Some Recent Developments in Acceleration Sensors: George Juraj STEIN

MEASUREMENT SCIENCE REVIEW, Volume 1, Number 1, 2001
SOME RECENT DEVELOPMENTS IN ACCELERATION SENSORS

George Juraj STEIN
Institute of Materials and Machine Mechanics
SLOVAK ACADEMY OF SCIENCES
Raianska 75, SK-831 02 Bratislava 3, SLOVAK REPUBLIC
on temporal leave at Department of Measurement,
Faculty of Electrical Engineering and Information Technology
SLOVAK UNIVERSITY OF TECHNOLOGY, Bratislava
Abstract
The contribution deals with recent development in acceleration sensors,
specifically on MEMS sensors and their use. The various noise sources presented
are described. Also a short description of so-called Transducer Electronic data
Sheet (TEDS) concepts according to emerging standard IEEE P1451 is included.
1. Introduction
Acceleration and associated vibratory variables belong to the vast amount of physical
quantities subject to measurement and analysis by various means. There is a continuous quest
for improvement of vibration sensors properties used in this diversified field of measurement
science. The basic physical principle is quite simple: the so-called proof mass or seismic mass
m is supported by an elastic element of stiffness c. This element may be in the form of a prestressed spring or a cantilever beam. The relative displacement x(t) of this linear oscillatory
system is damped by a linear damper whose damping b is proportional to the relative velocity
x&(t ) of the proof mass m in respect to the sensor body. Using Newtonian formalism, the
describing linear differential equation (1) can be derived in following form [1]:
x&&(t ) + 2 0 x& (t ) + 20 x = &&
y(t ) ,
(1)
where the so-called sensors natural circular frequency 0 = (c/m) and dimension-less
damping ratio = b/bkrit , with critical damping bkrit = (4cm) are introduced. In this way the
acceleration &&
y(t ) of the sensor base is converted to relative movement of the proof mass in
respect to the body. Then certain mechanical variables are subject to further transformation to
electrical quantities and their further processing in acceleration sensors [1-3]:
a. Force Kx(t) in the spring type element via the intrisinic piezo-electric effect,
b. Relative displacement x(t) by a relative displacement sensor,
c. Counter-balancing the elastic force Kx(t) by another force of non-inertial nature, hence
establishing a pre-set equilibrium position sensed by a sensitive null-detector.
The first transducing principle is widely used in standard piezo-electric accelerometers,
having well known properties - good linearity and wide dynamic range, ruggedness, etc.
However, being a charge source has also some disadvantages - very high inner impedance,
low output signal and no steady state response. Some miniaturised accelerometers are
equipped with build-in integrated pre-amplifier/impedance converter, and in this way above
disadvantages are partially removed. Such an accelerometer is connected by a standard coaxial
cable to a simple powering front-end unit where the ac vibratory signal is branched of for
further processing. This type of accelerometer is generally termed as the Integrated Electronic
Piezo-Electric sensor (IEPE sensor) and is marketed under various brand names.
183
Measurement of Physical Quantities G.J.Stein
2. MEMS acceleration sensors

The rapid development of semiconductor manufacturing technologies enabled development
and mass manufacturing of various sensors and actuators using state of the art technologies.
Devices, whose operation principle is based on use of miniature mechanical elements are
denoted as Micro Electro-Mechanical Systems (MEMS). Also in the field of vibration
measurement such systems have been developed and are marketed [4-7].
The above physical principles and
methods of transducing of mechanical
variables into electrical ones are used
[2, 3]. The principal sketch of MEMS
type sensor is in Fig. 1, where also the
first two mechano-electrical transducing
principles
are
indicated.
Differential capacitor principle is
often used in commercial applications
[4, 5]; the servo-accelerometer principle is used widely too [5, 6] (Fig. 2):
Fig. 1. Principal sketch of MEMS.
The servo-accelerometer principle is following [6]: the sensitive element (dashed) is a comblike structure of 46 differential capacitors arranged in parallel on a beam forming the seismic
mass supported by springs etched from silicon substrate. The differential capacitor forms a
capacitive half-bridge driven from a high frequency square wave generator by opposite phase
pulses. When acceleration is applied perpendicularly to the seismic mass the differential
capacitor is mismatched and a non-zero signal appears on the central plate. This signal is preamplified, demodulated in the synchronous detector, amplified and outputted as voltage VPR,
corresponding to the applied acceleration. The demodulated signal VPR is feedback via the
internal loop 3 M resistor to the differential capacitor's central plate, so providing the
electrostatic restoring force to move the beam to the original centred position. Further signal
processing is provided on-board by a build-in signal conditioning/band-limiting amplifier.
Further improvements [7] consist of addition of a temperature sensor to compensate for
temperature influence and of using an on-board voltage-to-duty cycle converter to improve
connectivity to micro-controller (no ADC is required; instead counter inputs are used).
Fig. 2. Abridged layout of servo type MEMS accelerometer (after [6]).
184
MEASUREMENT SCIENCE REVIEW, Volume 1, Number 1, 2001
3. Noise and drift issues in MEMS sensors

