ADIS16480 gyro issue

Thank you for your post.  It might help to use the EVAL-ADIS and IMU Evaluation Software to validate your observations in parallel with your effort to observe this IMU's behaviors with your embedded interface.  I am not sure I understand the problem associated with this part of your statement.  Since this is an inertial sensor, shouldn't we expect a response when the device is in motion?  Perhaps there are more relevant details that we can use to help you close this gap?

• ORIGINAL: There are continuous variable values unless the ADIS16480 has a fixed position without varying it.

With respect to the following, I would address this before doing anything else. 

• The values were good but there are some dropped readings.

This typically relates to some sort of timing or signal integrity issue in what the slave (ADIS16480) is receiving. Perhaps this blog may be helpful:

Connecting to Smart Sensors, it Takes More Than Code 

Dear NevadaMark,
Thank you for your kind response.
Sorry for not explaining the issue in a proper manner. I tried to check the gyro_x values using the evaluation board and IMU Evaluation software and the results were suffiecient.
But when reading using the break out board with my arm_cortex based board the gyro_x readings drift, and some readings dropped.

So I have two questions to ask:
1- What is the reason for dropping readings in gyro?
2- Why did the gyro readings drift ?

1- What is the reason for dropping readings in gyro?

Thank you for clarifying this question.  I sort of thought that is what you mean, so I would start with the information I offered in my previous reply.  

• This typically relates to some sort of timing or signal integrity issue in what the slave (ADIS16480) is receiving. Perhaps this blog may be helpful:

  Connecting to Smart Sensors, it Takes More Than Code 

2- Why did the gyro readings drift ?

No sensor, that I am aware of, has zero drift, so can you quantify the differences in your observation, between the evaluation system and your system?  In other words, how much is it drifting?  Can you share a plot or some data, which is easy for us to import into Excel for quick analysis?  To be honest, I would address the first problem (drop outs) before spending time on the drift.  Any observation and analysis of sensor behaviors will be handicapped if we cannot rely on communication integrity. We have run these units for days, at maximum sample rates, without losing a single data point, so when we encounter what appears to be a bad data point, we investigate it fully, before doing anything else.