Michael,
I've attached two logs and two flight plans from this morning's flying that show a few things I do not understand. Would you please take a look and comment on them?
The flight plans are a short loop around a local flying field. The first item is the resting altitude of approximately 87 feet. Google reports the field elevation as 47 ft MSL. GPS can be inaccurate but this reading is very consistent when the GPS gets 3D lock. Can you say why it might be "always" 40 feet higher than surveyed?
The flight plans call for an altitude of 200 feet and I assume that is AGL, not MSL. The first log shows a flying altitude of 330 ft +/-. Would this be an APM or GPS error? Again, the 330 ft altitude is consistent while the flight plan is set to 200 ft.
The second log shows another problem that comes up less often. The log starts waiting for GPS lock. When it gets 3D, the altitude comes back as 590 feet when the a/c is still at 47 ft. The actual flight was indicating about 800 ft when using the same 200 ft flight plan.
Or, is all this "pilot error"? Am I missing some setup, calibration or procedure error that's giving me these inaccurate readings? As a point, the airspeed has always been reported accurately from the GPS, the a/s sensor is not installed yet.
Thanks for taking a look at this.
Andy
Replies
Had some time for more investigation and here's what I found.
The HUD altitude tape is driven from the baro sensor. And in a typical startup, my baro sensor reports 80ft MSL at my location that's 40 MSL. This is a error constant of +40 feet and that's ok. The altitude tape moves correctly when the airplane is lifted up or moved down. I also moves appropriately when warmed by a finger touch or cooled by a breeze. Other controllers use the BMP085 baro sensor and some suggest enclosing their boards to minimize these effects. My APM is inside the airplane and is mostly shielded from air currents. And so we also now know the GPS altitude data does not seem to be used by the autopilot. Anyone familiar with APM coding, please jump in if that's an incorrect generality!
So there can be some randomness to the altitude due to environment factors affecting the BMP085 baro sensor. These could be reasonably in the range of zero to tens of feet depending mostly on temperature. But environment doesn't explain why the altitude would jump from 80 to 800 feet when the APM is reset or restarted.
There's a button on Planner's Flight Data screen labeled "Set Home Alt". When it was showing 800 feet today, that button would reset the tape to zero on the first click and then go back to 800 on the next click. This is reasonable since my home altitude was set to zero on the Flight Planner screen. Can someone say why this button was created? Does the APM report incorrect altitudes and so was the button added to override an incorrect reading?
Ryan- are you a Planner developer? It would help to know if you have insight to this great program. It _does_ make APM work much easier. Thank you.
1 - Google maps elevations have been reported here as being quite inaccurate in some areas.
2 - GPS altitude is usually less accurate than horizontal position by a factor of 3 to 10. It is not uncommon for the error to be a repeatable bias.
3 - Your waypoint altitudes can be msl or agl, depending on how you set them up. If you are using Planner look for the checkbox that says "absolute altitude". If it is checked you are using msl, otherwise you are using agl.