Servers by jDrones

Drift of quad in land mode


(Ashwani gupta) #30

ok sir i will follow this. Thanks for your valuable suggestion.


(Ashwani gupta) #31

Hi Rick,

i made a mission to execute two circles at an altitude of 20 m then go to another point and then come for RTL (point was defined for RTL). But what it did was completely annoying and out of the box. will you please see this attached log and comment that what is the reason behind this ?
https://drive.google.com/open?id=1XMgpMoekl4lpSEaW_cy9Mfwu6IUinYgC


(Ashwani gupta) #32

Dear Rick,
please suggest the way to resolve this problem and the level of diagnosis that i should perform to eliminate these things completely.


(Rick) #33

I see two problems in this log:

First, your transmitter failsafe is repeatedly being triggered. It looks like a bad connection with your RC transmitter. Each time the RC In channel 3 (throttle) drops is a disconnect. Eventually, it completely loses connection as the drone flies away.

This is a problem that you should fix, but it did not cause the fly-away. I think we saw radio problems with one of your earlier flights. Are you using the same transmitter?

The second problem is what caused the drone to fly too high: vibrations.
This is the first part of your flight, when the drone is ascending to 20 meters (24 meters EKF altitude).


In ArduPilot flight logs, you can see what the drone is trying to do by comparing desired versus measured attitude. For example, in this graph, the drone wants to make the red line (measured altitude) follow the green line (desired altitude). You can see that in the first half of the graph, this is ture - the red line follows the green line upwards until it gets to the correct altitude. However, in the second half, you can see that the red line begins to go upwards, even though the green line stays below it. This means that the drone is climbing upwards, even though the flight controller doesn’t want to. This indicates that the flight controller cannot accurately estimate its climb rate.

Climb rate estimation heavily relies on the accelerometers, which can be confused by excessive vibration. Here are your vibration readings:

In general, you want the vibration measurement to be less than 30. Also, the accelerometer clipping VIBE.Clip should never increase in flight. The very high Z vibrations and clipping are certainly the cause of the uncontrolled ascent by the drone.

Try to mount your flight control board to reduce vibration, and make sure that nothing is touching it that could rattle against it, such as loose wires. If you post a picture of your current mount, we might be able to suggest some improvements.


(Rick) #34

The short version is, you must do two things:

  1. Fix the RC transmitter/receiver so that you do not lose connection in flight
  2. Fix the vibrations so that they are less than 30, and the clipping does not increase while in flight (some clipping on takeoff or landing is okay).

(Ashwani gupta) #35

Sure I will post pic don’t arrange t.


(Ashwani gupta) #36

But one thing making me confuse is that when I made a mission with only 40 m of max altitude at way points. How it is taking this high altitude in the desired altitude (DAlt) flight log whereas mission is design only for 40 m ??


(Rick) #37

The desired attitude can get “dragged around” by the actual attitude. This is true for technical and safety reasons, which can prevent the drone from doing crazy maneuvers.

For example, in your case, if the desired altitude stayed at 40 meters, the drone would keep trying to descend by 100+ meters. But what if the measured altitude was wrong? The drone would try to fall into the ground. This “desired dragging” makes sure that the drone never tries to execute extreme maneuvers by ensuring that the desired and actual attitude are always reasonably close to each other.


(Ashwani gupta) #38

HI rick ,

       is there any way to check the capacity of my drone that how much wind it can take ? what kind of changes you are suggesting in parameters to increase the capacity keeping all the hardwares same ?

(Rick) #39

You can use eCalc to get an idea of your drone’s payload capacity. Otherwise, you can build a test stand to measure the maximum thrust of your motors.

There’s no real way to increase the drone’s payload capacity without changing hardware. It is limited by the motor’s performance and the weight of the drone.


(Ashwani gupta) #40

Hii Rick, Actually i don’t want to increase the lifting capacity of my drone rather want to improve the performance like fighting against the gust and even fly against the gusty wind, more stability, less vibration etc (keeping the power plant unchanged). Is there any suggestion from your side, link or some kind of reading stuff that you can provide me so that i may get some idea to incorporate the amendment that i mentioned above ?


