Bad AHRS fault, is the pixhawk faulty?

Attached is the first .bin file on the SD card there is a second .bin file which is too big to upload, I hope you can make sence of that because i can’t. maybe it will give a clue whats good and what’s not.
00000001.BIN (81.3 KB)

What version of the code are you using? Are you testing outside or in the house. Compass offsets look fine but Nsat’s and hdop are bad. At this point what is the problem with the vehicle. Only works in manual or what is it doing and not doing.

I think that .bin refers to earlier this year January I think and I am not sure what version I was using then. I would have been testing inside. At this point I have the pixhawk stuck down on a piece of plastic with the GPS mounted above it on one of the standard mounting things, its about 100mm (4") above the pixhawk.

I have just now had the most promising results so far. On compass calibration I unchecked the boxes in compass #1 and checked both boxes in compass #2. Calibrated and got a successful calibration.
Now when pointing north I am within a few degs and when I point east, south and west it now points in those directions. Bad AHRS have gone but I now have error Bad Compass Health.
The HDOP is 1.5 with 10 Sats, so if I calibrate outside that I hope will get better.
What does that tell us? is the GPG module faulty or is it the pixhawk??

I can’t calibrate any better and still have “Bad Compass Health” The FC drifts away from pointing north after a few minutes.
How can I determine what is faulty please, the FC or GPS module.

Hi,
Could you tell me where to find these parameters i need to change please. I think i then turn each one off or on and recalibrate and I will hopefully see a good result from one of them.
I am thinking of ordering a new FC and Gps module.
For what I need (gps controlled boat rudder) what would you choose?

Honestly I would get a cube + HERE GPS or a mRo Pixhawk 2.4.6 “cool edition” + mRo GPS. I’m not a big fan of the cheaper clones. People do have good luck with the clones most of the time but it’s not something I even consider. Plus it supports the organization which is important to me

@colindo

Try going to the initial setup --> compass tab in mission planner. that is where you can change which compasses ardupilot uses for ahrs. Compass 1 is the external compass in the GPS, Compass 2 is the compass in the pixhawk.

By default ardupilot uses compass 1, check in mission planner as above to see what yours is set to (it may not be the default).

Then I would in mp select compass 2 internal as your only compass, and then run the compass calibration in mp, and see what happens; ie whether after calibration your internal compass points in the right direction when you move the pixhawk. Compare the direction mp thinks the pixhawk is pointing to the direction an ordinary compass thinks it’s pointing.

Then try the same with compass 1 (the external one), after you run compass calibration again

After this you should know whether each compass is good or bad. Since AHRS relies on both GPS + compass data, if both compasses are good, logically it may mean that your GPS is bad.

Many many many times problems occur because the two compasses conflict and ardupilot doesn’t know which to use. By testing them one at a time you will hopefully be able to isolate what your problem is.

And lastly remember that magnetic interference is also a frequent problem so test outside away from large metal objects.

Thank you so much. I can only calibrate compass 2 and it points in the correct direction but gives error “Bad Compass Health” and I can’t ARM. this I expect is because the external compass is faulty. I have two replacements coming from banggood, so hopefully they will solve all my problems.

Do I need the battery connected to calibrate?
I thought I would try calibrating using The Wizard and I can’t get past “Missing BATT_CAPACITY param” something is wrong then I get “Failed to set battery size, check input”.
Could this be the problem??
When I see calibrating demo’s on youtube they don’t show any battery attached.

I’m not sure but try putting in the BATT_CAPACITY in. What ever mAh it is. I have never used the wizard

Have done that, still no success.

I believe the batt capacity param is part if the batt fail safe setup. It wouldn’t effect the compass calibration, but you will need to fix it for the fc to arm. In fact setting up the pixhawk largely consists of clearing up each of the many warning messages you’ll see (at least on my pixhawk it was like that). So, at least you’re making progress

I wouldn’t use the Wizard either. What version of Ardurover are you on?

I am waiting for new GPS module to arrive.

Hi, I had the same problem with neo 6m compass.

It can be many things and replacing it may not solve it. Here are the probabilities

  1. wrong orientation : sometimes the chip is 45 degrees angled on yaw, or 180 degrees on roll or even both. And you may be physically installing the module in a wrong way
  2. i2c faults: i2c may have inconsistencies with compasses. Back up your parameters to computer. Reset all of your parameters in FC. Wait 5 seconds. Plug the power off. Wait 12 seconds. Power on.

Try to calibrate now. Then you can restore the params again

  1. same sensors used in neo 8m and pixhawk causing issues because its on i2c channel. I have no idea if this claim is true but there are some reports.

  2. no Voltage regulator in the module causing the issue (3.3V 5V differences for apm/pixhawk)

  3. if your gps works (df13 connections are different for apm and pixhawk) but no compass on hardware id it’s probably dead.

I didn’t sleep tonight and googled the similar issue.

It has got nothing to the with EMF at this point.

Some suggests removing all the cables except gps/compass (without gps connection compass can’t be feed). Usb connection to your MP and try calibration now.

I hope you can solve this issue.

Hi,

Have you seen this https://github.com/iNavFlight/inav/issues/2619
These guys are cutting a chip out of the module and replacing it with a couple of wires.
They are using Inav. I don’t know enough to say if it would be a suitable mod for pixhawk and Mission Planner.

Well it appears to be driver issue im their case.

I think the reason is mostly the orientation or the pixhawk. Have you tried the re orientation?
Also check this topic

I ve soled my issue.

roll180 was the fix for me!

you may try roll 180i yaw 180 in order

don’t forget that each time you change orient, you must re calibrate accels then compass.

No compass variance error no stuck at 33 percent! nice!

I knew it wasn’t a HW error!

I have solved my problems also.
I have taken a gps module off an old drone that I know is good and it calibrates.
So the problem is with the cheap crappy Chinese gps modules.
Grrrrrrrr!!!

1 Like

Actually they have no issues. The problem stems from orientation. They don’t print what side is the mag is soldered. So we don’t change the orientation. It’s mostly 180 roll or 180 yaw.