2–102
Motorola Sensor Device Data
For More Information On This Product,
Go to: www.freescale.com
along the wiring harness. Once through the chamber wall, the
data is translated to RS232 and fed to a dummy terminal. On
the terminal screen, the A/D codes for the accelerometer can
be monitored for unexpected performance.
Ideally, when the accelerometer is at rest (no acceleration
applied), the output should be at 0g regardless of what EMC
testing the system may be subjected to. Depending on the
crash algorithm of the airbag module software, there is some
allowable offset shift that can be tolerated. Higher shift in
output could create errors in the crash analysis software,
perhaps causing the airbags to unnecessarily deploy when
there is not a crash or not deploy when there is a crash.
The Motorola accelerometer must be able to meet the
airbag system requirements throughout BCI exposure. It has
a sensitivity of 40 mV/g and an offset (0goutput) of 2.50 V.
During the BCI test, the accelerometer output should be
2.50 V at 0gwith as little drift as possible. A typical airbag
system may have software that can tolerate from as little as
0.5 gup to 2.0 g of deviation from the offset. The system would
then expect the accelerometer output to be within 40 mV of the
offset during the entire BCI test. Therefore, at any given
frequency of the BCI test, if the output deviates outside this
expected window of drift, it fails the test.
MMA2202D ACCELEROMETER BCI TEST
RESULTS
If a system has not been well designed for electromagnetic
compatibility, the accelerometer, as well as other devices, can
have performance problems. What has been found for the
accelerometer is that in some system applications, it suffers
from an offset shift when certain frequencies of BCI are
applied. For example, in one airbag system being tested at a
certain frequency, with the desired BCI current applied, the
offset is found to shift down by 60 mV. This would equate to an
error of 1.5 g See Figure 2. At other frequencies, this shift is
even higher. This DC shift plot was taken with an oscilloscope
using a 20 MHz filter to remove the high frequency component
of the signal. Probes are placed at the accelerometer in the
system application. The plot shows the accelerometer output
before and after BCI was applied (before and after the RF
generator creating the high frequency signal was turned on).
Figure 2. Accelerometer Tested Under High Frequency BCI
ACCELEROMETER VOUT
w/o BCI
ACCELEROMETER VOUT
w/BCI
VCC
This phenomenon has been determined to be system level
related. PCB layout and grounding for the accelerometer will
affect its performance. This was found by testing the
accelerometer outside of the airbag module. The device was
put on a test board by itself with only the supply decoupling
capacitor of 0.1
μ
F connected to it. To simulate the effect of
BCI on Vcc, a frequency generator was used to inject a known
high frequency sinusoid that caused BCI failure on to the 5.0 V
supply voltage. The device was first tested in small test board
with ground provided by one wire back to the supply. This
grounding reproduced the failure due to BCI seen at the
module level. The test board was then mounted down to a
ground plane provided by a copper plate and the
accelerometer ground was soldered to the plate (providing a
low impedance path to ground). With this setup, the offset shift
did not occur.
If a system does not incorporate a good PCB layout
providing a low impedance to ground, the accelerometer
output may shift at certain high frequencies. This output offset
shift was caused by a shift in the 0–5 V supply window.
Because the accelerometer has a ratiometric output, its offset
is dependent on the supply voltage. Any change in the supply
F
Freescale Semiconductor, Inc.
n
.