Not sure whether V3.5 Rc1 release covers the AUAV-X2.1.
Loaded Copter V3.5 Rc1 onto my AUAV-X2.1 flight controller via Mission Planner. Noticed in the HUD, pitch appeared as roll and vice versa. Changed AHRS-Orientation to ‘2’ yaw 90 degrees. HUD then appeared correct. The flight controller has been installed in standard orientation.
Problems encountered calibrating compass. Disabled internal compass and used the Drotek gps XL compass and did the compass dance. I did try to use the ‘Onboard Mag Calibration’ but couldn’t get it to work (probably lack of knowledge on my part). Offsets are high x -276, y -326 z -54. Although I clicked on ‘Externally mounted’ I’m not sure whether the Drotek or internal compass is being used. I left orientation as ‘none’ for the external compass. The arrow mark on the Drotek gps/compass points forward.
After setting everything up including ‘Motor emergency stop’ on channel 8 I took my quad to the open fields to test.
Obtained around 18 Sats and HDOP around 0.6.
Flew quite nicely in Stabilize and AltHold.
I have Loiter setup with Super Simple. In this mode pulling back on the right stick made the quad go forward away from me and pushing the right stick forward brought the quad towards me. Left and right roll were also reversed. Not sure whether this has anything to do with me setting the ‘AHRS-Orientation’ to ‘2’
Tried Roll autotune and it worked well - PIDs updated. Also successfully ran a Pitch autotune.
What did surprise me was that after running Log Analyser the ‘Max mag field length’ was 571.75. I’ve never had it this low for the same area before on my other quad.
Jon,
I guess this is the first time you’ve setup ardupilot on this flight controller? It would probably be best to use AC3.4 instead. The reason is just that it makes it easier to separate AC3.5 beta testing from the regular vehicle and board setup issues. Put another way, if you use AC3.4, because it is well tested, we can immediately know that any issues are most likely setup issue and not software bugs.
Randy,
Thank you for your reply. I did initially load AC3.4 but it came up with an error:
‘Check BRD_Type: no i3gd20 found’. I’ve just reloaded AC3.4 to double check.
I’ve now reloaded AC3.5 Rc1. In the parameters - BRD_TYPE it has No. 5 which relates to the PixHawk Mini. There is no number for the AUAV-X2.1. I assume this means that currently Arducopter cannot be run correctly on this flight controller.
This is the new AUAV board, right? It’s quite possible it isn’t supported. I think (but I might be wrong) that no one in the team has received one or even the schematics (I’ve looked in their website and it’s not available either). I thought that board was delayed, I wasn’t aware it was out now.
@OXINARF
Yes it’s the new AUAV-X2.1. I managed to purchase one and was hoping that it was supported in Arducopter. I appreciate that sometimes these things take time. I’ve got it installed on what’s now a test frame and don’t mind trying things out.
I think Tridge has one of these and I believe it should work because we have all the drivers for it. I’ll add it to our list of boards to test. Maybe Tridge can give it a quick test and if he sees the same problems as you hopefully we can sort them out easily.
Jon22,
Ok, Tridge tested his AUAV-X2 board and saw the same issue you saw. It’s unfortunately very difficult for us to automatically detect the board type because it’s got the same sensors as the PixhawkMini so we’re going to add a new board type that you’ll have to pick manually. This will be included in AC3.5-rc2 which will be out this week.
I’ve loaded AC3.5-rc2, changed board_type to 20 (Thanks Randy for including the flight controller in rc2). Calibrated accelerometers and external compass. I’ve managed to arm the ship indoors and motors work. I’m hoping to test fly tomorrow.
Copter V3.5 rc2. Initially had a few problems and ended up restoring the default settings. Calibrated accelerometers and external compass. Also used the Mission Planner module for calibrating the ESCs. Only short flight, Stabilize felt good, even on default settings. Went into Althold, worked fine but the throttle mid section felt just like Loiter and not as responsive as I’m used to in Althold. Changed back to Stabilize and crashed shortly after (pilot error). It’s a diy quad so hope to have repairs completed by the weekend, will then undertake further testing, but looking good.
Damage on the quad was a bit more extensive than I first thought - terminal. Transferred everything over to a cheapo quad.
Stablize and Althold worked fine. Went into Loiter and pitch became roll and vice versa. I’ve got my Loiter set up in SuperEasy mode. I wasn’t flying very far away from the launch point due to space restriction, so this may have been the problem. Landed. Unticked the SuperEasy mode, re-tried Loiter and all was fine.
Although I’ve got my Drotek XL gps/compass up on a stick my compass offsets are high X -189 Y -355 Z -55. I don’t know whether this would affect EasyMode for Loiter? In Mission Planner the compass offsets are in green and whilst flying both Vibration and EKF Status bar charts were fine, though I didn’t yaw very much due to restricted flying space. If the wind dies down I’ll try it up the fields later on and retry SuperEasy mode in Loiter.
The high compass offsets should be ok. If it’s not OK, the EKF should complain (at least the MP has an EKF health display, I’m unsure about QGC and other ground stations).
Yes, I suspect the super-simple mode issue was because the vehicle was very close to home. I think within 10m it doesn’t do anything. We could make that radius smaller if you like. GPSs are getting better and better…
Loaded V3.5 Rc3. Stabilize, Althold and Loiter worked fine (Didn’t try Super Easy mode)
Ran Autotune separately for Roll, Pitch & Yaw using Loiter mode - drifted a little bit but probably no more than fifty feet, all worked fine, PIDs saved. Flew quite nicely after the Autotune, felt quite tied in.