Replies to This Discussion

Permalink Reply by Paul Marsh on August 1, 2012 at 5:33pm

While looking into this, I first discovered that going back to Mission Planner 1.1.99 allows the simulation to work.  However, I have discovered some strange waypoint behavior.  In the middle of the simulated mission, the plane started to fly to a waypoint not in the flight plan.  It was actually going to a waypoint left over from a real mission.  In other words, the simulated waypoints did not fully overwrite the previous real waypoints.  This didn't happen until about waypoint five or six.  I can't seem to completely get rid of these ghost waypoints.  I reloaded the firmware for real flights.  Deleted all waypoints and loaded waypoints to clear them them, i.e. loaded with no waypoint set.  Reading back waypoints seemed to indicate that they were gone.  I played around with loading, reading and deleting waypoints in real mode and all seemed well.  I then reloaded the sim firmware and discovered that these ghost waypoints were still there.  Nothing I do can get rid of them.  I even pressed the reset button on APM2 hoping that might clear things, but it didn't.  This seems to be a problem with the sim firmware only, although I can't be entirely sure.

• ▶ Reply

Permalink Reply by Paul Marsh on August 1, 2012 at 5:51pm

Another clue to the problem, maybe.  The problems seem to start with waypoint 5.  In my previous real mission, I had set waypoint 5 to "Do Jump" to waypoint 1 for -1 repeats so that it would keep reflying the mission until I went to manual mode.  I noticed that whenever I set a "Do Jump,", that waypoint disappears from the flight planner and flight data views.  Nevertheless, the mission goes just fine.  Now in sim mode, my problems seem to start with waypoint 5.  I can't seem to overwrite it.

• ▶ Reply

Permalink Reply by David Jones on August 1, 2012 at 9:31pm

Have you tried to completely erase the EEPROM in cli mode? This should delete all "old" waypoints from the chips memory.

• ▶ Reply

Permalink Reply by Paul Marsh on August 2, 2012 at 5:01am

Hi David.  Thanks for the response.  I will give that a try.

• ▶ Reply

Permalink Reply by Paul Marsh on August 3, 2012 at 6:59pm

Still getting strange behavior in simulation mode.  I did both a reset and erase from the CLI.  On my first attempt to run a simulated mission, the plane went to the first few waypoints OK, but then skipped the rest and headed to home.  What I found was that after writing all of the waypoints to APM2 and then reading them back, not all waypoints were there.  Ultimately, I had to write the waypoints about three times in a row before starting the mission to get the simulated mission to complete.  Basically, there appears to be a problem with writing waypoints to APM in simulation.  Also, navigation is not acting as it did.  The plane does not track directly to waypoints--it takes a curved path, but does get to the waypoint.  Finally, the loiter is now an oval and not a circle.  None of this was happening before I updated Mission Planner.

If there is a problem, I hope this helps resolve it.

Thanks a great deal for this software, though.  It works great for the most part, and appears to be working great in real mode as opposed to simulation (at the moment--I'm sure this will get resolved).

• ▶ Reply

Permalink Reply by criro1999 on August 27, 2012 at 6:47pm

I do have the same "oval loiter" behavior, in real flight, see attachment, any idea?

Arduplane 2.6, MP 1.2.9.

Attachments:

• Loiter_oval.JPG, 78 KB

• ▶ Reply

Permalink Reply by Paul Marsh on August 28, 2012 at 5:29pm

Hi criro1999,

The oval behavior you are seeing is a lot different from what I observed.  It appears as though your aircraft is over-correcting.  However, it could still be related to what I was seeing as I would expect the plane to settle into a circle no matter what.  If it's trying to fly an oval and over-correct at the same time, it might end up looking like your photo, which resembles a peanut.  You might try lowering the navigation PIDs (the "P").  That's really just a guess on my part.

Paul

• ▶ Reply

Permalink Reply by criro1999 on September 5, 2012 at 5:33pm

Hi Paul,

I did apply little bit the correction and it seems it works, see attachment, maybe I need more. Should I lower more the navigation PIDs? Loiter radius = 40, Default Alt = 50, Loiter Alt = 50, MP 1.2.10, APM2, ArduPlane 2.6

What I did notice is the the Altitude during the Loiter is changing (red line) comparing to Target Altitude (blue line) and I am wondering why so many difference?

