Looking for some suggestions to try and troubleshoot an interesting problem. I have a standard ArduCopter kit (the most recent version from the DIYDrones store, + GPS / magnetometer / sonar, but no telemetry) and am running 2.0.23. In addition to the drift issues under discussion in another thread, I have noticed that in altitude hold and gps loiter modes, I have a great deal of difficulty with altitude.
When operating close to the ground (ie what I assume is sonar range) if I put the copter in a stable hover and then engage the elevation hold, it works pretty great, it holds through a range of about a foot or so. Engaging alt hold higher off the ground (6'-8' altitude approximately) causes the copter to head straight up into the sky at a fairly good clip. This is consistant and every time I've seen it I have aborted the hold and gone back to Stabilize as I don't want to lose the copter. This same thing happens when engaging GPS loiter mode; loitering at low altitudes works well, loitering higher than about 6'-8' causes it to hold position fairly well but progressively gain altitude.
Also not clear if this is related: even at low altitudes in hold mode, it can fly around fairly well and it will hold altitude, but if I start moving at a good speed (stick far forward, copter angling ~30 degrees) it will begin a rapid ascent as described above.
I saw this with the barometer uncovered and so added a small piece of foam from headphones over the sensor hole, and it doesn't appear to change anything.
Both the ALTITUDE and SONAR tests in the CLI give what appears to be decent data when I run them over USB, the sonar's a lot more precise but I believe that's normal.
Any thoughts? Any other information that would help troubleshoot this?
-Chris
Tags:
Thanks for the beta test report. Alt hold is something we're working on is being revised in 2.0.24. It should be pretty solid at all altitudes by the time we take AC2 out of beta in a week or two.
I assume you checked your altitude sensor readings in the CLI test menu and they looked fine.
Permalink Reply by vaeddyn on June 8, 2011 at 2:12pm Yup - well, at least as far as I can tell eyeballing it. Sonar is pretty accurate as I move the copter up and down from my desk/floor, not much fluctuation. Baro has a lot of noise to it but it does appear to be working.
I would love to be able to see what it was reading in the air. Guess telemetry is my next hookup.
Permalink Reply by Jim Phoenix on June 8, 2011 at 2:40pm
Permalink Reply by vaeddyn on June 8, 2011 at 8:55pm
Permalink Reply by Jim Phoenix on June 8, 2011 at 9:55pm
Permalink Reply by Al Ros on June 8, 2011 at 10:56pm Hi Guys
I was testing Alt hold tonight and loiter. What I found was the sonar worked much better if during power up it had a foot or more off the ground rather than sitting on the ground.
Also If i activated altitude hold when the quad was in the sonar range, it held altitude well, then I activated it when I was well above sonar range, and it also held reasonably well by baro. However, in one of the tests I think I was around the boarder between sonar and baro, and the quad did what you guys described. For a few seconds it held ok presumably with sonar, then it rose slightly to what I think was into the transition altitude zone, at which point it shot up at near full throttle
can you guys confirm if yours will hold well when low in the sonar altitude range and when fairly high in the baro altitiude, then check the middle transitiona altitudes?
p.s. I have mag and GPS enabled
Thanks
Al
Permalink Reply by vaeddyn on June 8, 2011 at 10:59pm in my case:
I haven't done an exhaustive test but that's my observation at this point. Tomorrow assuming weather cooperates I plan to test a bit more and also try running with the sonar disabled and see if anything changes when baro's all it's got to work with.
For what it's worth, I have never powered up when not sitting on solid ground and my sonar hold is great. In what way is it working better if you boot up a foot in the air?
Permalink Reply by vaeddyn on June 9, 2011 at 5:12pm Update: Today I removed the sonar (which, as I never specified, is an LV-MaxSonar-EZ0 from the diydrones store), disabled it in the CLI through APM Planner, and went for a flight. Altitude Hold and GPS Loiter now work at all altitudes I tested, I could not get the copter to take off vertically the way I'd seen before, it held onto its altitude well though with more fluctuation than I saw with the sonar at low altitudes. I would expect that to be the case.
In alt hold mode I was even able to maneuver a bit aggressively and it sustained its alt.
It looks to me like at least with my sonar unit, the copternever seems to pay much attention to the output from the baro. As much as I miss my rock-solid low altitude stuff I think the sonar's going to stay disabled for now until we know more.
Permalink Reply by John L. on June 9, 2011 at 6:22pm I have the LV-MaxSonar-EZ4 mounted and don't seem to be getting consistent altitude hold at lower altitudes.
I noticed that now in the manual it has the XL series shown.
Are the output scalings the same between the 2 series?
The advantage of the XL over the LV, is that when you power up the LV it calibrates and needs to be at least some distance away from anything (about a foot I think, see datasheet) which is a pain, because it means holding the Quad off the ground every time you connect the battery, which needs 3 hands. The XL doesnt have this restriction, plus it has other advantages, except for higher cost! Yes, it seems the LV has been quietly dropped.
Rob
Permalink Reply by Al Ros on June 10, 2011 at 9:42am Ok very interesting - I will try again - I just received an XL sonar
Al
Permalink Reply by Vincent Mees on June 10, 2011 at 12:11pm Also just received my sonar.
It works veru accurate but in the planner I get 33 cm when the sonar is against the floor. Also the alt err is 33cm.
It seemsit doesn't go below 33. Is this normal?
Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.183 members
674 members
1282 members
116 members
87 members
© 2013 Created by Chris Anderson.
Powered by
