Multiple APM 2.1 / HK GCS bugs

[GoogleCodeExporter]

What steps will reproduce the problem?

1.Again problems with altitude reporting in HK GCS
  Altitude oscillations at steady state from between -50 to +50 meters
  (The pressure sensor and GPS have been tested and thier altitutes  remain steady)

2. Waypoints incorrectly updating. IE the waypoints will randomly jump to the 
next waypoint even though the radius is at 30m and the waypoint is 300+ m away 

3. Throttle value in HK GCS respond to rudder changes not throttle changes

4.  Target altitude never changes from 0.  The loaded waypoints altitude vary 
from 35m to 50m.

During a test flight today once switched into auto mode the plane headed 
towards the 1st waypoint but tried to hold an altitude of 0m causing a crash

The control surfaces all check out. IE stabilize, FBW check out

Everything checks out in the simulator

What version of the product are you using? On what operating system?
Everything is updated through SVN
APM 2.1
Mission Planner 04.24
HK GCS 1.2.89
Vista

Check uploaded screen shots for a better idea of whats going on

I've tried anything and everything I can think of... Help is appreciated

Thanks 
-BP

Original issue reported on code.google.com by bdphilli...@gmail.com on 3 May 2011 at 4:00

[GoogleCodeExporter]

hmm, I got similar problem still in 2.012. Target altitude problem. FBW, 
stabilize mode checked and flight OK and after switched in return to lauch 
airplane just start losing altitude and tried to flight into ground. I could 
not replicate this in Xplane.It all works ok in simulator, but real airplane 
not ok. I am struggling with this 2 weeks now. I am not sure about this error. 
Could be also error by me. Will test it further.
I am trying now to set new PID settings.  I saw strange things once, when 
updating rtl altitude. Once I got -22000 meters in EEPROM, when parameters were 
read. I did not write negative value. Maybe absolute and relative altitude 
caused some confusion when writing altitude by HK GCS and reading with APM 
planner. I could not replicate that again. I will test 2.1 now.

Original comment by metod.iv...@gmail.com on 3 May 2011 at 6:50

[GoogleCodeExporter]

Original comment by dewei...@gmail.com on 4 May 2011 at 1:37

[GoogleCodeExporter]

1) A flight recording would be helpful. Use the Data File Tab bottom left, type 
in a new name and hit the record button, then post the .hko file here.

2) Radius removed from GCS in latest updates. Radius cannot be set per waypoint 
in APM's MAVlink implementation.

3) Throttle channel can be selected on File, Settings, General tab

4) I have been considering showing the next waypoint's altitude according to 
waypoint data in this field as only ArduPilot Legacy currently provides that 
data.

5) All altitudes were absolute up until the latest few versions of the GCS. Now 
there is a checkbox to specify if all altitudes are absolute or relative to 
home. 

Another thing that is not well explained is the fact that I'm converting all 
the parameter values and altitudes into whatever you've selected for units. So 
for example, if you've selected feet for your distance value, APM uses meters. 
I automatically convert from meters to feet which is what you see on the GCS 
and then back to meters when it's saved to the APM. This is why there are the 
two decimal places and it's not exactly what you typed in the field when you 
hit save and then read.

Original comment by paulbmather@gmail.com on 4 May 2011 at 1:51

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Paul - Thanks for the quick reply

Throttle value taken care of, I overlooked it in the general settings 

I've attached the .hko file from the GCS and the log from APM (for cross 
reference) from a test just sitting on the bench so all values should reflect 
zero (approx.) altitude, position and airspeed change 

You should be able to see what I was trying to describe in my earlier post.  

First the target altitude does not reflect the 100m set for the waypoint in the 
mission planner.  I am still a little fuzzy on what APM "sees" as the altitude 
value.  Is the GPS or pressure transducer set as default? I've seen the 
"altitude-mix" variable change in recent releases of the code. And which of 
these is displayed in the GCS? I think it would be a good idea to use the 
legacy's way of showing the next waypoint's altitude, but easier said than done 
with the upgrade to MAVLINK.

Second, the waypoint jumps to the next without being inside the waypoint 
radius, I was able to duplicate this in the attached log file.  By switching 
modes, between auto, manual, and stabilize it trigger the waypoint as reached 
and moved on to the next waypoint.

