Loss of height when in RTL resulting in crash


I was wondering if someone could help me diagnose a crash I had on the weekend.
Aircraft: Bixler 1.1
Flight controller: APM 2.5
GPS: MediaTek MT3329

The aircraft flew well in manual and stabilise, but when I switched to RTL mode it rapidly lost altitude resulting in a crash. The whole flight only lasted a few seconds.

The only thing that I can see that looks weird is that GPS.Alt indicates that the aircraft seemed to think it was much higher than it actually was. I.e. GPS.Alt indicates that the crash site was significantly higher than the launch site, which is not the case. GPS.RelAlt seems more-or-less correct to me.

I am not sure if I am on the right track or where I should go from here. I was wondering if any of the experts out there could have a look over the below dataflash log for me.

Thanks in advance,

Is that the right log. It shows that the copter did a lot of swaying back and fourth like bad PID tuning.

Also the PARM part of the log is missing. Don’t know what logging you have set or what PID values you have.

Try another test flight with it in ALT-Hold and see if stays in the air.



I can confirm that the file in the DropBox link is the correct log. It is for a plane though - not a copter.

What made you think it was for a copter? Is the file corrupted in some way?


I forgot to mention - when I reply the log it tells me I have bad compass health soon before the problems start. I am not sure why it thinks my compass is not healthy but I suspect it is because of poor layout of the wires to/from the power supply.

Could poor compass health cause the dramatic loss of altitude that I experienced?

I flew again this weekend with the same result.
The aircraft was behaving well on manual and stabilise modes, but when I switched to Loiter and RTL (i.e. GPS dependant modes) it dived towards the ground. This time the crash wrote off my airframe.

I have looked over the logs but I can’t see anything that I would expect could cause this behaviour.

I have posted the dataflash log from my second flight here:
dropbox.com/s/doxtcvw3jtzpq … 9.log?dl=0

If anyone would like to have a go at diagnosing what went wrong I would really appreciate it as I am stuck.

I have the same issue on my plane…

well almost, the plane only looses height in the initial turn when going to rtl, from there on its fine…

its sad you lost your plane because of it. I’m very very new to autopilots, so the only mode I feel completely comfortable in, is manual mode and I grind my teeth in rtl looking at the plane turning without my inputs.

I’m however thinking maybe you have an RC timer or HK v2.5 copy of the apm and you have an open board. with some sunlight hitting your barometer.

Hoppy 1977,
I have the same plane ( or I should say HAD) with almost to same deadly results.

1st plane when straight in going from RTL to Auto. It was an immediate level flight to a vertical crash. Demolished
2nd new plane a little less radical crash, but not salvageable. This time I was in manual then switch to RTL and at this time it banked towards HOME but when in to the ground.
$500.00 down the drain, that’s just in the plane.



Thanks for the reply.

I am using a genuine 3DR APM2.5, with the original case (including the foam over the barometer), all of which is tucked away inside the Bixler fuselage, so I don’t think that is an issue.


Its good to see others are having similar problems :slight_smile:

What GPS are you guys using? I am using an old MediaTek MT3329 (from 3DR) which I had used (successfully) in the past.


I looked at the log from your second flight. I noticed at the end, the pitch demand was positive, but that the plane responded by pitching down. The plane should have responded to a positive pitch demand with positive pitch.

Roll seems to be responding correctly, but the commands to the elevator out of the autopilot may be reversed.

Hi David,

Thanks very much for taking the time to have a look at my logs.

I was really interested to hear about the pitch discrepancy you found. I had it in my head there was a problem with the altitude from the GPS - I may have been going down the wrong path.

However, if the commands to the elevator out of the autopilot were reversed, wouldn’t it have been un-flyable in stabilise mode? I.e. stabilise mode would have been effected by the same problem. (In both flights the aircraft flew acceptably in stabilise mode before the crash.)

Is my understanding correct? If so, could something else have been causing the pitch discrepancy?