[revolt_randy]

Hi all, new to the forum, seeking a bit of help.

So I built a simple X-quad out of wood, and as one would guess, I've had problems. Quite a bit of problems. Most of this was documented on the OP forums where I had a WIP/build thread and received lots of help from the online community. I'm attaching a pic of the quad to this post. (Not sure how to embed pics in my post, so if someone can clue me in on that, I would be thankful)

Here's a video of the maiden flight (warning: it's not pretty)


It's totally unstable and has crashed a few times. I have to say, I try to be as safety minded as possible, and the crashes occurred on private farm land, away from anyone other than myself.

So I got side tracked on this project, but now I'm back to focusing on this and getting it to fly correctly in a stabilized hover. I'm using the OpenPilot GCS software that I downloaded about a year ago. Rev. 15.02.01 to be exact. Been thru the wizards and set everything up before the first flight. Been thru it again and again.... and again...

Recently, I discovered in the HITL tab of the GCS, there's a display of the controller's sticks. Whenever I apply throttle on my controller. the display shows I'm also adjusting yaw as well, but the display is delayed a bit. I even used a ruler on my transmitter's throttle stick to stop sideways movement of the stick to assure myself I'm not yawing the craft. I was using a Turnigy TGY-i6 radio and it appears to be faulty.  (as a side note, most of the hardware I bought for this was turnigy, and I have had problems with most of it)

Got a FlySky FS-T6 in it now, and have stable yaw while advancing the throttle.  In the GCS, I went into the configuration tab, stabilization settings and upped the P to double the stock settings. It now flies a lot better, but still wanders about instead of just hovering in no wind conditions.

Now looking at vibrations and the cheap props I have. Thinking vibrations from unbalanced props is my problem. I've balanced 4 props and will try them soon.

While reading up on stabilization, I came across this:
http://ardupilot.org/copter/docs/common-measuring-vibration.html#common-measuring-vibration

Is that what the scopes tab in GCS should report to me? I've looked in there, but never seen any data.

I guess my real question is how do I access the CC3D's gryoscope data in the GCS?

Thanks
Randy

[crof2003]

In my LibraPilot software, with the quad connected I go to the Scopes tab and I see the gyro information.  I've attached a screenshot so you can see.  What I did was let the quad lay flay and then picked it up, moved it from side to side a bit, and then set it back down.

The battery does not need to be attached for this to work, only USB.

My FC is a CC3D.

Slightly off-topic: building your own frame with wood? Pretty awesome

[TheOtherCliff]

I watched the video.

It looks like it might be drifting a little that you could correct with "Rotate Virtual", but honestly it looks like it is flying just fine.

What I see looks like inexperience but I could be wrong. 

First thing, get some color change between front and back.  Maybe paint the rear arms red and put red props on the rear and change the ESC color in front.  I usually buy props in green and red sets.  You could even paint and rebalance the props you have.  Second, practice motionless hover, right in front of you and chest high.  Third, don't fly it so far away.  Fourth, practice figure 8's close in.

Quads are like a warm skillet on a sheet of ice.  They don't stay in one place unless you force them to stay.

Maybe I misread the video... 

[revolt_randy]

Thanks crof2003 for your reply.

In OpenPilot GCS 15.02 all I saw was a blank gray display in the scopes view. I've installed LibrePilot 15.09 on my computer and my quad, and now I've got the scopes working, see attached pics.

The first pic is the scopes view with the quad sitting on my bench. The 2nd pic is the quad sitting on my workbench, then I full throttle the motors with no props on. Can anyone tell me if these readings are normal?

Seems like the gyrostate and accelstate increase as the motors are cranked up. What's considered a normal range for the gyrostate and accelstate?

Do I have a vibration problem?

Slightly off-topic: building your own frame with wood? Pretty awesome

Not off-topic at all, title says it's a scratch built wooden quad.....

Wood is easy to work with, without advanced tools. Having said that, I got side tracked on this project while learning CAD software, specifically http://www.freecadweb.org/, an open source CAD project. Then I studied carbon fiber construction.... So I've returned to this project.

I need this quad to be semi-stable before I mess with other more advanced toys..... 

Thanks for the help,
Randy

[f5soh]

Do I have a vibration problem?

Yes

Look at Accel scope, there is spikes around 5 so half gravity.
The green trace (Z) measure the gravity around 9.81 and has also big spikes while motors are running.

Solutions:
- Remove vibrations at source: motors, balance props...
- Add filtering in Attitude > Accel filtering with a value around 0.1, at least not 0 that disables filter.
- Mount board's case using double sided adhesive foam.

[TheOtherCliff]

If after balancing props, you still have a vibration issue, you may need to increase Stabilization -> Expert -> GyroNoiseFIltering

[revolt_randy]

Hey TheOtherCliff, thanks for your replies as well!

What I see looks like inexperience but I could be wrong. 