I am flying an easy star with APM, UBLOX, MPX7002DP Pressure transducer, 900MHz 
Pro Xbees, upgraded 2S li-po battery, everything else stock configuration.

Thanks - BP

Original comment by bdphilli...@gmail.com on 4 May 2011 at 4:06

[GoogleCodeExporter]

Right now, I'm not getting any data from APM about target altitude. The AP 
Legacy included that field and another auto pilot is planning on adding this in 
the future. I believe the APM developers' plan is for the GCS to figure out 
what the APM is doing for altitude. With basic missions this won't be a 
problem, but with more complicated missions this might be a challenge.

As for the values shown in the hko file, the drift on altitude is simply what's 
being sent to me by the APM. I am displaying the junk data I'm getting 
correctly. You'll notice that you've got 0 sats.... which can explain the 
altitude drift. 

As for changing waypoints, again, that's APM's decision to make. I'm just 
showing what the APM is telling me. I am unable to get my setup on my desk to 
duplicate what you're seeing. I can see the index advancing by switching 
modes....but I can't explain why the APM is doing that. What value do you see 
on the configuration tab for WP_RADUIS?

So far
1) APM's decision
2) APM's decision (my GCS could have a typo that's setting the radius wrong)
3) Setting
4) Not currently tied to any APM message (could just read the next waypoint's 
altitude and plug it in)

Original comment by paulbmather@gmail.com on 4 May 2011 at 4:52

[GoogleCodeExporter]

The 0 sats thing is an error with the APM and UBLOX communication.  I have 
never seen anytihng other than 0 sats. Again an APM problem:

http://code.google.com/p/ardupilot-mega/issues/detail?id=142&can=1&q=number%20of
%20sats

In the CGS it is correctly showing WP radius as 30m and altitude RTL at 100m 
(as set in the Mission Planner)

I don’t know why the WP is changing as it is. There is not much in the way of 
documentation out for the APM 2.1 code yet.  I guess the APM software guys 
should take a look at what's going on and see if they've got any suggestions.  
I'm sure I'm not the only one experiencing problems

Original comment by bdphilli...@gmail.com on 4 May 2011 at 5:13

[GoogleCodeExporter]

Do you know how to use uCenter? You should probably confim that NAV_SOL is 
enabled...

Original comment by paulbmather@gmail.com on 4 May 2011 at 5:19

[GoogleCodeExporter]

Are you able to access the GPS unit from uCenter through the APM board or is 
there an external USB adaptor needed?

Original comment by bdphilli...@gmail.com on 4 May 2011 at 5:41

[GoogleCodeExporter]

I have not had good results getting it to work.

http://diydrones.com/profiles/blog/show?id=705844%3ABlogPost%3A148030&commentId=
705844%3AComment%3A148074

You can try the attached file by writing it to your APM

Original comment by paulbmather@gmail.com on 4 May 2011 at 5:52

Attachments:

• uBlox_Program.pde

[GoogleCodeExporter]

Success!

The little code for transfering the data directly to the GPS output worked

Enabled the NAV_SOL from the uCenter

Number of sats reporting as 10 now in HK GCS

Original comment by bdphilli...@gmail.com on 4 May 2011 at 6:04

[GoogleCodeExporter]

I can't say if that's going to solve any of the issues above. When did you buy 
that uBlox and did you get it from DIYDrones? If so, you're the second person 
I've talked to that was missing the NAV_SOL messages enabled on his uBlox. It 
might be time to have Jordi check his configuration procedure.

Original comment by paulbmather@gmail.com on 4 May 2011 at 6:30

[GoogleCodeExporter]

I purchased the GPS from DIYdrones in Sept 2010.

As for everything else acting funny I think there are other bugs in the code 
but fixing the GPS was a start.  I'll do some more troubleshooting and see if I 
can pin point the issue

Thanks again for your help

Original comment by bdphilli...@gmail.com on 4 May 2011 at 12:48

[GoogleCodeExporter]

Please try again....never ending bug fixes....

Original comment by paulbmather@gmail.com on 9 May 2011 at 5:48

[GoogleCodeExporter]

Original comment by paulbmather@gmail.com on 4 Jul 2011 at 5:48

• Changed state: Fixed