RC India
Welcome Guest, please login or register.
 
Pages: [1]   Go Down
0 Members and 1 Guest are viewing this topic. Topic Tools Topic Tools 
Read
« on: July 10, 2018, 05:56:04 PM »
vaisakh032
Plane Lover
Active Member
**

Reputation Power: 1 
vaisakh032 has no influence.
Offline Offline

City: Thiruvananthapuram
State: Kerala
RC Skills: Intermediate
Posts: 9
Join Date: Feb, 2018




Guys Hats Off, i am new to the forum, but have been following them for a while. I've been working with copters for around 3 years now and have some basic knowledge over its working.
My present build is a **Hexacopter**, with specs:

Frame: Tarrot 680 Pro
Motors: EMAX MT3506 - 650Kv
ESC: Hobbywing X-Rotor 40A
Controller: PixHawk Cube
GPS: Here+
Rc Reciever: X8R
Rc Transmitter: Taranis X9D plus
Gimball: Tarrot 2axis GoPro Gimball
Lipo: TATTU 6200mAh 4s 35c.
Apm Version: Copter 3.5.5

Story:
 She was designed for Aerial Mapping, and was flying perfect. It was not her maiden flight that day, she flew great before then. Good control, nice autonomous flight, and great Altitude hold in Alt.Hold mode. But the day when i installed the Gimball, everything started to turn around. First i thought it was the issue with the gimball but then after digging through the logs, i found it was not the gimbal but something else.

Error:
 After plugging in the power we had around 3 to 4 flights. In the very first one, the altitude dropped in Alt.Hold Mode and i had to flip the switch before it was about to hit the ground. After that i tried 3 more flights. This time the drop in physical altitude of the hexacopter was rectified (i didn't do anything), but the Mission planner keep showing drop in altitude. That is, from the second flight onwards, thr physical altitude of the hexacopter was rock stable, but the data in mission planner showed a liner drop in altitude, which even went to negative values.

From Log:
  From the log it understood that, in the first flight the CTUN.DAlt (desired altitude) dropped and the hexacopter followed it (Dont know why the Desired altitude dropped, as there was no reason to drag it down for the controller). And in rest of the flights, the CTUN.DAlt was constant and she stood rock solid in the locked altitude (but this time mission planner showed drop in altitude).

So guys, please have a look at this log file, and help me find the issue Huh?
log:https://drive.google.com/drive/folders/1mFio8fdIlAOvsFK-nn94MGfH4vB4r27N

Screenshot from 2018-07-09 18-31-56.jpg
Stable Copter but Mission planner shows drop in altitude
* Screenshot from 2018-07-09 18-31-56.jpg (30.04 KB, 800x225 - viewed 232 times.)
This post has 1 more images(s)/attachment(s). Please login or register to view them.
« Last Edit: July 10, 2018, 06:08:10 PM by vaisakh032 » Logged
 

Read
« Reply #1 on: July 10, 2018, 09:57:48 PM »
kn8alpha
Plane Lover
Senior Member
***

Reputation Power: 2 
kn8alpha has no influence.
Offline Offline

City: Bangalore
State: Karnataka
RC Skills: Advanced
Posts: 170
Join Date: Feb, 2016



Is the hover throttle value higher because of more weight?

I've always observed Dalt to track actual altitude than commanded altitude, it's not exactly mapped to baro altitude.

Trying undoing the last couple of changes to the quad and see if this issue goes away to identify the exact cause.

And welcome to the forum! Happy building.
Logged
 

Read
« Reply #2 on: July 10, 2018, 11:01:27 PM »
vaisakh032
Plane Lover
Active Member
**

Reputation Power: 1 
vaisakh032 has no influence.
Offline Offline

City: Thiruvananthapuram
State: Kerala
RC Skills: Intermediate
Posts: 9
Join Date: Feb, 2018



Hai,
  I dont think its because of more weight, because according to design calculations she should be able to lift upto 5.8 kgs lets say at least 5 and i guess she is some way near 2.5.

