First Skyfun flight today.
It was a bit tricky because it was windy / gusty and the stock setup is very weak... but all went well.
I switched a couple times from normal to stabilize mode and every time the plane immediately started diving.
I am not sure where to start trouble shooting the issue.
I was replaying the session but can not find the time-slot when I tried stabalize mode.
I assume to analyze the issue I'd need to find that spot. Is there a trick to it?
What else can I do to figure out what the problem is?
Here is what I have:
Skyfun Stock motor, Elevons upgraded to balsa, wings covered with Monocoat
APM 2 with Gps on board
2200 MAH lipo
Dx7s & Spectrum 8100 Receiver
DR3 900 mHZ Radios ( Yes the work now :))
Current Sensor ( Sparfun)
Air speed sensor will be installed tonight...
Nope CG if Fine. In manual mode it flies just fine.
But I am analyzing my sensor data and there is something dead wrong.
VSI shows -1
AS 68 (GS 30)
Pitch angle 25
Measurements in KM/h
That makes no sense thet is no way this plane sinks at 70 km/h pitched up 25 degrees.
check, when APM2 initializes, to keep it steady and in the ballanced position, if plane looks down in initialization, then in stab mode it will try to keep it there. happened to me many many times..
i assume all your banging preflight checks are right...
I manually leveled the SKY Fun to sitting on a even surface. The cord of the wing probably points 1-3degrees upward in this position. Which would I believe be a reasonable flight position.
But I am now wondering if the APM is not properly initializing. I usually wait until the GPS is initialized before I move the plane ( take 5 min here on Oahu)
But I need to check the lights next time to make sure the APM thinks it is done ….
Sorry this took so long, work is keeping me pretty busy….
Having not seen any output from an Ardupilot before I have no reference, but the output ( logfile replay)seems much more nervous than what I observed flying the plane.
I flew 4 batteries 16 min each and tried to switch to stabilize mode several times. Most of the time, the plane would go into a 45-60 Degree dive. If I countered with ruder the plane would still be nervous around the roll axis. Once during these four flights I managed to put the system in stabilize mode and I it worked for about 3-5 minutes as anticipated. The Plane flew stable in one direction responded to control inputs and would level roll and pitch axis as soon as I let the controls go. Then - suddenly the plane went into a steep dive again and I had to switch back to manual. Despite several attempts I was not able to get the plane to behave normal in stabilize mode again L. I have no idea how to explain this behavior. The Ardupilot is mounted with double sided tape to the plane and has no play at all.
At 15:16:34 you can see a example when I switched to stabilize mode. Ardupilot “ thinks” the plane is pitched 24d upward (while it was really flying straight) as a result Ardupilot applies down rudder and the plane goes into a dive….
On the ground, in stabilize mode the elevon action is exactly as expected.
Could it be a problem with vibration?
The prop is balanced but the motor runs a bit uneven.
Bad output from the ESC(bec)?
I attached the log file and would greatly appreciate some feedback
PS: not sure what the difference between T and Rlog, or which one is needed so I am uploading both...
having had a case of "the leans" with a multirotor your problem doesnt sound too different, especially as you were able to fly as expected once with no changes.
perhaps try isolating APM from vibrations with some kind of damping foam, you could test this on the ground watching the artificial horizon and running the motor up through its speed range. if you find a speed that provokes a sudden change in the indicated attitude then thats certainly the problem.
That is an excellent idea, I'll try to find time for a " ground test" tonight.
If any body could take a quick look at my logs and tell me if the out put is "normal' that would be very helpful!
Sure, What do you consider long? 1 km ?
Thank you for checking it out! :)
I'd like to see a log from a "normal" behaving plane maybe Syfun.
The plane is very easy to fly
It does not feel as nervous as I would assume it would feel based on a replay of the logs.
I was able to replicate the behavior on the ground.
Plane is mounted stationary (CAN NOT PITCH ROLL OR YAW)
I start with the plane pitched down at -10 degree
At 21:22:55 I go to full throttle - watch the artificial horizon, it keeps dropping.
At 21:24:00 pitch is at -35 degrees I cut throttle and within 4 sec it is back at -10.
I repeated the same test with the plane mounted at +10 degree and the artificial horizon pitched up to +40 at full throttle.
I improved the vibration isolation and upgraded to 2.34
I did a very short test (to late too loud) and did not observe this behavior. (Horizon was still shaking a bit)
More testing tomorrow! The logs are included.
I broke my motor mount yesterday on takeoff.
I was using a new setup E-Flite bl450 870 KV and a 9/6 prop.
The plane went into a steep climb and lost airspeed.
I chickened out and and pulled the throttle back and landed flat. with the big prop attached that broke the mount.(not the prop)
I suspect the added thrust from this motor caused the step climb.
I will use a washer as a shim to angle trust a degree to 2 down.
Has anybody seen the behavior after a motor upgrade on the Skyfun?