(Ashwani gupta) #41

hi rick ,
i have a doubt about the vehicle performance. what will happen if i use a overpowered quad copter in auto mode (alt hold, pos hold) and in manual mode (stabilize and acro) ? How does it behaves ? If i want to forcibly use an overpowered copter what will be the consequence ?


(Hari .P) #42

Hi, Ashwini,
A perfectly tuned (i think it is a never ending process) Copter will perform well in any mode.
Pls post some photos & specs of the overpowered copter, so that the experts can chime in,
Best Wishes…

,


(Rick) #43

Stability and wind resistance is mostly a matter of tuning. There’s a long list of things that can improve the tuning of your drone; some of these things are quite advanced and are beyond most users.

  • Reduce vibrations
  • Tune PIDs and acceleration limits (Autotune will do this)
  • Tune filters (gyro filter, rate filter, notch filter. May require FFT analysis)
  • Characterize powertrain (MOT_THST_EXPO, MOT_SPIN_MIN, MOT_SPIN_MAX. Requires a motor test stand to do properly.)
  • Maximum performance tuning (THR_MIX_MAX, other output shaping parameters. Requires a lot of flight testing)

Most users do the first two, and those are usually the most important; as long as you get your vibrations low and run Autotune a few times, your copter should perform quite well. The other items require testing and analysis to do, but they can turn a good tune into a perfect one.

A very overpowered copter requires careful tuning to fly properly. This is because when the motors are running at very low throttle, the copter will have a limited ability to throttle down some motors when it is maneuvering or dealing with wind. This generally results in a drone that is a bit wobbly in wind and cannot perform precise maneuvers because its rotational acceleration will be limited.

It is possible to have a well-tuned copter that hovers at ~15% throttle, but it is difficult to do, and probably requires the 3 last items in my tuning list. If you want a powerful copter, is generally better to have a copter that hovers around 25% throttle.


(Ashwani gupta) #44

Thank you rick.
as per your suggestion i purchase a taranis Q X 7 ,transmitter with vibration damper and a carbon fiber frame with T-motors of 650 kv for 12-14’ of quad.
may be somewhere at the weekend i will assemble it and share a pic with you.
Thanks for your continue support. .


(Ashwani gupta) #45

Hii Rick,

what is the most stable firmware for pixhawk 2. recently i am moving slightly towards the professional drone with following component.
pixhawk 2
here plus rtk
tatto battery 10000 mah
taranis q x 7
turnigy 40 amp esc
vibration mount
pixy ir lock
rp lidar
px4 flow sonar
T motor 650 kv


(Ashwani gupta) #46

hi rich,
i attached here+ gps to the system but unable to do the compass calibration on AC3.5.5. Do you know how to fix this problem if you used it before ?


(Oldgazer) #47

Why don’t you tell us IN DETAIL what you are doing and EXACTLY what indications your are getting.

To give you a clue posting this: " i attached here+ gps to the system but unable to do the compass calibration on AC3.5.5. Do you know how to fix this problem if you used it before ? ", is like walking into a Doctor’s office and saying , “I don’t feel good.”


(Ashwani gupta) #48

ohk,

i got a here+ rtk gps , that i attached to the pixhawk cube. while calibrating the compass the external compass is not calibrating, somei=time reaching to 33 % and stop forwarding and sometime it is reaching till 99%, giving a sad tune and restarting the calibration of the external compass.
now during this some time it is not allowing the internal compass also to get calibrated. when i am removing the external compass and calibrating only internal compass it is not showing any problem. more over while going for the live calibration it is leaving some white dot behind and the entire sphere is shifting very far from center. Due to this i am getting the offset value of external compass in 2000+. because of that i am not able to use this important hardware. Just for one trail i changed the firmware from 3.3+ to 3.5.7 but there was no change and problem is still the same.
that what all happening.


(Ashwani gupta) #49

Hello,

I want to read about the description of options written in the temp option of mission planner. Do you have any link associated with that ? I want to read details description about the swarm, microdrone, moving base, follow the leader, inject ge etc. please send me any link if you are having ?