Hi.
I have a problem after updating AC to 3.2.1 from 3.1.5.
When in Loiter, the the speed of the motors rev up and down/pulsate quite rapidly. This was working fine when I was using previous versions of arducopter.
I have downloaded the logs, and have looked into accX, accY and accZ. These seems to be inside the range specified in the wiki at arducopter. (Log_vibrations.bin)
Please have a look at the Log.bin, and note the difference between ThrIn and ThrOut when I'm in Loiter mode.
Anyone having any idea what the problem may be?
Replies
Now here what I did.
Model: X quad. 850kV motors running on 4S with 10x4.5 props.
1. Disabled GPS at all.
2. On 3.2.1 firmware (binary log), made a couple of back and forward flights, then a short trip round and then did still hover.
Overall performance was satisfactory, but copter had that feeling that something's not right. PID algorithm should give swift and smooth adjustments. This one had feeling of analog servo working with bad bearing. You could call it smooth twitching.
On long forward accelerations there was slight "dive".
3. Flashed to 3.1.5, disconnected GPS and compass.
Did all the same tests. Well, I don't know what magic was used to write 3.1.5 but you should use it again. All elevation changes were smooth and altitude hold was solid without twitches. Long forward run did give me a bit "dive", but this could be airframe dependent effect.
Might be important, that altitude gain and descent seemed to me much more calm (I mean on full throttle down, copter was descending slower than on 3.2.1) than on 3.2.1 (could be somehow connected to sensor resolution, past and present).
That's pretty much all.
p.s. I think 3.1.5 doesn't even need EKF, just add that splendid features like RSSI scaling, GPS controls, etc, not harming copter behavior, and maintaining compatibility and you need nothing more.
2015-04-13 18-49-43 3.bin
2015-04-13 19-26-07 5.log
No problem at all. I only need 1.5 hour for battery to charge.
I'll make 3.2.1 log and then 3.1.5 log. What kind of flight is preferable?
I'd suggest:
1. Hovering in alt hold.
2. Linear flight "forth-back"
3. Else?
As my quad never pulsates while hovering but does it while there are pitch or roll inputs (it loses hight dramatically and pulsates till it reaches correct hight again), I suggest fast forward-backward and left right maneuvers for the test
For the analysis it might be also important if there was wind or not. Perhaps strong wind makes it worse. (I am not sure, because in our region it's always windy... :()
I guess there are two ways of resolving present problem:
1. Cutting out something and trying to fit in EKF part into 3.2.* firmware.
2. Return to 3.1.5 way of estimation, because 3.2.1 is not better at all. In fact 3.2.1 is worse than 3.1.5 because it's just not flying right. It was transitional buld from 3.1.5 to 3.3 with EKF (hope EKF helps after all), it was bad idea to leave it "last stable for APM boards".
I admit, that if you have no vibrations (for example on small drones with high kV motors), 3.2.1 behaves satisfactory, but if you run bigger drone it becoming a headache "catch the vibration".
Migrating to Pixhawk is not an option cause if this is the way problems are resolved I'd rather migrate to another project like dji, paparazzi or openpilot.
I'm sorry to complain, but it is really irritating, I haven't had so much trouble with controller because of vibrations ever and this included newly developed boards that didn't want to fly at all and needed to be tuned.
It's time to make decisions, at least I'll waste no time hoping it will get better.
Well, downgrade is pretty simple:
I don´t know exactly, wether my experience has the same reason, but perhaps it can help someone -
i have a 3dr Y6B with larger props and motors - the upper props are only 10 cm away from the pixhawk. When flying without the canopy the copter is very unstable especially in height - it bounces up and down about 2m. When I fly it with the canopy it hold heigh very stable even in gustling wind. Also the when braking from fast forward flying it doesn´drop as much as it does without the canopy. In the log i did the first part without, the second part with the canopy.
2015-04-11 18-54-17.bin
Perhaps your props near controller are affecting the pressure
I´m sure the props will affect the pressure above the copter - I thought that´s the reason for flying :-) .
My idea is that the reason for the copter drifting up and down is some kind of resonance in the control loop:
Copter drops - pressure goes up - throttle is increased - pressure goes down because of prop effect (not height) - throttle is decreased, but copter is not really climbing ...
As you can see from the GPS the peaks in the pressure are not only caused by the props but also by height.
Anyway, with the canopy mounted the problem is solved.
APM 2.6 3.21 firmware 810KV 10x4.5 Deadcat frame 3 Cell
Same problem as everyone here. Made two flights today one early with wind second flight late with very little wind. No changes to Quad between flights. My second flight was 90% better then the first. WHY ?
Any help take a look at my logs
Please guys, as Rob says, always report issues with logs so developers can check if there is a firmware issue or an user one, that helps us to have a better code to enjoy our flyes ;) ; I grumble a lot with this problem but only on one cuad and it was a sum of problems on my cuad, developers report that the next release is going to be less sensible to vibs but only for Pixhawk.
-
8
-
9
-
10
-
11
-
12
of 16 Next