Due to extreme miniaturisation the noise and temperature drift issues in MEMS sensors are of
paramount importance. The MEMS sensor's proof mass m is of the order of 1 g and is
influenced by the Brownian motion of the molecules of the surrounding medium. Following
sources of noise are present, which fundamentally limit sensor's performance [5, 8, 9]:
- Thermal noise due to agitation of the proof mass by the surrounding medium,
- Thermal noise in resistors of the semiconductor structure (Johnson noise),
- So called 1/f noise associated to any source of energy in the sensor's structure.
The proof mass thermal agitation can be expressed using thermodynamic approach as the
energy of simple damped harmonic oscillator of Eq. (1), excited by random force due to
thermal fluctuations of the surrounding medium at absolute temperature T. The power spectral
density frequency distribution GN(f) of this thermal noise is flat and constant (GN(f) = GN) [9].
This input mechanical excitation is shaped by the frequency response of the oscillatory system
[5, 8] and is manifest at sensor's natural frequency. According to the equipartition principle
the average potential energy and the average kinetic energy of the oscillator in thermal
equilibrium with it's surrounding are equal to thermal fluctuations (kB = 1.38 10-13 J/K is the
Boltzmann's constant):
1
2
m v 2 = 12 c x 2 = 12 k BT .
(2)
The average squared displacement x 2 of the oscillating mass due to the random thermal
force is equal to GN/(4bc) [8], while, according to generalised Nyquist's theorem, GN = 4bkBT.
Then the signal to noise ratio (SNR) due to thermal agitation is given as the ratio of the square
of the rms value of the force due to excitation by acceleration (ma) to the square of the rms
value of the force due to thermal agitation at temperature T [8, 9]:
SNR = (ma ) : (4 k BbT ) .
2
(3)
This source of noise is manifest at sensor's natural frequency, especially if in vacuum and
adds to the Johnson type noise generated by the sensor's electronic circuits [5]. For practical
purposes a SNR value of 1 is assumed. Then the combined thermal and Johnson noise floor is
given as equivalent rms acceleration value for the designed sensor frequency band [5-7].
The 1/f noise source seems to play an
important role in acceleration measurement
at low frequency [8, 9]. This noise source is
termed the Hooge's noise [9]. The power
spectral density frequency distribution GH(f)
depends on the biasing voltage across a
resistor Vb and the number of carriers n in
the resistor, is a constant (Fig. 3 after [9]):
GH ( f )= Vb2
(nf ).
(4)
From the structure of the accelerometer

of Fig. 2 follows, that Hooge's noise source
is present. This had probably impaired
envisaged sensor application [10]. If no
dc response is required this effect can be
eliminated by proper electronic design [5],
which improves SNR by order of magnitude.
Fig. 3. PSD of the very low frequency noise.
185
Measurement of Physical Quantities G.J.Stein
4. The Transducer Electronic Data Sheet (TEDS) concept