You are correct, inexperience, been about 25 yrs since I held an r/c transmitter in my hands. I quite playing with r/c cars about the time the first brushless motors came out. Back then, we used ni-cad batteries too.

First thing, get some color change between front and back.

I figured that one out real quick and bought some bulk packs of red & white leds. Drew a wiring diagram to figure out the size resistors I needed (been a few years since I've messed with electronics) and wired them up. It's like a car, white lights in front and red lights in the back. I am currently only flying it about waist height.

Look at Accel scope, there is spikes around 5 so half gravity.
The green trace (Z) measure the gravity around 9.81 and has also big spikes while motors are running.

I noticed that in the Accel scope, but didn't understand what I was seeing. Now that you mentioned gravity, I noticed units of measurements and understand what your saying. What I was more concerned about was the gyro readings.

Where can I find in depth documentation on the scopes view? What I've found for documentation really lacks sustenance. A simple description of the units of measure would be nice.

I'll try the software solutions both of you have presented first. See what that does, as it's easier than balancing motors, and report back when I have something.

Thanks for all the help so far!

Randy
 

[f5soh]

Did you look here ?

https://librepilot.atlassian.net/wiki/display/LPDOC/Scope+Tab

A simple description of the units of measure would be nice.

Like that ? Ok, looking to something more dense: https://en.wikipedia.org/wiki/Metre_per_second_squared

[revolt_randy]

As to documentation on the scopes tab....

Did you look here ?

https://librepilot.atlassian.net/wiki/display/LPDOC/Scope+Tab

Yes, I did look there when I first realized the scopes weren't working. And indeed, by looking a bit closer, I think I've figured out all the data of the default display. But perhaps a paragraph spent explaining what each scope is displaying and units of measure would be helpful to newer users. Anyhow, I'm not here to nag about documentation.

I replaced 2 broken arms on this quad. 5/8 inch square dowel from lowes, maybe $2.00 and my time cutting, drilling, etc...

I did look at this:

- Remove vibrations at source: motors, balance props...
- Add filtering in Attitude > Accel filtering with a value around 0.1, at least not 0 that disables filter.

I'm assuming changes made in Accel filtering won't been seen in the scopes view? The scopes view is raw data from the sensors, and not the end result of adjustments to the Accel filtering? Would I only notice a difference while flying it and not in the scopes view?

And:

If after balancing props, you still have a vibration issue, you may need to increase Stabilization -> Expert -> GyroNoiseFIltering

Would I only see the effect of that in flight and not in the scopes tab? I've not tried flying it after repairing the damage.

Instead, I did play with balancing the motors, but I need to do more work in this area. I used the scopes to view data running one motor at a time, using trial and error to find the spot the motor needed weight on, and used tape to weight the motor. The results weren't really all that good.

I added tape to what I found to be the light side of the motors and reduced the Accel values to half of what I started with. Overall, with all 4 motors spinning at full throttle, the Accel scope dropped from a spike of 5 to about 4.

Is this a good method of balancing motors?

One of the motors was really smooth, needing no work and I replaced a motor recently. So I assume the one smooth motor is the new one.

Right now, I'm thinking I should buy 3 new motors and replace the vibration causing motors.  I'm also looking at ways to cushion the FC. Thicker foam tape to mount it and maybe foam mounting the board that has the FC on it.

Any advice is welcomed!

Randy

[revolt_randy]

OK, so I upgraded from OP 15.02.01 to LibrePilot 15.09 and got the scopes tab to working, but now I have a different problem. No welcome screen when LibrePilot GCS is started, see attached image. The welcome screen always worked with OP 15.02.01. This is on my laptop, and I'm wondering if the display problems I'm having is because of the GPU on the laptop. My laptop's specs are:

Dell Latitude D630
Intel Core2 Duo CPU @ 2.2 GHz
2 GB ram
Mobile Intel 965 Express Chipset Family display adapter

I don't know if this is a bug in LP and I'll have to try this out on my desktop which has a nividia GPU.

Since my last post, I erased my CC3D and went thru the setup process again to ensure all values and settings were reset to the original settings. It flies decent without much adjustment to gyro & accel filtering. Yes, it still has plenty of vibrations which I will address shortly.

Right now, the main thing is the craft wants to tilt to the left on take off.


I was reading thru another post, and will follow the advice here:
https://forum.librepilot.org/index.php?topic=2171.msg15892#msg15892
to try to correct the problem.

Any other advice anyone can give will be useful.

Randy

[TheOtherCliff]

Try with the left side facing down hill.  Does it always drift left or always uphill (to the right this time)?

You are taking off correctly, but you should be aware that it winds up the "lower" motors in the time between motor start and take off.  It does this because it tries to make it level, but it doesn't have enough power to lift the bottom side yet.  So the lower motors are spinning faster (more correctly are even being commanded to spin faster because of PID "I" term windup) at lift off and it goes up hill.

If this is the case, then the solution is to take off as quickly as possible after starting the motors or make a level platform to take off from.