Posted by Luke Wiseman on December 11, 2014 at 5:03pm
Anyone else have issues with the Iris+ occasionally pointing in random directions while flying autonomously? I usually make missions with spline waypoints, and it usually flies smoothly, just doesn't always face forward while flying. I've also made missions with a region of interest, and sometimes it will focus on the region and other times it won't. Focus comes and goes during flight. I should also note that I recently upgraded the firmware to 3.2. I loaded the firmware, set params to iris+, performed a compass calibration, accel calibration, radio, and ESC. I took it outside, loitered for 5 minutes, landed, disarmed, re-armed, took off again. Loaded a test mission (no ROI). Flew it fine minus the random headings it had through the whole thing. I then loaded a test mission with a ROI,started off facing towards the ROI, then it just started facing whatever direction it wanted to. Any thoughts? Should I redo my accel and compass calibrations? Debating on going back to the previous firmware that my iris came with and see if the problem is still there.
You need to be a member of diydrones to add comments!
Yeah I may try recalibrating the compass again and see what that does for me. I would have to guess that is the issue. Even when I was doing my first initial 5 min loiter, after about 3 minutes, the Iris started slowly rotating. I chalked it up to the compass fine tuning itself?? I should note I was using droidplanner to make the missions. Mark, I'll have to check what version it came with when I get home from work this evening. I received the "factory" firmware directly from 3DR about a month ago. I had uploaded a beta firmware to the Iris before I had a little more experience and successfully screwed the whole thing up haha. I never flew it with the beta, but the pixhawk was not happy with it. Kept getting a low battery alarm even with a fresh full battery. Anyway, yeah I'll see if I can find out what version that is that they sent to me. Thanks for the replies guys. Who knows, I may take the Iris back out today after work and it may be fine. I'll let you guys know
Sorry, I have an upgraded IRIS and started with 3.14. In that version ROI would only last for that way point.
That latest version 3.2 changed that and the IRIS+ came out with a modified version of 3.2 before it was fully released.
I think there is a note out to do a firmware upgrade to the current 3.2 version.
Otherwise will need a log of the issue to determine what it was doing at the time.
Mark Prince > Michael BurmeisterDecember 12, 2014 at 6:02am
I also have an upgraded Iris. I have been running 3.1.4 in the Iris since last January 2014. About 4 weeks ago I installed the hardware upgrade and then flew it for a few weeks with 3.1.4. A week ago I made plunge and upgraded to 3.2. Haven't finished all of the testing yet due to work travel but I made a list of all the tests I wanted to perform in order to decide that 3.2 is stable enough.
I had an absolutely flawless platform with 3.1.4. In my profession firmware upgrades are very thoroughly tested for long periods of time before they are put in production for boiler and turbine control systems in power plants. We never upgrade unless there is overwhelming reasons. We don't take the vendor's work either, we use test beds that represent the production environment which is really what we do as open source users.
I haven't done the exact mission described above by Luke but I did use droid planner to set an ROI and then did the circle command around the ROI instead of several waypoints. It worked as advertised. I did test a circle of waypoints without an ROI but that wouldn't duplicate Luke's situation.
Michael has some good info above that I wasn't aware of.
I'm not near finished with my list of tests yet but so far the only negatives I've found with 3.2 are:
1) One of the incentives for me to upgrade was to be able to download uncorrupted log files via USB mavlink but that doesn't work at all. I get the error "cannot get log list". Disappointing but definitely not a show stopper. I'm not complaining, glad to contribute to the cause:)
2) I seriously doubt this has anything to do with the 3.2 upgrade but I have more testing to do. One of the items on my list was fly the Iris with a front mounted go pro and then do an autotune. When I put the go pro in record mode the copter goes berserk in loiter mode. Just takes off in random horizontal directions. I think it's some kind of GPS interference but I didn't have it before. Today I'll will focus on that test and I'm sure I'll figure it out. My Iris has a GPS mast that protrudes through the top shell which could invite some go pro interference but I never noticed any problems before the upgrade.
Version 3.2 change the way ROI works. Previous it would only last past one way point. In 3.2 they persists until you cancel them with an ROI of all zeros for Lon,Lat,and Height.
Have not seen where the copter points in random directions while flying an auto mission except after using Droid Planner and the Follow Me function.
There is also a parameter you can set to change how it flies between way points but that's about it.
Replies
I too had the issue with the copter slowly turning until it flew off on its own and crashed.
I found out later that through mission planner I had set the compass orientation incorrectly.
After fixing that it was rock solid.
I have been running the beta code for a couple of months now and I have found little things that ether got fixed or I can work around.
Like it so far.
Sorry, I have an upgraded IRIS and started with 3.14. In that version ROI would only last for that way point.
That latest version 3.2 changed that and the IRIS+ came out with a modified version of 3.2 before it was fully released.
I think there is a note out to do a firmware upgrade to the current 3.2 version.
Otherwise will need a log of the issue to determine what it was doing at the time.
I also have an upgraded Iris. I have been running 3.1.4 in the Iris since last January 2014. About 4 weeks ago I installed the hardware upgrade and then flew it for a few weeks with 3.1.4. A week ago I made plunge and upgraded to 3.2. Haven't finished all of the testing yet due to work travel but I made a list of all the tests I wanted to perform in order to decide that 3.2 is stable enough.
I had an absolutely flawless platform with 3.1.4. In my profession firmware upgrades are very thoroughly tested for long periods of time before they are put in production for boiler and turbine control systems in power plants. We never upgrade unless there is overwhelming reasons. We don't take the vendor's work either, we use test beds that represent the production environment which is really what we do as open source users.
I haven't done the exact mission described above by Luke but I did use droid planner to set an ROI and then did the circle command around the ROI instead of several waypoints. It worked as advertised. I did test a circle of waypoints without an ROI but that wouldn't duplicate Luke's situation.
Michael has some good info above that I wasn't aware of.
I'm not near finished with my list of tests yet but so far the only negatives I've found with 3.2 are:
1) One of the incentives for me to upgrade was to be able to download uncorrupted log files via USB mavlink but that doesn't work at all. I get the error "cannot get log list". Disappointing but definitely not a show stopper. I'm not complaining, glad to contribute to the cause:)
2) I seriously doubt this has anything to do with the 3.2 upgrade but I have more testing to do. One of the items on my list was fly the Iris with a front mounted go pro and then do an autotune. When I put the go pro in record mode the copter goes berserk in loiter mode. Just takes off in random horizontal directions. I think it's some kind of GPS interference but I didn't have it before. Today I'll will focus on that test and I'm sure I'll figure it out. My Iris has a GPS mast that protrudes through the top shell which could invite some go pro interference but I never noticed any problems before the upgrade.
What version did your Iris+ come with? IOW - what is the previous firmware that your Iris+ came with?
mp
Version 3.2 change the way ROI works. Previous it would only last past one way point. In 3.2 they persists until you cancel them with an ROI of all zeros for Lon,Lat,and Height.
Have not seen where the copter points in random directions while flying an auto mission except after using Droid Planner and the Follow Me function.
There is also a parameter you can set to change how it flies between way points but that's about it.
That's a compass calibration thing. I'd calibrate it again.