
15
20
25
Cross axis
acceleration
0
Output (V)
2.1812
2.1823
10
2.1834
15
2.1845
20
2.1855
25
2.1866
30
2..1874
30
35
104
Since the sensor is symmetrical along X-axis and Y-Axis the crossaxis test is done along one direction only.
105
2.1880
2.1860
2.1840
2.1820
2.1800
2.1780
2.1760
-20
20
40
60
Temperature deg C
80
106
Temp (C)
Nominal output
(V)
-20
2.1773
2.1786
20
2.1812
40
2.1837
60
2.1860
80
2.1871
Table 4.4
The zero g offset voltage is measured at -20, 0, 20, 40, 60, 80C.
The zero g offset voltage at -20C is subtracted from the value obtained
at 80C. The resulting value obtained is divided by the accelerometer's
FSO to express the change in output in terms of % of FSO or alternately
it can be expressed as ppm also.
From Table 4.4, it can be seen that over the operating temperature
range - 20C to + 80C
107
108
sensitivity mv/g'
62.04
62.02
62
61.98
61.96
61.94
61.92
61.9
-20
20
40
60
80
temperature C
Temp (C)
Sensitivity (mV/ g)
-20
61.906
61.923
20
61.948
40
61.965
60
61.989
80
62.006
109
From Table 4.5 it can be seen that the scale factor variation over
the temperature range is a maximum of 0.1mV. Hence the scale
factor stability is 1612 ppm. This stability figure is adequate for
control class aerospace applications.
110
111
112
Power
Supply
Accelerometer
Data
Logger
Shock
Tester
11msec. Fig. 4.18 gives the input shock spectrum and fig. 4.19 is the
response shock spectrum of the accelerometer, the graph plots the
response of three different accelerometers with different sensitivities. Also
the shock response spectrum provides a measure of response time of the
sensor.
113
- Sensor 1
- Sensor 2
- Sensor 3
114
From the test results, it can be seen that, all the three
accelerometers have responded in a similar way to the applied
shock and the response is along the lines of input shock.
The hysteresis and bias stability values measured are less than
0.15% of FSO and meet the sensor specifications.