Also, one other information: when in Stabilize and trim the plane to fly horizontal and stable, if I switch to Auto and the back to Stabilize, the plane is not more flying stabilize (it loosing the altitude) and points to ground, which means the previous stabilization is lost. Is any thing with the baro altimeter? How should I check if baro is working OK? Which parameter from tuning refers to baro altitude and which one to GPS altitude? If baro altimeter is so precise (10cm), why so big oscilations?

Thanks again Paul for all your solutions.

Attachments:

• loiter1.JPG, 60 KB
• loiter2.JPG, 67 KB

• ▶ Reply

DeveloperPermalink Reply by Michael Oborne on August 4, 2012 at 12:41am

Paul i will check this out. last time i checked all was good

• ▶ Reply

Permalink Reply by Paul Marsh on August 4, 2012 at 3:50am

Thanks Michael.  In case any of this applies, I'm running Windows 7, 64-bit.  Also, I'm running MP on one PC and X-Plane on another.  Prior to this I had been running both on the same PC.  I haven't tested again with both on the same PC.  Regarding flying from one waypoint to another, what's happening is that the orange line points off to the side of the given waypoint.  The plane arcs out, but as it nears the waypoint the orange line comes closer to pointing directly to the waypoint.  I thought that perhaps simulated wind was pushing the plane off course, but I have all wind at all altitudes turned off in X-Plane.

I'll be flying for real this morning.  I'll watch carefully, obviously, but writing and reading back waypoints with real-mission firmware loaded works perfectly in MP.  All waypoints come right back.

Thanks again,
Paul

• ▶ Reply

Permalink Reply by Paul Marsh on August 4, 2012 at 10:32am

The odd behavior I'm seeing has occurred at the field now as well when flying in real mode.  After the plane initialized, the icon would not point in the right direction.  It was almost as if APM2 were installed backwards.  I double-checked that both the compass and auto-declination were enabled and then re-initialized.  No joy.  The only way I could get the icon to point in the right direction was to disable the compass altogether.  I went ahead and attempted a flight.  The flight went OK, but the orange line never pointed directly at the waypoint in question.  I have tracking set to 0 degrees so that plane will simply make a bee line to the next waypoint.  Just like in sim mode, this does not seem to be working quite as well as before.  There was some other odd behavior.  For example, the first time I check the flight surfaces in stabilize mode they seemed OK.  I checked again at one point and neither the elevator or rudder were responding.  I checked again before I actually flew and they were all working as expected.  Other than re-initializing, I didn't change settings.

Maybe I've done something differently than before, but I can't say what that would be.  The laptop I used at the field is running MP 1.2.1.  I still can't get sim mode to work in the latest upgrade.  I had to drop back to MP 1.2 for that.

One other thing--when I played back the log of the actual flight, none of waypoints were displayed on the flight data page.  I loaded the waypoint file in flight planner and could watch the flight playback there over the waypoints, but when I switched over to the flight data page the waypoints would still not display.

I hope this helps.

Thanks.

• ▶ Reply

Permalink Reply by Paul Marsh on August 5, 2012 at 6:38am

While troubleshooting the recent problems I'm having, I made an interesting discovery.  First, I went back to a previous version of Mission Planner to see if the problems would go away.  I'm on 1.1.92 mav 1.0 now.  The strange behavior with writing and reading back waypoints in sim mode remains.  However, regarding waypoint following, here's what I found.  Since my real flight paths are so short, I set crosstrack correction to 0 degrees.  Prior to this, it worked great in both sim and real modes.  The plane would make its turn and make a bee line to the next waypoint.  The orange line would point directly to the waypoint center the whole time.  Now, even with MP 1.1.92, the orange line points off to the side and the plane flys a curved path to the waypoint.  I tried an experiment.  I put back crosstrack correction to 30 degrees.  In sim mode at least, the plane now flys along the flight path and directly to the center of the waypoint after making its corrections.  However, the orange line continues to point off to the side.  The plane now flies the green line and seems to ignore the orange (see image below).  Based on my flights yesterday, this behavior exists in real mode as well.

I'd like to test rolling back firmware, but I don't know how to do that.  If someone can provide the steps, I'd like to give it a try.

Thanks.

• ▶ Reply

• ‹ Previous
• 1
• 2
• 3
• Next ›
• Page