Phantom takes off then becomes more erratic in flight

[aprilmedia]

Hi,

Had phantom for a few months now. Have recently Changed gimbals as the Tarot one we bought was faulty and put the old one back on (XAerocraft). Have also put on a new carbon fibre landing gear. (they aren't straight down but at 30 deg splayed out)
The compass is attached to that so now at an angle.

Now the Phantom is out of control. I have disconnected all gimbals etc, Have checked compass settings on software and all fine and the Phantom calibrates first time every time doing the hor / ver dance.

It takes off and hovers fine and flys OK, then after30 seconds it starts to become more and more out of control, although under control of the transmitter, but requiring severe controller movements to keep the thing even still and eventually impossible to keep airborne and crashes. The same happens in Attitude or GPS mode.

I'm thinking a compass issue maybe? Would a faulty compass cause this? Must it be fixed so its always in 180 / 90 deg position like on the stock phantom legs?

Any ideas welcomed.

Many Thanks

John

 

[Gizmo3000]

I something similar happen to me once, it would take off but then at some point get all weeble-wobble on me.
Upon landing, I discovered that one of my motors was _really_ hot. guessing a bearing was having issues. (I then replaced it).
But something to check after flying a bit.

 

[PhantomFan]

Good call. I bet there's filth inside one or more of the motors from a prior crash that is impacting RPM's in an irregular or haphazard way causing the OP's symptoms. Excellent diagnosis, Gizmo!

PF

 

[miskatonic]

You could have a short in one of your motor wires. The same thing happened to me after I took the Phantom apart and then put it back together. I accidentally crimped one of the wires with one of the mid arm screws. This caused a wobbly Phantom and eventually a crash.

 

[Driffill]

The compass is attached to that so now at an angle.

Must it be fixed so its always in 180 / 90 deg position like on the stock phantom legs?

Yes, must be in that exact position. On an angle, it won't work properly!

 

[aprilmedia]

Thanks all for your ideas. Have run it hard on a table holding it down for a few minutes and none of the motors get hot. none of them are binding and have taken it apart and nothing damaged inside or wires crimped.

What does seem to happen though is the "throttle" seems to 'stick' in GPS and ATT mode, so if I rev it up in any direction and then let go, the revs don't immediately drop off. but almost keep on going. not sure if this is correct? does anyone know?
I know it will need to maintain some down force revs or will just drop out of the sky! but seems odd that this also happening on the L/R throttles. This doesn't seem to happen every time either (in my table top test!) but intermittently.

If this isn't right than this would explain why it becomes harder to keep under control over the minute its airborne as trying to make larger corrections on the controller to keep it in check as it keeps over accelerating and sticking in general directions.

Ref Compass not at 90/180 degrees, surely by doing the calibration dance tells the Phantom which way the compass is facing? Or maybe not?! And if so, then this could be causing an issue, although it will hover and manoeuvre for a while and seems to know where it is until gradually becomes more out of control !

Again, Ideas most appreciated! Thanks

 

[miskatonic]

Did you do the advanced calibration?

 

[EMCSQUAR]

My money is on the compass location. I had similar issue when putting on different landing gear, it felt like it was in att mode & was erratic. I made a vert bracket and mounted compass in similar to OE position, did GPS dance & advance cal & it's spot on now.

 

[aprilmedia]

Thanks - I'll try that and advise of the outcome.

 

[OI Photography]

Yes, the compass DOES need to be in the same position and orientation as the stock mounting. Having it level and square to the ground is most important but the distance and orientation in relation to the NAZA is important too. Failure to keep it in the right location leads to the "Toilet Bowl Effect", which usually results in gradually more erratic behavior ending in a wide sweeping descending turn or roll and ultimately a crash.

If you have aftermarket landing gear (like I do), you'll need to find a way to get the compass mounted properly again...no amount of calibration of the compass or IMU will overcome that. Here's a thread discussing the issue, and with pictures of the solution I'm using: http://www.rcgroups.com/forums/showthread.php?t=2024859

 

[aprilmedia]

Yes, the compass DOES need to be in the same position and orientation as the stock mounting. Having it level and square to the ground is most important but the distance and orientation in relation to the NAZA is important too. Failure to keep it in the right location leads to the "Toilet Bowl Effect", which usually results in gradually more erratic behavior ending in a wide sweeping descending turn or roll and ultimately a crash.

Thank you so much for posting this 'OI Photography'! This is exactly what's started happening after we changed the landing gear for a 3rd party one. Even after squaring up the compass, the same problem occurring, so we are now about to put the old legs back on and should sort it. Amazing how the 3rd party leg manufacturers make no mention of this at all! and their photos don't show the compass in them!!
Also your fix by using the old leg is a fantastic idea.

I'm confident this is the issue and will post back once tested.

 

[aprilmedia]

Just reporting back, and yes that did fix it! The compass MUST remain in the position and orientation as per original fixing on legs. Thanks all for your help.