Lots of HUD movement when stationary

Good day everyone!

So without to much jibber jabber here is the story so far:

I’ve been at this for about two weeks now. One week crawling the interweb for forum posts, wiki entries and the like, and another week doing the same thing but also emailing the 3DR support guys. I have now come to ask for the help of a larger population of individuals more skilled than myself.

I am at my wits end. I have the following: Pixhawk. 3DR telem radios. 3DR uBlox GPS/Compass. 3DR Airspeed Sensor. Along with all the peripherals such a the really loud but small buzzer, switch, power distro cable I2C adapter etc…

I have connected everything to the Pixhawk with the exception of the airspeed sensor and my traditional RC gear.

I am using Mission Planner v1.3.7. Arduplane firmware 3.0.3.

I have noticed the following: While stationary - as the topic suggests - the HUD is moving all over the place. My roll, pitch and yaw and varying by degrees of up to 2, sometimes as high as 15 and sometimes as low as 0.2.
This ONLY happens when the GPS gets a fix and the degree of the error seems to be proportional to my GPS hdop.
I made a foil hat to selectively kill the GPS to check this and my theory stands - when the GPS is dead, the IMU data is stable as a house.

Is this normal?
Does the GPS data affect the inertial orientation?

When the sensors are behaving their absolute best I get swings of about 0.2 degrees, is that normal operation?

I have many other questions relating to the Pixhawk in general and specifically the operation in a fixed wing but I feel that this is perhaps the most important thing I would like to investigate.

Any help would be greatly appreciated.

I apologise for both emailing 3DR support and posting here but I wasn’t really getting any answers from the email option.


I have a test bench Pixhawk complete with GPS/Compass, R/C receiver, and a telemetry radio, mounted on an ABS plastic board that sits on my level work bench, and I have seen this kind of HUD display once the GPS gets a 3D lock.
It seems to get more stable as the number of acquired sats goes up and the HDOP goes down, but does just sort of drift around slightly at that point.

Yup, my setup is pretty much the same but I’m using a wooden clipboard… and I don’t have the receiver plugged in to avoid any and all interference.

Well at least I am not on an island with this then! Thanks for the input TCIII.

So why is this the case though? I did not think that the GPS influenced the INS at all. My understanding is that the GPS merely provides the position and ground speed.

Having done a few more tests, I completely concur with your observations, TCIII. It is not only the number of satellites that one can see but the HDOP as well. I can’t confirm which one has more of an influence as I can’t accurately vary the number of satellites as I can the HDOP value.

All this aside, why is this the case? How does GPS influence the INS? More importantly, in the event of a cloudy day, what happens then?

@TCIII, by how much does your pitch, roll and yaw vary in perfect conditions? Am I being crazy expecting better than a swing of 0.2deg on the most perfect Saturday ever?

I really have not made an objective measurement, but I would say a degree or two at the most. Just little wiggles and slight bounces.

@TCIII, do you have any idea how and why the GPS has such an influence on the inertial measurements??

Anyone? Not that it matters at this point, I am simply curious.