Question about gyro adis16266

Thank you for your post. Perfect calibration is not possible, so this isn't a very helpful place to start evaluating possible solutions, such as the ADIS16266.  Typically, we have seen others start off with a specific angle displacement accuracy goal over a specific time period, while offering more detailed insights on the application that could enable error estimates based on "best calibration," along with an assessment of condition-dependent errors. If we were to assume a perfect bias calibration and zero movement, the in-run bias stability specification would tell you that your angle accuracy will be 16 degrees (one sigma) after 2 minutes.  Understanding the actual motion would allow consideration of the sensitivity, linear-g and perhaps, temperature-dependent error sources as well. I hope that helps!

Best,

NevadaMark

Hello, thanks for your answer.

My application is for a UAV,The UAV will aerobatic maneuvers. And also the UAV follow a target in the air (in motion). But only for 2 minutes. All automatically.

Do you know if these movements will have more mistake gyroscope?

calculating the orientation is important, as this calculation the path of UAV (since not use GPS). and the control algorithm for guiding.

For your specific UAV, do you have a feel for the maximum range needed on your rate measurements?  I suspect that it is no where near the 14,000 deg/sec available in the ADIS16266.  With that information, we might be able to offer a better suggestion, but for starters, I would consider the ADIS16448, which offers a fully-calibrated, 3-axis device. 

Best,

NevadaMark

If your vehicle does pitch and roll as well , one gyro will not be enough. It all depends on how accurate you need to keep the heading, how much space and power you got and how much money you are willing to spend.

If money is not an issue then you need a DTG, RLG or FOG, but be prepared to spend thousands and most of the good gyros are ITAR controlled. Sorry, but MEMS gyros are not good enough for anything but the least demanding applications.

I would argue, that it cannot be done with any degree of accuracy unless you use three gyros, sample very fast and use Euler rotation math to get the real heading (vector) of your motion. I could imagine, that the roll of your vehicle is pretty severe.

KurtBecker comments are related to where I suspected this conversation would lead. While RTG, FOG and other mature inertial sensing technologies still hold some performance advantages, advances in inertial MEMS technology has enabled a very broad range of UAV development, many of which are just now entering the market.  In particular, the size, weight, cost, power and in some cases, ease of integration have been enabling for these types of systems.  The trade-off, as mentioned, is in the performance of the sensors. Granted, that performance gap has narrowed quite a bit in the past 3-5 years, but complete replacement is still a ways off.  Therefore, there is value in being able to determine requirements, understand key sensor behaviors and design systems that enable optimal performance.  As more details are available, we are happy to review it and help you determine if inertial MEMS can offer value with respect to your conditions and objectives.

Very best,

NevadaMark

If MEMS are good enough for this application is only half of the question.

More important is the fact, that ANY single gyro is not good enough for the application.

It is fair to assume, that in a UAV application the attitude of the gyro changes. So it will be necessary to measure the attitude. This is done using accelerometer to measure the gravity vector. You need 3 accelerometer, but even this is not good enough. Any accelerometer which measures the gravity vector is very sensitive to horizontal acceleration. To distinguish between horizontal acceleration and gravity, you need two more gyros... for pitch and for roll. With other words you need 3 gyros and 3 accelerometer to do it right.

While the bias of the pitch and roll gyros can be estimated using the gravity vector (and a Kalman filter), there is no way to estimate the yaw-gyro bias without additional aiding like GPS.I do not know about the accuracy requirements, but without additional  aiding like GPS, I suspect the vehicle's heading will be quite a bit  off... even after as little as  2 minutes.

Just to follow-up on KurtBecker comments, I would offer the following two options, as a starting point: ADIS16448 or the ADIS16488.  The provide the function that KurtBeckerdescribes at two different price/package/performance points.  While continued posts and discussion are encouraged, I will likely mark this as "answered," until next-level insights are available on the original question. Many thanks to KurtBecker, for his contributions to this discussion!

Very best,

NevadaMark