I'm currently building an Alien 680 Frame and looking around for ideas on mounting the ESC's, FC etc., I noticed several using xHawks and they mount it inside the frame with the power wires surrounding it underneath the frame.
I've always mounted the Pixhawk et al on top of the frame well above the power wires. How are these guys flying without compass issues with the FC sandwiched inside the frame right above the power wires?
Are they not doing compassmot or something? Their copters seem to fly quite well. It seems the EKF would be doing somersaults.
Replies
Power wires near the compass will definitely cause significant (as in crash or flyaway) interference with the compass either external or internal if you are using flight modes which depend on the compass. This is very easy to test and should be part of your preflight checkouts after the build. Just look at mx (and mx2) with the quad in your shop. Rotate rotors one clockwise and flip them over, then tape down the quad. Run the motors up and look at what the mx and mx2 signals do as you ramp up throttle, also look to see if the heading changes. ANY significant changes (more than10-20 in my experience) to mx or a few degrees in reported heading and you have big problems. mx2 should be your internal compass so you can compare the performance between the two. I agree with your findings that the pixhawk will never actually use mx2 (internal) if the external is present. It sure doesn't hurt to run compassmot at this point, but if at all possible to solve the interference with distance, that is the preferred method.
Are you sure these guys are running in modes which require a compass? (Loiter, Guided, Auto)? It seems to me if they are using these modes, they'd also have a GPS, and those often (3DR) have a compass included as well.
Bottom line: If the parameter is set to external compass, the internal compass is NEVER used.
That is where the confusion on my end arose. A few months back I had a flyaway with the test quad and posted in the 3.3 beta thread. Randy suggested moving the compass further away from the power wire. I took that as moving the internal compass because the external is a good 6-8" above the power wires, and the case is all plastic including the screws/nuts. So I installed a vibration dampener which did raise the FC up higher; not a total waste because vibration levels are now near non-existent.
The flyway is still a mystery, but Randy did come back and mention the AUAV-X2 FC had a hardware bug to be fixed via software in a later 3.3 rc. I ended up grounding the quad until it was fixed because there was no logical explanation for the flyaway, at least from my perspective. Since the fix, EKF monitors and all other data show no problems.
Regardless, I had thought when doing compass calibration it included the internal compass in the process. Wow, did I have this totally misunderstood.......after over 1 1/2 years. I thought it was important to keep the Pixhawk far away from power wires, yet continually saw setups with it buried inside the frame and wondered how they weren't causing flight issues.
Alrighty then, inside the frame it goes. Thanks all. BTW, this FC is a CUAV Pixhack for something different to try out.
There is no compass in the Pixhawk. Compass is with the GPS. As long as the GPS puck is on top you should be fine. Not sur e what you saw; maybe you can share links for us to see what you are talking about. You can also use copper foil beneath the GPS puck to shield it from the magnetic interference from the ESC wiring. If you check out the new Solo, everything is in the body too and there is also a piece of copper foil under the GPS/compass.
@Vinnie,
According to the ArduPilot Wiki the Pixhawk has a built-in ST Micro 14 bit accelerometer/compass.
The APM2.6 does not have the built-in compass.
Regards,
TCIII AVD
Maybe, I never use the internal one even if there is one. I am sure it was mention somewhere it is not recommended to use the internal one. 3DR GPS already comes with the external compass to use so you would not be using the internal one.
I'd nearly always recommend using an external compass instead of trying to rely on the internal one. Nearby power wires definitely sounds bad to me!
Hi Randy,
I was debugging a compass inconsistent issue on Plane 3.5, and eventually got it working after a 3rd full compass calibration (on pixhawk).
However, before that, I just had to disable the internal compass or I was not able to arm (the external compass felt more trustworthy) (COMPASS1_USE1 0)
But then I came to wonder, is there even a good reason to use the internal compass that is so close to so many wires and sources of interference? Any reason to not just disable that compass forever?
Thanks.