Skip Navigation. Press[2]key.

FAQ for gyro | Silicon Sensing Systems Japan - MEMS ring rate gyro

FAQ

HOME>FAQ

Please choose the category.

Question

Q.Bias and Drift
A.Bias is the gyro's output when the gyro is not rotating, i.e. zero point.　 True for any sensors, the bias moves by itself without input due to sensor's instability, which is the "drift". In addition to drift under constant temperature, it is normal to see extra drifts due to, for example, temperature change or external vibrations seriousness of which depends largely type of gyros. If the repeatability is good (i.e. drift is quite repeatable under known environment), it can be compensated externally. If the repeatability is no good (e.g. histeresis is big) external compensation will not be easy. Methods of compensating the drifts depend on application (and typical dynamic movements thereof), therefore will be the user's task (hence user's know-how).

File>>

PAGE TOP
Q.Resolution
A.How small the angular velocity can be detected, means how small the gyro output can be read: Therefore it is determined by the SN ratio of the gyro output. For example, if the gyro's dynamic bandwidth is 10Hz and noise within 0 to 10Hz is less than 1mVrms, and the Scale Factor is say 20mV/(deg/sec) for the power supplly of 5.00Vdc:

1mV / 20mV(deg/sec) = 0.05 deg/sec This is the readable signal for the bandwidth of upto 10Hz

Similarly, if the signal is ideally filtered for the bandwidth of 0 to 20 Hz,the noise can be considered about twice, hence the readable signal is 0.1deg.sec, and in the same logic　for 0 to 5Hz the noise is less than half therefore better than 0.025deg/sec. Here we are assuming typical noise.

PAGE TOP
Q.Why is this gyro strong against outside vibration and shock?
A.Our gyro's resonator (the sensing element) has no fixed support but is suspended in the space, and the vibration(primary and secondary) is of synmetric deformation, which are unlikely to be caused by external means. A Buddhist temple bell hangs under the cradle. Hitting the bell gongs the bell. Hitting the pillar of the bell-cradle does not gong the bell.

PAGE TOP
Q.Can the gyro be installed in any attitude?
A.Yes. Our gyro can be installed in any attitude. (Be aware other gyros may have limitations due to the effect of gravity, for example)
PAGE TOP
Q.Could two or more gyros close-by affect each other?
A.Our gyro element is very small and is vibrating synmetrically, therefore any resonance effect, if any, can be considered negligible.
This may not be so with other types of gyros.
PAGE TOP
Q.What's ratiometric?
A.It means the output bias and Scale Factor are proportional to the supply voltage.

For example, for CRS03-02, when the supply voltage is 5.00Vdc, the Scale Factor is 20mV/(deg/sec)
when the supply voltage is 4.90Vdc, the Scale Factor is (4.90/5.00) x 20 = 19.6mV/(deg/sec)
Similarly for Bias , if the supply voltage is 5.00Vdc is B, the Bias of the same gyro
when the supply voltage is4.90Vdc will be(4.90/5.00)*B

It is difficult and expensive to make a power supply that sustains 5.00Vdc in all situations (temp range, for example), but by sharing the B+ of gyro and AD converter it is easier to sustain good reading of gyro output.

PAGE TOP
Q.Is CRS03 affected by magnets near-by?
A.True that CRS03 resonator uses magnetism to vibrate, but such magnetic flux would have be very strong AND close (as 100 micro-meter order) to be affected. Therefore effects of mangets nearby can be neglected. If you have "peaking" type magnetic noise, like high current DC motor adacent to CRS03, please test enough, being careful. Note also that CRS03 does have magnetic flux from itself.
PAGE TOP
Q.Any interface necessary?
A.The gyro itself works without any extra components. Just supply 5.00VDC and the output DC (analogue) comes at approximately 2.5VDC center. This is in many cases enough for measurements. Only when you like to get rid of noise, please try the following.
The higher band noise contained in the gyro output is unrelated to the dynamic accuracy of the gyro. It is simply good to remove the noise with low-pass filter, at the bandwidth of your requirement (for example, CRS03-02 dynamic bandwidth is 10 to 20Hz). It is good to put an anti-aliasing analog filter before feeding the signal into AD converter in which digital filtering will be done. The anti-aliasing filter should be of small phasing effect. The noise on the gyro output remains high upto around 800 or 1kHz, therefore unless your sampling is really fast, anti-aliasing analog filter is useful. On AD conveter, considering that AD analog input range is typically 0 to 5VDC (gyro's dynamic output is 0.5 to 4.5VDC), linearity, and the gyro's accuracy, recommended is 14 bits or more. For digital filtering, you may like to try taking moving average. For details, see the file below.

File>>

PAGE TOP
Q.Please define Non-Linearity
A.Input vs Output is ideally perfectly linear. In reality perfectly linear is not possible.
Draw the best fit line on the actual input-vs-output curve, and we define this slope as the actual Scale-Factor.
The biggest gap between the actual curve and the best-fit-line, divided the full-scale in % is defined as non-linearity.
See the file below.

File>>

PAGE TOP
Q.Is there any sensitivity of the sensor to AC components on the power supply? I understand the ratiometric characteristics of the sensor and the feed through of AC signals as a result, but are there any other considerations. For instance, will there be some affect on the operation if there is AC noise near the oscillation frequency? If so, what is the oscillation frequency and how low does the noise have to be so as not to have an effect?
A.The sensor element is operating at 14kHz clock therefore if any noise near this frequency exists on the power line, you will see the "beat" on the sensor output.
Power supply recommended (3-pin regulator) is series type and not switching-type. For example, a commonly available 78 type 3-pin regulator. Recommendation is to add a 47uF tantar-capacitor (or more) on the output of 3-pin regulator which may be the common power supply to plural gyros, AND add a 0.1uF ceramic capacitor on the each power supply line of each gyro.
In the past there have been cases where 3 gyros were used together and we did not see any apparent output noise increase.
Note that each gyro takes about 100mA or more as initial current therefore you might like to use a 78 series of about 1 amp. Gyro's PLL may not lock up well if the supply voltage does not go up to nea 5.00V immediately while 100mA is flowing.
Noise of about several hundres Hz or less would remain on output somehow as ratiometric output, dependent on frequency response. therefore ripples on power supply should be carefully managed. Common impedance on the power supply or ground line will look like cross axis sensitivity (if your gyros are used on different axis: which is normal).
PAGE TOP
Q.What are the AC components of the load on the reference supply? Again, is there an AC component of current at the oscillation frequency (or any other frequency) and what is the magnitude of that load?
A.Since CRS03 has a filter circuit that includes tantar capacitor internally, you will not see any significant AC component in the output. But if you look at the output carefully with FFT analyzer you might see non^significant amount of 14kHz components and its harmonics.
PAGE TOP
Q.What are current applications?
A.Our family of gyro products are used for: Vehicle dynamic control for automobile, ITS, bipedal robots, AGV, unmanned heli, satellite-aided-compasses, marine radars, marine auto-pilots, stabilised antennas, inertial systems, agricultural vehicles, autonomous robots, SegwayHT, etc. Many of current users chose our gyro for its resistance against external vibration, besides good temperature stability.
PAGE TOP
Q.Where are the gyros manufactured?
A.All CRS03/07 are made in Japan at Silicon Sensing Products (productions dept) which is at the same address as Silicon Sensing Systems Japan, inside Sumitomo Precision complex.
PAGE TOP