ArduPilot 2.28 - twitching aileron

[GoogleCodeExporter]

The ardupilot is flashed and programmed using MP1.1.34  - "reset" done.

1.Manual mode gives (naturally) smooth aileron operation
2.Stabilized mode, - compensates correctly for APM roll.-smooth and nice.

However: in stabilised mode, I can move the stick to one way, and roll nicely, 
but at the end of the other way, the aileron goes absolutely bananas, jumping 
around randomly, while I hold the stick steady.

I cannot find any reason for that - channels min and max pulsewidth is 
correctly set up.

Original issue reported on code.google.com by andre.kj...@gmail.com on 10 Feb 2012 at 12:30

[GoogleCodeExporter]

please see this video - observe the OSD for mode and AHI information. 
http://youtu.be/0NVJV3AlLQw

Original comment by andre.kj...@gmail.com on 10 Feb 2012 at 12:40

[GoogleCodeExporter]

- This happens only when throttle is all the way down !
*IN* CH1 is erratic then 

Original comment by andre.kj...@gmail.com on 10 Feb 2012 at 12:50

[GoogleCodeExporter]

Solved by moving the throttlerange up some % on transmitter  it's now 1129-1927 
(vs default 1100-1900)

no more glitching on IN1 when IN3 *and* IN1 is low.
This may be a capture problem ?

Original comment by andre.kj...@gmail.com on 10 Feb 2012 at 1:25

[GoogleCodeExporter]

Andrejs,
    Could you provide some dataflash or tlogs logs?  I'd like to see if it's an issue with the radio channel read or the output to the servo.

Original comment by rmackay...@gmail.com on 15 Apr 2012 at 2:20

• Changed state: Accepted

[GoogleCodeExporter]

Sorry, too long time ago the logs are deleted.
anyway - it was not much to see, values in Mission Planner proved there were 
*no* confusion about the incoming PWM, they were steady, the out-PWM was steady 
too, which led me to believe that the problem was caused by my PWM and/or 
channels, but rather the PWM-handling microcontroller.
-I had channel 1,2,3,4,8 connected. 
If you have the same problem - just try to change input signal for some 
channels, see if the behavior changes.

Original comment by andre.kj...@gmail.com on 15 Apr 2012 at 8:04

[GoogleCodeExporter]

Andre,
     can you tell me more about your hardware?  Is it an APM1 and can I ask when you bought it?  At some point about 6 months ago we updated the standard 328 firmware shipped with the APM1.  The latest version can be updated using the instructions on this wiki page if you have an atmel programmer (which many people don't have :-( ).
    http://code.google.com/p/ardupilot-mega/wiki/Encoder

Original comment by rmackay...@gmail.com on 15 Apr 2012 at 8:50

[GoogleCodeExporter]

APM1 , Purchased 11/28/2011 PCB version  1.4

First I thought you asked (comment4) because you had the same problem.
now It seems you do not have the problem but try to help me ? (thanks)
at least you made me suddenly  aware of that when lost signal, I may still have 
a shaking elevator (PWM is then <950uS) - but it flies well - just not very 
good/efficient if it jumps around.
Well, I have programmer that can be used, I guess there's a chance my firmware 
may be old, and you will suggest I reflash it ? 

Original comment by andre.kj...@gmail.com on 15 Apr 2012 at 9:33

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

in addition to comment above: 
I've found ArduPPM_v0.9.87_Arduplane-APMv1.4.hex 
and ArduPPM_v0.9.87_Arducopter-APMv1.4.hex 

Why are there two versions ? yes, arduplane have 8channel support, why would 
arducopter be any different ?

Original comment by andre.kj...@gmail.com on 15 Apr 2012 at 9:53

[GoogleCodeExporter]

Andre,

   That's not very old so I would have thought that it has the latest pwm firmware but I'm really not sure.

   The latest pwm firware can be found in the downloads area.  Have a look for, "ArduPPM_v0.9.87_Arduplane-APMv1.4.hex".  Note the "Arduplane" in the name (there is an arducopter version that doesn't have the channel 8 pass through - you don't want that one i think).
        http://code.google.com/p/ardupilot-mega/downloads/list

   Yes, I'm very interested to hear if this resolves your problem.  My guess is that you're right that it's the encoder..and it's probably something to do with the very short pwm/ppm signals when your channels are at zero (i.e throttle down, rudder fully left).

Original comment by rmackay...@gmail.com on 15 Apr 2012 at 9:56

[GoogleCodeExporter]

For copters there's no need for the channel 8 override because a human has no 
chance of reacting quickly enough to keep it in the air.  Not having the 
override frees up another channel for some other use (although we don't make 
use of it yet..it's there for the future).

Original comment by rmackay...@gmail.com on 15 Apr 2012 at 9:59

[GoogleCodeExporter]

not sure if I can readout the current firmware to a .hex using avrdude, then do 
a compare, to know if there is any version difference.  do you know if that's 
possible ?
Anyway - will flash later today, and tell you the results.

Original comment by andre.kj...@gmail.com on 15 Apr 2012 at 10:03

[GoogleCodeExporter]

     I suspect there is a genius somewhere who could tell you how to download the current firmware but I don't know him I'm afraid :-)

Original comment by rmackay...@gmail.com on 15 Apr 2012 at 10:09

[GoogleCodeExporter]

Reprogrammed.
I am unable to reproduce the problem anymore, using default throttle travel 
(1100-1900).

The other, problem I thought was related to PPM decoder is still there.
http://code.google.com/p/ardupilot-mega/issues/detail?id=566
in short:
Ch8 PWM is 1705uS - Mode5 should be effective, - APM, OSD and FlightPlanner 
confirms "stabilize" - but the plane is really in manual (bypass?)
No AP mode programmed into mode5 (AUTO/Stabilize/Loiter) does anything, while 
working just fine if trigged by mode2,3 or 4.  To mee, it seems like mode5 
(even it's middle, 1705uS)is within that the PPM controller thinks should be 
bypass.

Original comment by andre.kj...@gmail.com on 15 Apr 2012 at 1:18

[GoogleCodeExporter]

Andre,
    Thanks very much for your contributions towards nailing this twitching issue.  This is a great outcome as it shows where the issue is.  As you've noticed this 'twitching' has appeared in a couple of places - we will see if this older ppm firmware is the cause of all of them.

    I'm going to close this issue now but the others will stay open.

Original comment by rmackay...@gmail.com on 16 Apr 2012 at 1:09

• Changed title: ArduPilot 2.28 - twitching aileron
• Changed state: Fixed