In testing and monitoring of large aerospace and engineering structures vast number of sensors
are distributed within the structure with star-like cabling connection to the analogue preprocessing unit. The cabling time and costs are rising and errors in sensor's identification may
arise. By the emerging IEEE P1451 Standard [12] use of low-cost proven communication
means, e.g. Ethernet LAN technology or RS 232 and RS 485 interfaces are standardised.
Various sensors with associated pre-processing electronics (e.g. filters, ADCs and preprocessing tasks) are clustered around an Interface Module, which then communicates with
the remote PC in standardised way [12]. The task of identifying and calibrating of a particular
sensor is furnished by the Transducer Electronic Data Sheet (TEDS), also defined by this
Standard. TEDS is a non-volatile memory chip built-in into the IEPE type sensor. The built-in
memory contains in 128 bits manufacturers data on sensors type and serial number, as well
as, a unique sensor identification code. Further 256 bits are in form of EEPROM, which can
be programmed in users calibration laboratory to store sensors sensitivity, calibration date
and parameters, location, etc. So the sensor is uniquely identified. Measured signal is
processed in same way as from an ordinary IEPE sensor, whereas digital data are retrieved by
reversing supply voltage and connecting the reed-out data to PCs RS 232 interface.
Processing software can use sensor identification and calibration data, sawing cost and time.
5. Acknowledgement
This contribution was compiled during author's short stay at the Department of Measurement
of the Faculty of Electrical Engineering and Information Technology, SLOVAK
UNIVERSITY OF TECHNOLOGY, Bratislava being on leave from the Institute of Materials
and Machine Mechanics of the SLOVAK ACADEMY OF SCIENCES, Bratislava. The kind
support of both institutions is gratefully acknowledged.
6. References
[1] ao, S., Kreidl, M.: Sensors and measuring circuits (in Czech). VUT Publishers, Prague 1996.
[2] Stein, J.: Microelectronic acceleration sensors. in: M. Karoviova and A. Plakova (eds.)
Proceedings of the 4th Intl. Workshop "Measurement '95" (Smolenice castle, May 1995),
Institute of Measurement of SAS, Bratislava, Slovak Republic 1995, p. 44.
[3] Stein, J.: New trends in acceleration sensors (in Slovak). Automatizace 38 (1995) 175-177.
[4] Berter, T., Kubler, J. M., Cuhat, D.: Kapazitiver Miniatursensor zur Messung nieder-frequenter
Beschleunigungen. KISTLER Instruments AG, Switzerland, 1993.
[5] Bernstein, J., Miller, R., Kelley, W., Ward, P.: Low-noise MEMS Vibration Sensor for
Geophysical Applications. Journal of MEMS 8 (Dec. 1999) 433-437.
[6] Hutyra, M.: Monolytic accelerometer ADXL50 (in Czech). Automatizace 37 (1994) 225-227.
[7] Doescher, J.: A High performance Surface Micromachined Accelerometer. Analog Devices Inc.,
Norwood, Massachusetts, USA, 1999.
[8] Gabrielson, T. B.: Fundamental Noise Limits for Miniature Acoustic and Vibration Sensors.
Trans. of ASME - Journal of Vibration and Acoustics 117 (Oct. 1995) 405-409.
[9] Harley, J. A., Kenny, T. W.: 1/F Noise Considerations for the Design and Process Optimisation
of Piezoresistive Cantilevers. Journal of MEMS 9 (June 2000) 226-235.
[10] Stein, G. J., perka, M.: Position Measurement By Micro-Electronic Accelerometer. in: Osanna,
P. H., et al. (eds.): Proceedings of the XVIth World IMEKO Congress. Vol. II (Hofburg, Vienna
Sept. 2000), Austrian Society for Measurement and Automation, Vienna, Austria 2000, 385-388.
[11] Schiefer, M. I., Moses, J., Freudinger, L. C.: Networkable distributed digital accelerometers for
aerospace and civil applications. in: Proceedings of the 17th Intl. Modal Analysis Conference,
(Kissimmee, FL, Feb. 1999), USA 1999.
186

Some Recent Developments in Acceleration Sensors: George Juraj STEIN

Uploaded by

Copyright:

Available Formats

Some Recent Developments in Acceleration Sensors: George Juraj STEIN

Uploaded by

Document Information

Original Description:

Original Title

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

Some Recent Developments in Acceleration Sensors: George Juraj STEIN

Uploaded by

Copyright:

Available Formats

MEASUREMENT SCIENCE REVIEW, Volume 1, Number 1, 2001

SOME RECENT DEVELOPMENTS IN ACCELERATION SENSORS

Measurement of Physical Quantities G.J.Stein

2. MEMS acceleration sensors

Fig. 1. Principal sketch of MEMS.

Fig. 2. Abridged layout of servo type MEMS accelerometer (after [6]).

MEASUREMENT SCIENCE REVIEW, Volume 1, Number 1, 2001

3. Noise and drift issues in MEMS sensors

From the structure of the accelerometer

Fig. 3. PSD of the very low frequency noise.

Measurement of Physical Quantities G.J.Stein

4. The Transducer Electronic Data Sheet (TEDS) concept

You might also like