AQ6 stable flight without gps

[johannes3154]

Hello Guys,
i built a quad, based on the Tank Quad. Basically the HL Frame from flyduino.net with 25cm CF arms. I used kiss escs and T-Motor MN3110 700kv with T-Motor CF-Propellers (rudimentary balanced ).
The AQ6 stands on stiff nylon standoffs and the escs are placed in the prop wash.
The whole system is driven by 2 4s 3300 mah lipos which give a AUW of 1900g. (hovering at 50% throttle)
My radio is a FrsKy Taranis with a homebuilt S-Bus converter.
This combination flies very well in my opinion, without a connected gps antenna. As soon as i connect a gps antenna and it has a 3d fix, it completly goes crazy. All operations are done under the normal manual flight mode.
I did a tare and a onboard magnetic calibration. (I think the result is within the tolerance)
The attached log contains data of 3 flights. First flight was done with connected gps but prior 3d fix, the 2nd and 3rd were cancelled takeoff tries, because of an unflyable state.
If the Problem occurs, the motors sound different (higher frequency of speed changes) but uncontrolled and without system. The quad seems to flip the next secound. To prevent this i land immediately but its in a highly unstable state then.
If there is no gps fix i can fly very stable even in very windy conditions (manual height control). I have to admit that my garden is probably the worst testing ground, because there are high trees which are causing immense turbulences.^^
Perhaps u can have a look at my logfile and point me into the right direction.
Thanks in advance
Johannes

[LPR]

Johannes

First of all GET OUT OF THE GARDEN you're a menace and you are going to cut down all the vegetables.

Really get away from trees that can block the GPS signals. A 3D lock is just a start of having a good stable GPS signal.

If you have good manual control, a bad GPS signal should not be able to cause your quad to flip. Usually a week signal can make a multicopter go in one direction and than another.

Looking at the log I would guess that you are taking too long to get in the air and by the time the quad is taking off the poor GPS signal has moved the WP that was set when you first went to APH.

If your not comfortable with hovering in manual and switching to APH when raise the throttle till the quad is almost off the ground and switch to APH while slowly raising the throttle. If you have a good GPS signal you will not have any problem.

If your signal is poor and you take a few seconds getting in the air, bad things will happen.

Take your computer out to the garden and connect it to the AQ board. Watch the telemetry for the value of hAcc. When the value of hAcc gets down to 3 your AQ board has a 3D fix. A good signal is when the value of hAcc is less then 1 and it can go lower then .5 and that will take a few more minutes if you're in an area that has a clear view of the sky.

Larry

[johannes3154]

[aBUGSworstnightmare]

Hi,
which GPS antenna do you use (please let me know exact part number / source where you got it from is no is not known). also please post a picture how you've mounted it.
As Larry pointed out already: your GPS signal is very poor.
And also please post a video of the motor reaction when it 'goes crazy'.
Have a look at this video:

Listen to what happens at 1:07min and starting from 1:38. Hopefully it doesn't sound like this (cause I've only noticed these spikes on a M4 so far).
As far as I know they are related (somehow) to a bad GPS_VACC. Never noticed them in flight 'cause the craft flies excellent.
Joerg

[LPR]

Johonnes

I see I missed the part were you said that you were in manual mode. GPS has no effect on manual mode flying and the compass has no effect.
I would look for something else causing your problem. Did you solder the DIMU yourself?
A few people have had problems with the DIMU and soldering.

Larry

[johannes3154]

First of all, thank u for the quick response:-)
The thing that makes me stumble is, that if i fly in manual mode without gps antenna, the quad flies nice and smooth, as on all those youtube videos^^. I certainly could have less acc noise, but its flying nicely.
If i plug in the gps antenna, everything is normal since there is a gps lock. If there is a gps 3d lock things go crazy very fast. I can even start a normal flight, go up to 2 meters for example and if i get a gps lock the quad starts to drift spin dive and tumble.
But the point is, why does the 3d lock has an effect, i'm still in manual mode?
I have to admit, till now i didnt really care about the gps, because i planned this as the next step after successfully flying manual.
The antenna is the one that is sold with the Autoquad. I started to build the gpsshield around the antenna, but because of those problems i left it to be just a piece copper;-).
I have soldered
@Joerg This video is getting close to the behavior of my quad. I can do some in flight footage tomorrow:-). I used the pids from the tankquad thread;-).

[johannes3154]

pic of my setup
Gps antenna without any signs or letters on it.

[aBUGSworstnightmare]

[LPR]

Johannes

GPS getting a 3D lock while your not in APH should not have any effect. A log from your SD card will be needed to track down your problem.

The only change your should have from a 3D lock in manual mode is that the blue LED goes on.

If you soldered your DIMU I would re-solder all the connection just to be sure it's not the problem.

Larry

[johannes3154]

I will do a new log tomorrw, meanwhile in my first post there is a log from my sdcard. The first sequence in the log shows a normal very short flight, without gps lock, the second and 3rd sequence is with gps 3d lock.
Tormorrow i will do a longer hover flight with disconnected gps and spin the motors while having 3d lock.

I will resolder the dimu thats no problem.