I recently destroyed an oiplan in a crash.
Ordered a new one and built it along with a new compass.
This time I have connected the compass via the i2c cable.
Trouble is it is very slow to show changes in direction in the mission planner.
The heads up compass display takes forever to show the new heading.
In the CLI compass test it is very fast however in the HUD it takes forever to settle at the new heading.
I have another / friends oilpan with the compass soldered to it and is is very quick to update the HUD compass value.
Is there something different needed if connected via the i2c connector?
some diagnostic suggestions:
take some closeup photos of the cabling and post on this tread
take a photo to show how the magnetometer is oriented relative to the APM and post on this tread
unplug the magnetometer and boot up again, is it faster?
Try setting "Use GDI+" in the Mission Planner settings, see if it runs faster with/without that set
Hope this photo shows the orientation.
When the board is flipped over and screwed down it should be pins forward components down.
Like I said in the CLI compass test it is very quick and shows new headings straight away.
But when looking at the status page on the flight data screen it takes 10 or 20 seconds before the yaw value changes to reflect the new orientation.
It's like there is some dampener affecting the yaw.
ok, how about the other suggestions?
This investigation started when my quad started to spin with no yaw input.
Just now I had a look at the raw sensor values.
If I just look at the gyro data then the Z gyro appears not to be working correctly.
Maybe this would cause the compass display to be slow but it certainly looks like the Z gyro range is very small compared to the other oilpan that I am testing against.
Z should be showing gravity. What values are you getting?
Power the unit off and unplug the magnetometer. Does it work as expected?
Ops, sorry, not the gyro, the Z accelerometer, that is. What is the Z gyro doing? Maybe we should focus on one issue at a time. Take a look at my first post, look at those suggestions, what do you observe?
If you run the CLI test, do you also see very little Z gyro movement if you orient the craft such that the Z axis is horizontal/perpendicular with the force of gravity, then rotate the craft around the Z axis / yaw axis?
Try doing this without any other sensors attached (GPS, Mag, current, etc.) Also, if you get less than expected performance, try gently flexing the IMU near the gyros... it may be that one of the gyro leads is broken/lose, and a reflow often fixes this.
Can you post some gyro output, showing yaw of the craft while it is "flat" and while it is pitched forward 90 degrees (and rotating around yaw)
Here is a screen capture of the CLI output.
I was shaking the board about and twisting about the z axis while this was running.
I presume Z gyro is the 3rd column and as you can see it does not move much compared to the first two columns.
I tried pressing the IMU board around when the z gyro is but it appeared to make no difference.
Thanks for all your help.
I would just like somebody to be able to confirm that the z gyro may be damping the compass reading.
forgot to attach the screenshot
I'd recommend an eeprom reset, see how the numbers look afterwards.
Tried resetting and erasing but to no avail.
Thanks for your help