Dutch In English
Home page Organizational Structure Launches Projects Rocket motors Pictures NERO History Join NERO Definitions Safety Links Sitemap Search
NERO Research Gyroscopes

Characterization of the Murata solid-state gyroscope

1 Introduction
2 Test setup
3 Results
4 Conclusion

Since ten years the company Murata from Japan produces compact low-budget solid state gyroscopes, to be used for picture stabilization in video camera's and in car navigation systems. Because of their very small size these gyroscopes are ideal for application in amateur rockets. Last year the price of these devices has dropped dramatically and therefore made it possible to use this device in amateur rockets.




Introduction [Top] [Contents]

There is a problem with the Murata sensors: the datasheet of the manufacturer gives very little information on the behaviour of the device, as compared with (for example) the ADXL-50 accelerometer of Analog Devices. An attempt to get data from the manufacturer via the Dutch distributor failed, possibly because they consider application in amateur rocketry not commercially interesting. For application in amateur rockets it is of paramount importance that the sensor can function during or at least survive the high vibration and static acceleration levels generated by the rocket motor. According to the data sheet, the sensor is specified for a sine vibration level of only 1 G (10-1000 Hz) and half-sine shocks up to 100 G. If this would be a hard limit for operational use of the sensor it can not be used in an amateur rocket, simply because during motor operation vibration levels between 5 and 10 G (up to 1 kHz) are realistic values. (Assuming the sensor can be protected against the effect of pyrotechnic shock.).

Blue=new, Red=visited Follow this link for more information about the Murata gyroscopes
Blauw=nieuw Rood=bezocht Read about the use of gyroscopes in our rockets.

The relevant specifications of the ENC-05EA are:
Category Subject Specification
Power supply Voltage 5 V
Current consumption max. 5 mA
Output Full scale range +/- 90 degree/sec
Sensitivity 1,11 mV per deg/sec (+/- 20%)
Bias 2,3 V
Linearity Linearity 5 % max
Temperature Operational range -5 to 75 degreesCelcius
Drift by above +10% tot -20% van de volle schaal
Physical data Mass 2,7 gram
Size 22 x 9 x 8 mm3

The Murata ENC-05EA is a rate gyro: the output signal is proportional to the angular velocity about the (single) input axis.Other characteristics of the sensor for which we lack information are its cross sensitivity, i.e. the sensitivity for rotation about axes perpendicular to the input axis, and its thermal drift as a function of temperature.



Test setup [Top] [Contents]

I t would be desirable to have a kind of turn-table test bench for accurately measuring the scale factor, better than the specified 20% anyway. In order to get, as a first step, an indication of the vibration sensitivity of the sensor a vibration test facility has been built. A description of the test setup, test results and a discussion are presented below.

The vibration test setup consists of a heavy 300 W bass loudspeaker which has been transformed into a shaker facility for small masses. In the centre cone of the speaker a polyurethane foam platform has been glued, providing a mounting plane for a sensor head and instrumentation. The instrumentation on the platform consists of a small accelerometer mounted on a PU foam cube (2x2x2 cm3) together with the sensor under test and a conditioning amplifier. The bass speaker is driven by a 60 W audio amplifier which has a sine-generator at its input. The level of vibration is measured by putting the accelerometer signal on an oscilloscope. By observing the gyro sensor output simultaneously the disturbance of the sensor can be determined. During this test the sensor is not subjected to rotation, i.e. angular velocity about the input axis was zero, and the vibration axis is perpendicular to the input axis. The sensor pcb is provided with an AC-coupled conditioning amplifier suggested by the manufacturer. The amplifier has a bandwidth of approximately 1 kHz. The output signal of the sensor after the conditioning amplifier is 11,5 mV per deg/sec.

With this set-up two experiments have been performed: (1) A resonance scan from 20 Hz to 650 Hz at a constant sine vibration level of 2,5 G amplitude. The intention was to determine for which frequencies the sensor is most sensitive. The resonance frequency is supposed to be the frequency for which the sensor is considered the most susceptible to vibration damage. (2) A linearity measurement at the frequency of resonance. The intention is to determine the highest level of vibration that the sensor can be subjected to, without causing damage.




Results [Top] [Contents]

It appears that the gyroscope behaves like a microphone: mechanical vibrations are picked up and translated into proportional disturbances superimposed on the regular (zero) output signal. The result of the resonance scan is shown in figure 1.

Figure 1 - Resonance scan ENC05EA

A low Q-factor resonance is observed at a frequency of 109 Hz (+/-10 Hz). The amplitude of the disturbance varies between 0,06 and 0,26 deg/sec, which is less than 1 promille of the full scale signal of the gyroscope. The result of the linearity measurement is shown in figure 2.

Figure 2- Liniarity resonance peak ENC05EA

The amplitude of the disturbance is practically proportional to the amplitude of the vibration, and this result is also reproduceable. Up to a sine vibration amplitude of 29 G remains functioning. For this frequency this G-level is also the limit of the test set-up. The disturbance at this level is still less than 2 deg/sec. Taking 30 seconds for each step increase in G-level, the overall duration of the linearity test was approximately 5 minutes. The functionality of the sensor is checked by means of a simple test: the sensor is rotated about its input axis, by hand, and the order of magnitude of the output signal is determined. When "measured" this way, no difference is observed between the behavior of the sensor before and after the vibration tests. Of course it is possible that the vibrations causes some subtile damage to the sensor, however that can only be determined with a more accurate rotation- or pendulum set-up..




Conclusion [Top] [Contents]

The Murata ENC-05EA is likely to resist sinewave vibrations with amplitude up to at least 29 G maintained during at least 30 seconds, when applied at the resonance frequency of 109 Hz. Other resonance frequencies below 1 kHz have not been found. The disturbance in the output signal is proportional to the applied (sine) vibration amplitude and the proportionality factor is maximum 0.07 deg/sec per G. If the vibration levels in the rocket at the position of the sensor are known, the disturbance in the output signal can be determined. For this reason, and because the sensitivity for disturbances is minor, the ENC-05EA sensor is suitable for use in amateur rockets. Note that, due to the "microphonic effect" mentioned above, the sensor can as well be used to measure on-board vibrations.

These measurements have been performed in the summer of 1998. It is intended to continue the experiments by subjecting the sensor to random noise and shocks, and to measure the temperature drift coefficient and cross-axis-sensitivity in a calibration set-up. If anyone has other sources of information on the environmental characteristics of these sensors I would be pleased to know.