I have tried everything from
1) Repositioning the GPS
2) Recalibrating (about a billion times)
3) Added ferrite ring
I just don't know what to do at this point but maybe buy another GPS
Its a uBlox GPS
Does this in Alt Hold or anything GPS related,
Stabilize works great.
APM Mini 2.6
uBlox GPS + Mag
Its low in the pic below but tried much higher too, no reposition seems to make a difference.
Classic toilet bowl - https://www.youtube.com/watch?v=y61tZiGiI7Y
Replies
I have a 4 in 1 esc directly below the pixhawk under that top piece of plywood. I sprayed the bottom of that plywood with this, ;et it dry, then went over that with copper foil tape. The offsets are 200 ish 200 ish and 400 ish. Good enough to pass the new 500 offsets threshold. Hard acceleration does not induce yaw from mag interface and the logs indicate a pretty consistent mag field throughout flight. So I guess it seems to work pretty good :]
I recently had a insanely high offset 52,000,000 !!! [ impossible ] after a crash. I could not fix this by recalibrating... tried it a dozen times. Finally did 'reset' under the full param list. Cleared up the issue right away although that did mean starting over settings etc although you can save those first to compare after. Re-installing the FW won't work gotta do a complete reset.
@Randy
Here are the logs, and you will be able to compare, COMPASS_ORIENT looks the same on both (8)
This is before It started flying right
download.hurricanesoftware.com/files/legodrone/logs/2015-01-19 08-2...
This is when it started flying great, after recal/ferrite ring/remove screws
download.hurricanesoftware.com/files/legodrone/logs/2015-01-19 18-1...
Im not sure what the difference is that could have cleared it up.
@Jay, thanks I'll twist the wires too, this is a small quad(lego) so its hard to get away from the battery, I did have it about 2" higher but that did not help, lowered it to add the ferrite ring as the twists made my gps cable shorter, if I can get a longer cable Ill raise it again.
108
141
215
Not sure if thats good or not
No, I don't think those offsets are too high. A total length of >500 is where we have our pre-arm checks set now. So this vector is 278 (i.e. sqrt(108^2+141^2+215^2)).
What's more likely a problem is the COMPASS_ORIENT being set incorrectly, the compass being physically oriented incorrectly or interference from the power wires.
It's difficult to know which without looking at the logs I think.
Fixed, did a recal and removed the screws, but screws were not there before the only other thing I added was the ferrite ring, all fixed up, at least for now,
Thanks guys.
even tho its flying fine now, I do wonder if these values are too high?
Shouldn't they be less than 100?
I am afraid that your four screws are ferrous material and magnetized.
ty, but that was recently added, all flights without them also had it, but I will get rid of those and use black tape to hold my housing