I am trying to use ADIS16480 . Only yaw angle which is read via SPI keeps drifting. Roll and Pitch angles are stable. Do you have any suggestion about this situation?
I would start by issuing a TARE command (see GLOB_CMD register) while the unit is stable. Let me know what you observe. NOTE: I am out of the office so my replies might be delayed.
I tried TARE command. Nothing changed. I also reset ADIS16480, changed EKF reset bit and disabled magnetometer(EKG_CNFG) but it is same. I also observed that last element(q3) of Quaternion keeps drifting.
The yaw angle is most dependent on the magnetometer inputs, since the accelerometers don't change with orientation around this axis.
How fast is the angle drifting? Does it stop, slow-down or speed up? Can you post a data file that contains the accelerometer, gyroscope, magnetometer and angle outputs during the drift? Also, can you post a register dump for all configurable values?
NOTE: I am not going to be able to do much more with this for another week, as I must focus on some important family matters, but I will see if any colleagues are available to take a look at this & offer insights.
I attached video file which show outputs values during drift. I am not using EVAL-ADIS board.
Thanks for the video. Can we expect more posts that provide quantifiable observations in answering my previous questions?
Sorry for the short answer.
- Configurable Register Values
As i said before i do not use EVAL-ADIS board to get data from ADIS16480.
So I can not use your IMU Evaluation Software. Because of this situation getting all configurable
values takes time. I'm working on it right now. Actually , i am gonna compare all configurable values
with their default values that are indicated at ADIS16480 datasheet. If there is a difference between
any of them , I will inform you.
- Yaw Angle Drifting Problem(Observation)
When ADIS16480 is standing , I power up and get the values. In this case, they are almost stable
but when i observe values carefully, I see that the yaw angle is decreasing approximately monotonous.
This drifting is too slow(less than 1 degree/minute ). After that i change position of ADIS16480 and
it responds quickly. For example when it shows 20 degree , I change position and i get 70 degree
immediately. Everything starts from this point. Yaw angle value starts to change while ADIS16480 is
standing. However speed of change is depeneds on how fast i change the position of ADIS16480.
I compared configurable user register with their default values that are indicated at the Table 10 in ADIS16480 datasheet. I could not see any difference. All values match. So, I can not see any problem here.
Is there anyway to go back factory settings for ADIS16480? Do you have any suggestion to try?
It sounds like the device already matches the factory default. There no way to go back to the factory default settings with just one command.
So, what can i do right now in order to get meaningful output? Do you have any suggestion?
With respect to helping, I will be glad to hep, but I still don't feel like I have a clear problem definition or data that helps me understand the behavior and the environment that the device is operating in. Without that, it is tough for me to do much.
I posted a video that shows behavior of ADIS16480 and I checked all configurable registers. What kind of data you want from me? If you write them clearly , I will try to supply them for you.
PROBLEM : yaw angle is not stable(changes) when IMU is standing.
If you were in my shoes, how would you use that video? I want to help but that isn't what I asked for. In post #3, I asked for a data set that captures the raw sensors (accelerometer, gyroscope, magnetometer) and the angle drift together. By data set, I mean something that I can analyze in Excel, something that I can use to clearly identify the behavior to my colleagues. If the drift follow a specific maneuver, make sure that maneuver is is part of the data. A video of the maneuver itself, with some annotation that helps us correlate to the data might help as well. I hope that makes sense and look forward to your reply.
I added two data file. They are recorded while ADIS16480 standing and moving. Files include raw sensor values and in order to show values hex representation is used. First column of data file is timestamp in millisecond. It is added to help you while examining data file.
Actually we have same problem sometimes. Also we do not use eval board.
If you have conclusion to this problem, could you kindly share with me.
I am sorry that I have not responded but we seem to be having a hard time connecting on what information we need, in order to help you. I asked for data, which I can easily import into Excel. The data files that you posted contain text information that has hex in a format that I am not able to put into Excel (0x0000) easily. Perhaps I am doing something wrong, in my importing process, which you can help me with. Any chance of putting this in an Excel-friendly format?
hi I have same problem.
hi Nevada. could you send example file what you want. I will send a file to you for this problem
Thank you for your post. Can you start a different discussion thread? We are happy to help you and want to keep these discussions clear for your (and our) future use, so we are trying to keep these discussions focused on one specific question/topic, which is associated with one specific requester. I thank you for this, in advance and will look for your post!
My suggestion is to imagine that you are in my shoes and have zero visibility into your observations, setup or the things that you have tried to influence the behavior. I do have familiarity with the part and I do have access to the people who designed it. They are willing to help, but have no time to chase relevant information, so I am going to try to help gather what they need, assuming I cannot solve the issue (my solution capability is ~90-95% but this topic typically requires expertise that I am still developing). So, I encourage you to quantify everything that might be relevant, using the following list as a starting point. NOTE: It is fine to share a starting post, while you work on this in parallel, with the information you have right now, as well:
Thank you, very much, for trying out the ADIS16480 and for offering us an opportunity to help you improve your experience. I will do my very best to help you and only ask for this information to help speed the overall process, especially if I need to ask for help. I look forward to your new discussion thread!
I am terribly sorry that we seem to have dropped this communication. If you are still working on this, I will be glad to take a fresh look at the data that you posted. Many apologies!
Could you have a time to tend my last post? I could not make any progress. So, what should i do in order to solve this problem? We have deadline on this IMU project and it is coming. If you have another suggestion to try, I will try them.
Retrieving data ...