I’m in the early stages of flight testing a custom large (~20kg) quadcopter platform and doing my initial tuning and log analysis etc.
Everything is flying well at the moment and the progress is good. However, the logs show gaps in the IMU data, and when I try to use the online filter analysis tool, I get the warning message ‘Not enough continuous IMU data available’, and the tool provides no insight into the vibration levels as a result.
See the below image for GYRZ data. For the first 20 seconds or so of the flight, the data is missing a lot of samples and so is kind of useless when trying to analyse this for performance. Sometimes a full half second of IMU data is missing. It does seem like it pulls itself back together in the last half of the flight but there are still definite gaps.
It will do the correct configuration of that and other parameters for you, and you will always see exactly which parameters get changed and why they are getting changed. Try it out.
User clearly has some knowledge and background and asked an informed technical question. That software isn’t the only answer, and you gave no indication of the actual problem at hand. “Just use AMC” can’t possibly be the single answer given for all questions.
What has occurred with other users recently is they have configured for Batch Logging and came across Raw IMU logging, enabled it and didn’t disable Batch Logging. That is what is configured here. Also Raw IMU is set in the Log bitmask along with Fast Harmonic Harmonic Notch Logging hence the “check all the boxes” comment.
Unless I’m mistaken here, and previous discussion has been a bit confusing so it’s certainly possible, perhaps there should be a note added to the RAW IMU Logging Wiki page about this.
When I use Raw IMU (INS_RAW_LOG_OPT) I use default for the Log Bitmask and disable batch logging (INS_LOG_BAT_MASK).
@Yuri_Rage Yes, it’s not the single answer for all questions.
But it does solve a lot of questions if followed.
@dkemxr INS_RAW_LOG_OPT is recommended by @iampete for big vehicles with powerful enough processors. INS_LOG_BAT_MASK is legacy and should be used in other cases.
AMC automatically selects the correct one based on processor and prop size.
@dkemxr the problem with the wiki is that most users do not read it. AMC automates it so that even the users that do not read the wiki, get the correct settings. For example, this user clearly did not read the documentation.
Thanks for sharing. I have tended to stay away from these tools through my work with AP. I am typically involved in the more ‘unconventional’ configurations: using engines, no ESCs, no electric motors, and so on which makes these tools quite difficult to interact with. I am messing around with the tool for the upcoming project as it is more ‘conventional’, though I find it much easier and better for my understanding to work through the documentation as and when it is relevant.
Just a little feedback, it’s quite difficult to use with popups covering text:
Not quite. I have read the docs on this previously, and after reviewing these docs again just now, I see the line you’re mentioning and I must have missed it or forgotten about it. I am a human and I make mistakes.
I would argue that it can be quite confusing. Between INS_RAW_LOG_OPT, INS_LOG_BAT_OPT, LOG_BITMASK, there are a lot of similar sounding options of logging raw data, sensor rate data, multiple mentions of pre- and post-filtering… I’m not saying the documentation doesn’t address this, simply that I can see why the dicussion around this raises confusion and mistakes amongst users.
Please don’t make assumptions about how much of the documentation I have or have not read. Accusations with no understanding of peoples experience or circumstances will not solve technical problems.
I have been reading various aspects of the documentation as needed over the course of the last four years. Like I say - I am a human and I make mistakes. Things get forgotten, text gets skimmed, buttons get misclicked.
I will adjust my logging options before my next flight to see if this addresses the problem. I’ll report back after next flight tests.
