Naze 32 board, random yaw issues.

[JJBTEXAS]

Hey guys, ran into a problem with my new Acro Naze32 board and was hoping y'all could help.

Quad:
Acro Naze32
Multistar Elite 2204-2300KV Multi-Rotor Motor
Q Brain 4 x 20A Brushless Quadcopter ESC
Gemfan 5040 Multirotor Propellers
The Frame is of my own design but is very similar to the electrohub racer built in the video (including the dihedral).

The Problem:
When flying the quad will randomly yaw CCW quite frequently. Sometimes its quick and short, sometimes it will last several seconds and turn all the way around and sometimes it doesn't happen. This happens in every mode (angle, horizon, head lock, etc...). The quad doesn't yaw CW very well either.

Data:
I flew the quad while still plugged into the computer. The software showed the quad rotating accurately and the accelerometers showed the yaw. I was also able to determine the receiver wasn't receiving any random yaw inputs.

Attempted Solutions:
I have calibrated the speed controllers individually and all together as done in the Flite Test video to no avail. I have flow this same setup with a KK2 board and did not have these issues (this has been retested and still holds true).

Question:
Any ideas?

Thanks,
JJBTEXAS

 

[severed]

Could be several things. Could be a motor issue. I've read issues with the Multistar motors. Also, it sounds dumb, but make sure your props are installed in the correct direction.

 

[severed]

Also make sure your motors and ESCs are calibrated properly in Baseflight/Cleanflight.

 

[JJBTEXAS]

Could be several things. Could be a motor issue. I've read issues with the Multistar motors. Also, it sounds dumb, but make sure your props are installed in the correct direction.

Because I've flown this setup with no issues with my KK2 board, I Dont believe its the motors or speed controller. The props were also checked as that was the first thing I came up with.

Also make sure your motors and ESCs are calibrated properly in Baseflight/Cleanflight.

I calibrated the ESCs as they do in the Naze32 video in clean flight which I assume is the proper way to do it.

 

[C0d3M0nk3y]

Do you get the same behavior in rate mode?

Are you trying to fly in modes like mag and head lock on an acro board with no sensors for those, or was that just a test?

 

[Mustang7302]

I imagine having mag or head lock modes turned on will cause some erratic behavior given the acro board doesn't have a magnetometer. Disable all flight modes, except for Angle if you need auto-level, and give it a go again.

 

[JJBTEXAS]

Do you get the same behavior in rate mode?

Are you trying to fly in modes like mag and head lock on an acro board with no sensors for those, or was that just a test?

I imagine having mag or head lock modes turned on will cause some erratic behavior given the acro board doesn't have a magnetometer. Disable all flight modes, except for Angle if you need auto-level, and give it a go again.

It was my understanding that head lock works with Acro boards by using the accelerometers which is less accurate, but works. This happened in rate mode as well. I also didn't spend much time in rate mode once I confirmed the yaw was still happening. I've been running most of these tests in a big empty room due to weather constraints unfortunately.

I have noticed that dialing in quite a bit of trim (+20 on my DX6) helps initially but after flying for a minute or so the tinging will suddenly start spinning CCW. After 30 more second it will have turned competently around, sometimes a full 360. The spinning is not constant either. It will spin to 90 then stop, a second later it will spin 45 degrees slowly, then spin quickly another 90 degrees or so, with no consistency it the spinning frequency, speed or turning distance (time, theta and omega are all inconsistent). I'm going to switch back to the KK2 again to see to ensure the Naze is the problem. At this point, I'm just assuming something must be wrong with the board.

 

[cranialrectosis]

Do any of the motors get hot?

What are your PID values?

 

[JJBTEXAS]

I just tested her out with the KK2 board again and noticed a few things.
-no random yawing during through an entire 1300mAh pack of continuous flight
-the KK2 board was much easier to fly overall

This is strange because all reports show the Naze should be a more stable board but I didn't find that the case. Neither board has had any tuning so maybe the settings out of the box for the KK were just better suited to this quad but I find that unlikely to make that big of a difference

Do any of the motors get hot?

What are your PID values?

The motors were all evenly warm after the flight, next time I'll measure them properly. I used the defaults as FliteTest said that was the best place to start.

 

[severed]

Did you adjust the motor number orientation?

 

[JJBTEXAS]

Did you adjust the motor number orientation?

I made sure the motors numbers matched the diagram in clean flight if that's what you are asking

 

[severed]

I made sure the motors numbers matched the diagram in clean flight if that's what you are asking
View attachment 41041

Yeah.

 

[jhitesma]

On MW/BF/CF headlock isn't coupled to the flight mode it's in addition. So even in rate mode you could have headlock enabled and with no mag that could be causing the issue. Do you have headlock fully disabled in all modes? I'm pretty sure headlock only works with mag, the code tries to hold heading already without it the mag just attempts to enhance that hold but has to be enabled.

