Discussions

Toilet Bowl Effect

Posted by Terry Brooks on September 5, 2013 at 9:14am

I need some advice on where to start twaeking the PID's to get rid of the toilet bowl effect I'm experiencing

 

Arduflyer 2.5, flies lovely and smooth just how I like it, spider style quadcopter frame 3.01 firmware

In both LOITER and RTH In a mild wind (approx 7mph) the quad starts getting into the toilet bowl and continues to get worse if left, leave it too long and it gets really wild in RTH

 

Anyone know what PID's to change to start with ?

 

Thanks in advance

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –
Follow

Replies

  • Terry Brooks September 8, 2013 at 11:50am

    OK

    I have now re-wired the quad, re-calibrated the compass and ran the compassmot

    Largest offset was 138 so within parameters

    Compassmot was on first test 10%, second test was 12%

    That's with all the FPV gear as well as the GoPro all switched on and running

     

    Can't test for TBE now as the wind has dropped and it's turned dark here now

     

    Fingers crossed it will eliminate the toilet bowl effect in stronger winds than 8mph, we'll see

     

    So thanks to those with balanced non patronising/insulting advice

  • Warren Purnell September 8, 2013 at 3:33am

    Here goes......

    I'm flying n MR4 by Bask industries with a APM2.5 Mega (has everything on-board). I was never able to really tune any of the automatic modes.

    I was able to kind of get it to fly but not very well.

    So with this Board I got the Compassmot down to mid 20's and the Compass Offset on the Y Axis was up at 125. Others were about 60's or so. Vibration though was well within range (another story getting there with that too). All parameters were acceptable. Having reached this stage I could tune the flight parameters and actually see the effects and the Quad wasn't so sensitive to PID changes (everything was wound back) nor as sloppy in the air. So it became quite tuneable.

    Then I went and added a Metal Gear servo to the underslung camera mount. Nothing else was changed. The Compass Offset P went from 125 to 156. Out of the acceptable range of -150 to 150. Bring in the Toilet Bowling. It was also changing its orientation in an anti-clockwise direction while flying in any Mode.

    Solution:

    I swapped out the Board to an APM 2.5 with a seperate uBlox compass. The compass was mounted only about 3/4" above the board in top of the canopy. Also while I was at it I put an extra layer (already had 1) of Aluminium Foil Tape under the Board on the Quad's top deck. Due to a different Telemetry Port on the swapped out board I had to do calibrations with a USB cable connected to my laptop. The proximity of the laptop or my limited freedom of movement may have affected the Compass setup so had to fine tune slightly with Chanel 6.

    I ran all the set up routines recommended.

    Live Compass gave me Compass Offset P now of X:-38  Y:3  Z:-26

    Compassmot came in at 12% (run 3 times with same result). Anything up to 30% is good.

    Now what I will say that with Firmware 3+ DO NOT set Declination off the Web (unless prepared to tune in with Ch 6), using Logs or Auto-Dec. I had Auto Dec set and the Quad just gradually got worse to where it was eventually spiraling at warp speed while losing altitude.

    So all I can say is do not overlook any of the setup routines. Each one is important!!!

    Compass (Max -150 to 150)

    Compassmot < 30% (through the Terminal Window) 

    Accel Calibration

    Vibration (switch on Log Bitmask to Default+IMU then graph results after flying)

    All of this is in the Manual but sometimes we don't understand the importance of some things as we haven't got a full understanding of the interactions of settings and parameters etc. So don't even both trying to tune the flying parameters until the rest is sorted.

    Hope this helps & sorry it was so long.

  • Cam Donaghey September 7, 2013 at 8:33am

    Terry, I understand the frustrations that you are having, but I think people need to understand that APM's are not a simple plug in and off you go device like a NAZA, they are for hobbyists who are prepared to spend the time to get it right, but once you do get it right you end up with, imo, a superior product to NAZA for a lot lower cost.

    I actually feel a little sorry for Wessie, he was trying to help you out but it sounds as though you are not willing to take on his suggestions but rather curse at the product.

    If I may suggest something, buy a $3 external magnetometer, mount it up and away from all other electronic components on your quad and you should see the performance significantly improve.

    Best of luck in resolving the issues you are having.

  • mP1 September 6, 2013 at 8:31am

    Can someone tell me why nobody talks about shielding  when it comes to keeping the compass etc blocked from those nasty EMI from other devices onboard the copter ?

  • AlexP September 5, 2013 at 6:29pm

    I never had toilet bowling before doing compassmot. After compassmot, it does... does that make sense? Compassmot yields close to 60%, so I'm going to order the external compass.

  • Terry Brooks September 5, 2013 at 11:40am

    Just to update

    I tried adjusting the LOITER PID from 1.000 to 2.0000 but it had no noticable effect

    When the wind drops the quad loiters and rth perfectly

This reply was deleted.

Activity