External compass interference?

Hi. I have a custom built hexacopter (1000mm, 5kg dry weight) with Pixhawk (3DR original) and GPS/Compass module (also 3DR). The flight controller is mounted on the PDB. External GPS/Compass is 40cm away from FC and PDB (from motors also). Internal compass is turned off and only external is used.

Compassmot shows 40-50% interferance. When I fly up it loses orientation in several seconds. Logs show compass drift after certain current consumption (40-50 Amps). How can it be possible? Compass is so far away.

Any suggestions?

Best regards,
Roman

1 Like

So, I installed APC 3.4.5 and now it shows me 5%. I had 3.3.3.

Firmware?

We would need a log file of your flight to analyse the issue. I assume you have COMPASS_USE2 set to 0 - is that how your disabling the internal compass?

Thanks, Grant.

“The flight controller is mounted on the PDB”!!!..also,how did you manage to get 40!!cm??

I use only the external compass. Why is it a bad idea to mount it there, besides internal compass interference? It is mounted inside the frame. GPS/Compass is mounted on long carbon stick 40cm long. The whole drone is quite big and it doesnt look weird, like it could sound :slight_smile:

I am getting very strange behavior here. Testing now, will post some logs later.

Because it behaves strangely, I can not let it fly anymore (almost broke my 18" TMotor props several times). So, I am holding it in my hands over my head while it goes to hover state, then I loosen the grip and let it fly (hands on handles, but not pushing). It stabilizes itself, but starts rotating. In PosHold it wants to fly away.

Without the logs real experts wont answer your questions…
FC on PDB can cause all kinds of problems ,not just for compass…try to remove FC few cm up from PDB
Using only one external compass is also unsafe,its time to ask yourself;what is more important,tidy look(all inside frame) or redundancy on almost most important sensor on your craft?
Also,long pole(40cm) for GPS can be prone to vibrations…giving you strange results…
How do you start your arming sequence? holding it in your hands???or you arm it and than pick it up?that metod also can lead to wrong sensor data…
Sorry for all criticism but i just try to figure what’s wrong with obscure data you provided…
And i am no expert for logs,but have years of experience with multiple pixhawk crafts…

First, thank you for your time and answers. I appreciate your help, and criticism, if there is any, is even better!

The Pixhawk is mounted over the PDB PCB on plastic 1 cm spacers. I tested interference with both compasses also, but something strange happens here. I get inconsistent results. Sometimes it shows huge interference, sometimes less than 10%. In theory, the PDB is designed in a way that it should not interfere with the FC. Nearest current path is 6cm away from the FC. Sometimes testing shows that even at full throttle (160Amps) it gets low interference even with both compasses. The problem is the FC behaves inconsistent with every test (firmware and parameters change). I am trying to find some dependence here.

I arm the drone this way: Drone is on the ground in steady state, I connect the battery, wait for initialization ang GPS lock, then I lift it over my head with my hands and my friend arms it. Then he rises the throttle until is hovers, I losen my grip and let it fly where it wants. Sometimes is hovers, sometimes rotates, sometimes tries to fly away. Haven’t caught the reason yet.

The GPS is not vibrating, tested it also.

I will testfly it today and will post the logs here.

Thanks again, and any type of advice is appreciated.

Best regards,
Roman

I tried to fly today. After the drone went up it lost orientation and position and went sideways rotating. several seconds in the air. Please find the log attached.

Also, I was doing compassmot calibration several times and everytime did the same procedure (connect the batteries, wait init, GPS, raising over head, starting compassmote from MissionPlanner, raising throttle slowly up and then down), but got very different results Please see the screenshots. I am flying Pixhawks for 2 years and I think I know all the procedures, but who knows. What can I do wrong here?





And here is the log from test flight.
351.BIN (488.0 KB)

Also, the frame is my custom built Hexa Alta6-style with folding arms. I was flying with it 7 times (6 months ago), everything was ok, except compass interference issues. Mechanics are the same, electronics changed.

Any advice and help is welcomed.

Thanks,
Roman

Hi. I found one big mistake. I connected 2nd motor to 3rd, and 3rd to second. It will explain rotation and flyaways. Will resolder it and test tomorrow.

What about the compassmote results and the varying interference? I tested compassmote several time on my second drone (quad) and it shows under 2% interference with the external compass and under 5% with both turned on.

Why results on my screenshots may vary so much?

Thanks,
Roman

Ok. Drone flies well.

1 Like

Great news that its working well. Good effort on the self diagnosis. Often the hardest problem to find is the one staring you in the face :slight_smile:

Thanks, Grant.

Hi,
here is some more general information which might be helpful:

Check the entries around 30.9.20