Hi everyone. I'm new to DIYdrones. I'm looking for some help with my Iris+ issue. the Rep at 3dr has looked at my flight logs and determined everything looked fine and there was no issues. I'm trying to escalate the issue with 3dr but for now I wanted to see if the collective here has every experienced this issue. I had a few experiences with what I thought were flyaways.. turns out it may be a real thing. here is a video I made to send to the tech at 3dr who keeps telling me there is nothing wrong based on the flight logs. Please do not watch the entire video as the first few mins were waiting for the issue to present itself. I have the Iris+ connected to the planner and have it sitting level. I even move the Iris around to show that the horizon goes back to level and its actually active. at 7:57 the horizon suddenly tilts about 45 degrees with out the iris moving. I can try and tilt it in the opposite direction to compensate it but when I go back to level, the horizon is still 45 degrees off. the Yaw deg goes from over 350 value down to 0 and then back to 44. if I disconnect the planner and reconnect it, the issue is still there. only if I actually unplug the batter and then re hook it back up, does the issue go away temporarily. so what happened to me, if that I was flying and then I for some reason on a wide open area lost gps, then the thing suddenly took off to one side and I ended up crashing it vs causing damage to a house. I haven't owned it log and since march have only flown about 5 times because I cant trust it.
please watch the video and if anyone can tell me what they think, or how to approach this with 3dr that would be great as they don't seem to believe there is anything wrong.
https://www.youtube.com/watch?v=O2NJkqJT2fc
I have been back and forth with 3dr and am getting no where. I feel there is a warranty issue and don't know why they cant see the issue in the flight logs. the tech even said that that one of my videos I posted on line March 29th corresponded to a flight log for may 9th? not sure how he thinks that since the one video was uploaded in march?
Replies
I may have got a data log of the horizon spinning. Im not sure what Im looking for on this but these are the data logs from me plugging in the battery today. the larger file should be it, that's the one where is sat a bit ok. I tilted the iris to just make sure that it was reacting ok. then the horizon suddenly started spinning. I turned it off and then back on again and it was still doing it. see if you can tell me if you see it doing a barrel roll.
I hope this helps and you can figure out what its doing.
2015-05-25 13-54-34.bin
2015-05-25 14-05-25 45.bin
Hi Randy. the time that this occurs is totally random. sometimes when I turned it on, it would already be spinning. some times it would take 5-10 mins and sometimes not at all. Im just glad I was able to data log it. I was playing around trying to look at the different parameters and I saw one where the x axis was going like a wave on the graph. it was wild looking. I cant remember which one I was looking into right this moment. I just wish I had realize there was something wrong before the last time I took it out because that time it broke an arm and the gimbal when it went off kilter. I didn't post that video, but in it, you can see like my other flyaway, the arms of the quad all of a sudden show up in the footage and you can tell it was about to haul ass off, had I not downed it to prevent it from hitting the building. the tech at 3dr just sent me an email yesterday again saying that he didn't feel this was a hardware issue but they will start the RMA process. he also didn't seem to happy that I used the beta firmware because he said it was glitchy.
I really appreciate you looking into this. if you all can figure out any more info on it, please let me know. even though they are replacing it, I would still like to know what happened to mine. if not for anything, but for at least piece of mind.
It certainly looks like a hardware issue but we will ask Tridge and Paul Riseborough tomorrow if they have any ideas.
In these two logs it starts rolling about 1 minute after the board has booted up, but in your video I think it took much longer so perhaps it's just a coincidence that in these two logs it happens at roughly the same. Do you notice that it stats to roll after a certain amount of time or is it quite random?
Thanks Erick, I see in the logs a similar scenario to that reported by someone else with this problem (try graphing IMU->GyrY and IMU2->GyrY). Hopefully someone with more knowledge than me can figure out whether there is anything that can be done about it.
Erick,
I think there are two separate issues going on here.
The near fly-away is caused by GPS glitches. At the time of take-off the vehicle only had 6 satellites which is very low. I'm slightly surprised that the GPS was reporting a good position (i.e. low HDOP) so this is a lesson perhaps that hdop cannot always be trusted. There were multiple GPS glitches in the log including at the time of the near flyaway.
Re the video with the bizarre sudden change in attitude - I strongly suspect this is a hardware issue with one of the accelerometers and you should RMA the whole vehicle with 3DR. If you purchased it between June 2014 and Feb 2015 there's a known manufacturing issue that can cause one of the IMUs to fail. There's no 100% reliable software solution so replacing the Pixhawk on the vehicle (or the whole) vehicle is the best thing to do. 3DR has been pretty good about accepting RMA of boards/vehicles built during this time.
To be sure it's an IMU failure we need a dataflash log of it failing. Below is a screen shot of how to set the LOG_BITMASK to "All+DisarmedLogging". Make sure to push the "Write Params" button after changing the drop-down.
Thanks again for your report, the video in particular of the attitude shifting 45degrees was something I'd never seen before.
If that doesn't sort it out, contact customer service and they can give you a RMA for a replacement.
Hi Chris. thanks for the suggestion, I will see what I can do with my limited knowledge base to update to 3.3. My concern is why is this giving the bad data? I mean should there even be " Bad Data" ? My concern is that 3dr is telling me there is nothing wrong and I am worried about liability if this crashes on someone and they are telling me its ok? In my march 29th video, here https://www.youtube.com/watch?v=RHu8v2rUmTk at min 7:23 you can see it was heading right for a group of kids. I was finally able to fight the bad yaw and at least get it to level out. it was lucky it didn't crash into those kids.
Sensors vary both in performance and under certain environmental conditions, which is why IRIS+/Pixhawk has redundant sensors. The updated code does a better job of discarding data from sensors that are misreporting at any given time, and instead using the ones that are performing better.
Do you use a Mac or a PC? I can walk you through the update process either way.
-
1
-
2
of 2 Next