 Yeah that,s what. Dalt is mostnof the time set to particular value and if you notice in the second flight DAlt is a perfect straight line. But in this case, its dragged around by thr baro. But i did see a thread that was posted 3 years back saying its not a bug, but a design decision taken by the team as if the DAlt cannot be maintained the FC will itself take a decision to drag it to a possible height. But still the question remains what made it take the decision?

Yeah, i am waiting for a day with good and stable weather. For the past 3-4 days it raining continuously Cry

What made you think the throttle is way to high?...and by the way mavlink did report ekf error mid flight many a times
Logged
 

Read
« Reply #3 on: July 11, 2018, 05:46:36 AM »
kn8alpha
Plane Lover
Senior Member
***

Reputation Power: 2 
kn8alpha has no influence.
Offline Offline

City: Bangalore
State: Karnataka
RC Skills: Advanced
Posts: 170
Join Date: Feb, 2016



First and foremost I hope your baro is isolated from any sources of light and the altitude gets tracked well in MP.

APM altitude controller sends an initial throttle value which it thinks can counter the weight of the quad. Check whether this is the same all the time and also if it's sufficient. Don't remember if there's a param for this, but check it out.

From telemetry logs check that you don't have excessive vibrations. I don't remember the exact param name but APM docs has a page on that. Z axis vibes can lead to ekf error, and that can affect z axis attitude Dalt calculation.

Other parameters to check: Throttle deadzone (too less and you maybe commanding altitude change without even knowing it), PSC_ACCZ_I AND P (Alt hold tuning)

Some pics of your FC placement and video of the behavior would help forum members give you more pointers.

Logged
 

Read
« Reply #4 on: July 17, 2018, 09:17:36 PM »
vaisakh032
Plane Lover
Active Member
**

Reputation Power: 1 
vaisakh032 has no influence.
Offline Offline

City: Thiruvananthapuram
State: Kerala
RC Skills: Intermediate
Posts: 9
Join Date: Feb, 2018



Hey,
 I am very sorry for the late reply, I was out of station for a while, that's what happened.
 tripple checked
Yep the baro is isolated, it is a PixhawkCube that I am using, so that won't be a problem and about the vibrations,... i've triple checked them.

However, i did find the error and now everything is back to perfect.

Error Found:
  In AltHold mode, the throttle should be in between a value of 1400 to 1600. ANd if the throttle value is less than 1400 the copter descends and if its above 1600 it ascends. After cross-plotting all the graphs, digging around to find an anomaly, i found that it was never a an anomaly nor an error. Everything was working as it was meant to be, and its just an error from the pilots side. Hats Off

@kn8alpha, thank you so much for your time. I gues we'll check this the first the next time Roll Eyes.
Logged
 

Read
« Reply #5 on: July 18, 2018, 08:25:13 AM »
ambattuhari
Plane Lover
Senior Member
***

Reputation Power: 2 
ambattuhari has no influence.
Offline Offline

City: Kottayam
State: Kerala
RC Skills: Intermediate
Posts: 175
Join Date: Oct, 2017



Hi, as you have posted this in Ardupilot Discuss, I was waiting for somebody there to respond...!

Automatic Learning of Hover Throttle
Copter-3.4 (and higher) includes automatic learning of mid throttle (aka “Hover Throttle”). The MOT_THST_HOVER value will slowly move towards the average motor output whenever the vehicle is holding a steady hover in non-manual flight modes (i.e. all modes except Stabilize and Acro).

If you wish to manually set the MOT_THST_HOVER value, it is best to download a dataflash log and set the value to what is seen in the CTUN.ThO field. The value should be between 0.2 and 0.8.

If for some reason you wish to disable learning, you can set the MOT_HOVER_LEARN parameter to 0.

  I don't think it is "Pilot error" !. Whenever you change weight, or even different batteries , the hover throttle changes.
 Flying in Alt hold mode for 30 sec will let it learn & Save the new value.
 Happy Flying.

Logged
 

Pages: [1]   Go Up
Jump to: