Naza V1 - gain tuning issue

B

Bigandy

Member
Hi folks,

I wonder if I could bounce a few ideas of everyone here. I've built a quadcopter, and I'm having a few issues with tuning the gains. I've set-up approx 15 quadcopters so far using Naza flight controllers, with no issues at all. This one is giving me some greif though.

The quad I have built is nice and rigid, and is around 900mm span. It uses Axi 2820/18 motors, and Gemfan 14x4.5 wooden props. Speed controllers are DJI 30A opto ones. The flight battery is a 4s, 9000Mah pack (actually two packs in parallel). The airframe is all carbon fibre, so fairly light, but very strong indeed. The flight controller is a Naza v1 on the latest firmware.

The problem I am having, is setting the basic gains. For some reason, I cannot get them dialled in well enough to eliminate any wobble. I have the basic pitch and roll (and yaw) gains set on a series of switches on my TX, so that I can increment the gains in 2% steps (or decrement!). I've tried over about 5 flights now to get them dialled in, but I can't seem to get rid of the wobbles. With excessively low gains, I get hard to control slow oscillations. With very high gains, I get fast wobbles. In between, in flat calm weather, I can get it sort of stable, but the moment a small gust of wind comes (or I try doing anything other than a hover) the oscillations return. My knowledge of tuning smaller craft suggests that the wobbles are fast enough to lead me to reducing the gains. This then introduces the low frequency wobbles, and sluggish control response.

Has anyone got any ideas? I'm half wondering if the flight controller is damaged (it is a V1 unit bought about 18 months ago, but was working fine on the TBS discovery it was removed from).

I'm about to resort to installing a KK2.1 board to try out, to see if it is an airframe (or motor/esc) issue, or a tuning issue.

Cheers
Andy
 

PeteDee

PeteDee

Mr take no prisoners!
I wonder if the setup you have built is close or just past what the Naza can handle, the earlier firmware was not very good on bigger than 450 sized quads.

What are the settings you are starting with? I would be thinking to go to high gains to start with, 180-200.

Pete
 

B

Bigandy

Member
I'm wondering the same thing. I've actually loaded the latest firmware (4.02), and the problem is still exactly the same.

I started with the naza default gains settings (100) but the closest I have got to it being stable has the basic pitch and roll gains at 140, and the yaw gain at 120. The vertical gain is at 100 (haven't touched that one yet), and the atti gains are also at 100 (haven't touched those either). I have tried the gains up as high as 250 on the pitch and roll, and it just results in a fast oscilliation, that gets worse the higher the gain goes. I have tried the gains down as low as 60 on the pitch and roll, and this is pretty much uncontrolable.

Even with the gains where I have them now, there is still a horrible oscillation when the hover is disturbed (either by me on the sticks, or a gust of wind). It isn't the low frequency (gain too low) oscillation, but it is slower than the "gain too high" oscillation. It's almost as if the controller cannot keep up with the airframe, and is lagging a little behind what it is doing.

I think I'm resigned to trying out a different flight controller, to see if it is the Naza being pushed beyond its limits on a larger than normal platform. I'm half tempted to get another 450 kit built to see if the controller works on the smaller airframe....

Cheers
Andy
 

LeeT

LeeT

Wannabe AP Dude
DJI recommends not going past 650 on the Naza. Agree you are past the limits of it.
 


B

Bigandy

Member
Have you got a reference for the 650 dimension LeeT? I couldn't find anything in the instruction manuals online.

Would adding a payload have a similar effect to using smaller props? I.e. increased payload = higher hover rpm values? I am currently flying with a 4s 4800mAh pack, but I can put dual packs on in parallel, and also add a mass to represent my Nex5 camera that is intended to go on there too.

Cheers
Andy
 

LeeT

LeeT

Wannabe AP Dude
See page 6 of the manual, the fine print below the multirotor configurations.

Note: The NAZA-M V2 flight control system doesn't support Gimbal function when used on the Octo-rotor aircraft. For big aircraft that is larger than 650 or with heavy load, WKM is recommended.
 

A

anwar3000

New Member
Hai everyone
***** Urgent help required ******
i wonder if someone can help me with perfect gain for my TBS Discovery Pro

TBS Discovery pro
DJI Naza V2 + GPS + iosd mini
DJI E300 prop system (E300 2212 920Kw motor + 20 ESC + 9449 propeller) DJI tuned system
4s 4500mah battery
1650 gram total weight

I tried with many combination, my main concern is when i initiate fast yaw the qud became unstable and it reduces its height
and trend to uncontrolled.

current gain
pitch - 160
roll - 120
Yaw - 180
Vertical - 180

attitude - 100
Roll - 100
( not good for my setup)

earlier i used the following setting for my DJI F450 it was perfect and i was almost happy
125/125/160/160 - 125/125
 




A

anwar3000

New Member
Tahoe Ed said:
The Discovery is not that stable in yaw because of the arm placement. That is why I sold mine.
Click to expand...

Hai Tahoe

thanks for your replay

i faced terrible yaw issue with dji f450 with naza v2, I contacted DJI support staff they said it might be the problem with naza V2 so send it back to dealer. I was wondering why? I got couple of accidents as well with dji f450
one time almost flyaway.
finally I found its the problem with underpowered E300, it does not have sufficient lift power while yaw
ie, while making yaw two motors are almost no spinning (diagonal opposite motors spin@ very low rpm) that is not sufficient to lift the quad in air; hence it is dropping its altitude. Finally I found a good gain level and changed the battery to 4s the problem was solved to an extend but not eliminated - you can search in forums "naza left drift"
initially the problem was too bad because I sticked with DJI recommendation gain level and battery (3s).
See the dji f450 video - http://youtu.be/3UrJPw9XnzM
that's why I shifted to tbs disco
#why this is happening in left direction only? Need to suspect the DJI naza algorithm (no answer)
i changed all my motors ESCs, my CG was at pin point center

tbs discovery is more stable in air than dji f450 with their recommended gain level
i think if I fine tune tbs I may get a very stable flight, that's why
otherwise I suppose to go for Pixhawk... I know several Naza seems very stable in air, but several not

anwar
 

You must log in or register to reply here.
Top