Interestingly, the copter does try to yaw, it seems that Compass1 has a significant drift on the X axis.
Could you try changing to Compass2 ?
Hey Andras.
IS it possible that the Compass1 yaw (which should be the external compass) is yawing due to current demands as the craft gets airborne. I have not done a compassmot so I am wondering it thatās the case.
Not sure how to graph that out.
Not sure if you changed anything but both active compasses are still Internal.
1 Like
Damn I keep telling it to use the uavcan compass
Got the boot delay set?
I assume you mean this setting. BRD_BOOT_DELAY,0
IF so noā¦but if your calling it outā¦me thinks I best set it.
Hi Ricky, I donāt see where the external compass is installed. It gives me the impression that it is inside the frame, if so try to place it on a mast, well away from the esc and batteries.
In any case, everything points to a problem of calibration or mechanics of the compass.
A good test to rule out the malfunction of the compasses is to deactivate them completely and test in āstabilizeā mode.
I imagine that your radio control is not with any imbalance in the Yaw signal, it is still good to reset it to make sure
Hey there.
So first @xfacta I set the bootdelay to 1000ā¦thatā's I think about 1 second.
Also this is what Mission planner says regarding the compass.
Note that the graphic shows compass 2 enabledā¦is now disabled and compass 3 has the external unchecked.
I have no issues with the radio. Same results with a different radio.
Patricioā¦how can you see which is internal versus external.
I have added the boot delay and see three compasses. Now I admit I canāt tell how you know in mission planner which is the external versus the internalā¦ Which compass number is assigned to what id.
IS devid 0 compass 1 and dev 1 compass 2 and so on. IF so I have told it to use an internal compass and not the external.
OK probably set that boot delay to 3000 - just to be sure (more reliable). Disable Compasses 1 and 2, enable Compass 3 - get outside, get a good GPS 3D fix then do the compass cal dance.
Thanks Shawnā¦So I am right dev_id is compass 1, dev_id2 is compass 2 and id3 is 3ā¦so I have set the wrong compass as the external compass. Will make that change tomorrow.
Will also up the delay time.
Still doesnāt explain the yawā¦but its possibly the cure for one of the two issues.
I am working on a mod to my compass mot jig to see about ensuring the motors are all square and level.
Wa hooooooā¦
Yeah, see that compass_dev_id3 is UAVCAN - most likely thatās the one you want to be relying on.
The other 2 compasses will be internal.
The compass_dev_id1,2,3 should map to Compass or Compass1 or Compass2 in parameters, or to Compass#1 , #2 , #3 in the mission planner settings page you have shown.
Mission planner assumes as internal the devices connected to the internal SPI bus. therefore the decision of which to put as PRIMARY is important, since it must be the one that has the best performance and, clearly, it must be the one that is most free from interference. The external Compass installed away from the power cables and batteries should therefore be external.
I suggest disabling the internal compass and performing the calibrations only with the external one and, once adequate performance is obtained, adding another internal one, which in my opinion is not necessary, except for a possible backup? I donāt know if it works properly. there is a great discussion about it.
Very important that the external compass, working alone, (only the main compass) is well oriented in relation to the orientation of the frame (the front to the front and thus with the correction of orientation to zero degrees or that corresponds)
Hi Ricky, after reviewing your log, I conclude that compasses, despite being poorly programmed, have nothing to do with the yaw problem. That if I observe anomalous performance in the ATT, role, pitch and yaw graphics.
I suspect that the horizontal level and / or alignment of the flight controller is out of tune with the frame. Or, it may be that the motors or some mechanical condition is sufficiently misaligned that the inertial measurement unit (IMU) cannot correct. It would be necessary to review the programming of the IMUās to verify that their closed loop is working well.
Hi @PAPO
Yes I agree that the compass and the yaw are unrelated. However I have been having issuesā¦Starting to understand now why and what needs to be done to resolve. Tomorrow I will get the bird outside and let it sit for a bit to warm and ensure there is a good 3d fix. Then I will do the compass calibration on the external GPS/compass.
After that I will get a compass mot done.
I have a jig that holds the quad in place for the calibration and am just making a few mods to it to help ensure that all the arms and motors are square. Itās been suggested that perhaps a motor is tilted resulting in the yaw. IF I look at the logging I can see the motors 1 and 2 are pegged or nearly pegged which in my mind is telling me that the craft is trying to compensate for the CCW yaw but is being unsuccessful.
So I will careful review the frame and do my best to rule that out.
I will also redo the Acelerometer calibration however I donāt see that as a problem. But not confident enough to not redo it.
I have a small hand held level I am going to use to get a lot more precise about it.
As for the IMU programmingā¦thats way above my understanding.
Before calibrating compasses, I suggest doing a good accelerometer calibration,
Good luck
I donāt know if this has been covered, but based on your 08-05 flight I found something that might be limiting the aircaft. Donāt know if this is causing the Yaw issue - but who knows.
You seem to have the following parameter:
MOT_BAT_CURR_MAX = 50
However, during your flight you regular exceed 50A and bounce down in amperage (I am assuming this is automatically done because of the aforementioned parameter). Maybe try setting this parameter to 100A or disabling it for now?
To avoid compass confusion, I recommend to update to 4.0.3rc3 and the latest MP beta, you got a bunch of fixes for compass priority and a nice interface for setting the priority. Especially for CAN compasses.
Hey good morning everyone and thank yo all for your input.
This morning I did an Accel calibration using hard surfaces to ensure it was actually on itās side or its nose in a properly square state. Check @PAPO
I took it outsideā¦nice morning here. Let it sit for a few minutes. Got a good sat lock and then did a compass calibration on all three compasses. I figure I might as well have offsets for them all.
I then disabled all but the external and did that compass calibration again to be sure.
I see the hud follows true direction. I checked using a compass on my phone. While not super accurate. They both reported within a couple degrees of each other.
So calibration Check @xfacta
I updated the Boot delay to 3000 per Shawns ( @xfacta )
I will update the MOT_BAT_CURR_MAX setting becuase your right the machines theoretical max amps is 20x4=80 plus any overheadā¦so around 85 amps.not 50. @manavgandhi17
@Eosbandi I updated to the 4.0.4 betaā¦and the beta MPā¦wow where has that been all my life. Made compass work so much easier. I redid the compass calibration and while it seemed to take longerā¦just me I think. It was so much easier to set up the right compass. Thanks man.
Next I will finish the mods to my jig and then go over all the motors and arms to ensure they are as square and level as possible. Then back to the field for another test.
I imagine everyone is wondering why a jig. Itās because the arms on the drone fold up as itās so largeā¦so making sure the motors and arms are true is a little harder.
Ok @OldGazer
I trapped the quad in my jig and using a long shop ruler I checked for level across the motors to ensure it was good. While I cant say if any where off, as I had twisted each to make sure they where loose. They are all level now. I shone a light from behind to see if there was any gap under the rulerā¦nadaā¦zip
I also loosened all the bolts that secured the arms, the locking mechanism and pivot components. I have them all clamped level and square. I am leaving them over night clamped down so it all relaxes and ensures they are all square before I tighten them up again.
Then I will double check the motors to make sure nothing went out of whack.
Once tightened up I will do the compassmot.
Then off to the field to try again
If thereās still a yaw issue then itās got to be frame twist under load or something like that.