So I'd say make sure mag and headlock are fully disabled in all modes.

 

[JJBTEXAS]

On MW/BF/CF headlock isn't coupled to the flight mode it's in addition. So even in rate mode you could have headlock enabled and with no mag that could be causing the issue. Do you have headlock fully disabled in all modes? I'm pretty sure headlock only works with mag, the code tries to hold heading already without it the mag just attempts to enhance that hold but has to be enabled.

So I'd say make sure mag and headlock are fully disabled in all modes.

Once I confirmed that the issue continued with headlock on I removed it. The problem appears to get worse the longer the quad is flying. I believe the accelerometer might just be getting a false yaw reading for some reason whether it be a faulty unit or bad vibrations. I'm going to calibrate everything again and have the quad sit motionless for half an hour or so and see if the software records any extreme changes in yaw position. If it does then the accelerometer is probably to blame. If that doesn't net any results then I'll try resetting the calibration in clean flight, fly the quad, then plug it back in in the original orientation and see if I can gleam anything from that.

Update:
Just ran the second test with no real results. The orientation was only slightly of CCW while the quad managed two full rotation after about 2 mins of flight time. Trying the first test now but I'm doubtful much will come of it. I might try switching to baseflight. I've already tried cleanflight 1.7.1 and 1.7.2.

 

[JJBTEXAS]

Update 2:
Running the first test now and finally getting somewhere. The cleanflight software shows the quad slowly rotating at roughly 1 degree every 4 seconds. Working on setting up baseflight now.

Update 3:
Baseflight gives me similar results. I'm fresh out of ideas.

 

[brizrobbo]

I have a similar problem on my naze32 setup. It's my first one, so I am really learning as I go ....but one thing I learned after watching the FT episode on setting up the naze32 is to have your flight controller positioned as close to the CoG as possible. Mine is not, so I am wondering if that could a contributing factor.

I'm about to start troubleshooting my setup, so I will post any discoveries.

 

[jhitesma]

accelerators don't get used in rate mode - only gryos. So it could be a faulty gyro. That's also why it's best to debug in rate mode with nothing else enabled - that way you're only dealing with one sensor (well, technically 3 but they're all in the same package just pointing different directions.)


You could try doing a manual calibration on your gyro - you are letting it sit still while you power it up correct? (It attempts to calibrate the gyros as soon as power is applied so it has to sit still for a few seconds.) You can do a manual gyro cal by letting it sit still and assuming you fly in mode2 it's right stick down and centered and left stick down and to the left. HOld that for a second and you should see an LED blink in confirmation.

 

[JJBTEXAS]

gave the manual setup a try and it seemed to help a little, but the problem still exist. It really seemed to just delay the inevitable. The longer the flight times are the worse the problem gets. This seems like an error stacking issue to me at this point. I sent getFPV an email to see if they can get me a new board or something along those lines.

 

[nilsen]

I had a board from getFPV and it was great but unfortunately I fried it by pluggung in the VBat lines in the wrong order so I switched it out with a naze32 from hobbyking.
What I noticed is when it's connected to cleanflight with no battery connected the onscreen model sits perfectly still, but as soon as I connect the flight battery it slowly rotates clockwise on the screen. In flight I notice this as well. It's very slow and controllable but annoying. What made my craft flyable is using PID controller 2 and it's a lot better now (still a slight rotation but the controller seems to iron it out).

 

[RC-Flyer63]

I has a similar thing happen to my 250 quad, I am running a Naze32 acro board in it with 1804 ZMR motors. When I built it the motor mounting bolts were just barely long enough to attach the motors but i thought it would be good enough so I proceeded to finish the build and start flying. Anyway I had several flights LOS on it and during one it yawed to the left 90* with no input, I remember thinking that was odd but it did no further unusual movements so I kept flying. the next flight I began to FPV it and I took off flew about 100 feet and it suddenly yawed 90* left again, I stopped in the air and began to think about what had happened when the image in my goggles began to flip, ground, sky, ground, sky, grass, you get the picture? LOL Anyhow I told my wife, who was watching me at the time, I'm down. She saw where it hit the ground so we walked over to survey the damage and it turned out that one of the motors on the right had come loose and had come off the arm. I reallise this is a long story but the point of it all is I think the motor had loosened and was maybe causing a vibration in the frame that confused the gyro, I have since repaired the damage, replaced all the motor bolts with longer ones and after many more flights have not had any more problems of a similar nature. So you may want to look at all the mounting points and recheck your prop and motor balance. By the way my FC is attached to the frame with solid standoffs so it would have felt even a slight vibration. Just my thoughts and experiences.