Beta Testers requested to fix Compass Issue

[ianwood]

Back and forth 3 or 4 times. By the last run there was an increase of hook, about 4 or 5mtrs, but direction was maintained very well.

Tried IOC. Forward stick applied and it immediately started drifting off about 30 degrees to the left.
Tried HL and got similar results. Wouldn't maintain a straight path.

30 degrees is less or equal to what you saw on the non-beta?
Strange that I am not seeing CL issues as much as others. xrover, what are your CL results? Sorry if I didn't see them before.

And you saw an increase in the hook later in the flight? That's no good.

In GPS mode I set out and observed, as before, a 20deg or so off heading track.

Was this in course lock?

As for the un-commanded yaw, I reviewed my video from yesterday and it looks like it might be mostly limited to throttle, i.e. ascending or descending. And I reviewed some older video with 3.02 and the H3-3D and did notice some slight but maybe lesser un-commanded yaw when descending and sometimes when ascending. Wondering if maybe it's not related to 3.05 but possibly 3.02+ and the H3-3D gimbal.

 

[ianwood]

At least two others with the P2, H3-3D and earlier firmware are seeing some yaw instability similar to the un-commanded yaw we've been talking about here with 3.05. That may be related to the H3-3D and not 3.05. Hopefully some more non-beta users will come forward to confirm it.

 

[Sanababit]

At least two others with the P2, H3-3D and earlier firmware are seeing some yaw instability similar to the un-commanded yaw we've been talking about here with 3.05. That may be related to the H3-3D and not 3.05. Hopefully some more non-beta users will come forward to confirm it.

I am experiencing that on firmware 2.0, phantom 2, h32d gimbal, mini iosd, boscam transmiter, it started 3 days ago

Sana

 

[John Shaw]

I have stock Phantom 2 Vision and have not seen any uncommanded yaw or yaw instability - none that I noticed. I did a hover test specifically to look for it - about 17 minutes just hovering in light gusts - no yaw at all. I planned to do more with movements but the rains set in.

 

[John Shaw]

I finally made the effort to set up a Youtube account and played around with very basic video processing.
I have uploaded a video I made on my second day of testing 3.05 and commented on earlier in this thread (entries for both first day and second day). I have captioned this video to explain what is going on and basic observations. A few times on the return legs I induced some wandering flight.
There are...
Three runs with IOC off
Two runs with Course Lock using the course set at takeoff
I reset course in flight
Three more runs with Course Lock

In addition I don't see any uncommanded yaw or strange altitude changes

My previous testing was on an adjacent field and runs were to the south. These runs are headed to the west. I do think it worked out well that running west in the morning put the Phantom's shadow in view and I think adds to the "immersion".

So give it a try
http://youtu.be/wHQqBNTa8tM

 

[John Shaw]

I finally made the effort to set up a Youtube account and played around with very basic video processing.
I have uploaded a video I made on my second day of testing 3.05 and commented on earlier in this thread (entries for both first day and second day). I have captioned this video to explain what is going on and basic observations. A few times on the return legs I induced some wandering flight.
There are...
Three runs with IOC off
Two runs with Course Lock using the course set at takeoff
I reset course in flight
Three more runs with Course Lock

In addition I don't see any uncommanded yaw or strange altitude changes

My previous testing was on an adjacent field and runs were to the south. These runs are headed to the west. I do think it worked out well that running west in the morning put the Phantom's shadow in view and I think adds to the "immersion".

So give it a try
http://youtu.be/wHQqBNTa8tM

 

[Andy T]

I finally made the effort to set up a Youtube account and played around with very basic video processing.
I have uploaded a video I made on my second day of testing 3.05 and commented on earlier in this thread (entries for both first day and second day). I have captioned this video to explain what is going on and basic observations. A few times on the return legs I induced some wandering flight.
There are...
Three runs with IOC off
Two runs with Course Lock using the course set at takeoff
I reset course in flight
Three more runs with Course Lock

In addition I don't see any uncommanded yaw or strange altitude changes

My previous testing was on an adjacent field and runs were to the south. These runs are headed to the west. I do think it worked out well that running west in the morning put the Phantom's shadow in view and I think adds to the "immersion".

So give it a try
http://youtu.be/wHQqBNTa8tM

excellent demo!

 

[wkf94025]

I reset course in flight
Three more runs with Course Lock

John,

Helpful video for those of us not running 3.05. However I am confused about one aspect of your testing. What exactly does "with course set in flight" mean? Did you reset home position before each of those runs? Or reset course lock vector at the end of a previous run? Or... ?

Sorry to be slow,
Kelly

 

[nzvideoguys]

30 degrees is less or equal to what you saw on the non-beta?
Strange that I am not seeing CL issues as much as others. xrover, what are your CL results? Sorry if I didn't see them before.

And you saw an increase in the hook later in the flight? That's no good.

I never used the IOC before installing 3.05 so I don't know if there was any difference. Another problem I was getting with 3.04 is that when I yawed the P2 around, either left or right, and this is giving only yaw, it would drift off, sometimes as much as 5 to 10mtrs, before slowly coming back to the start point. I found it very disconcerting trying to do a rapid yaw, esp if I was near trees etc because I couldn't predict where it was going to float off to.

 

[Xrover]

I downloaded an app that reads Magnetic interference and gives me a (uT) Magnitude reading at the current position where I am holding the phone.
What number would be a threshold to indicate that it can start influencing the Compass on the P2 (non Vision)

Thanks

 

[John Shaw]

I reset course in flight
Three more runs with Course Lock

John,

Helpful video for those of us not running 3.05. However I am confused about one aspect of your testing. What exactly does "with course set in flight" mean? Did you reset home position before each of those runs? Or reset course lock vector at the end of a previous run? Or... ?

Sorry to be slow,
Kelly

Kelly
I am using NAZA mode and IOC as needed.
At takeoff, the Phantom automatically sets a home point and a course aligned with the way it is facing on the ground.
The first three runs have the IOC off but at that point I select the IOC Course Lock with the left switch. The two Course Lock flights before the caption says "with course set in flight" use that course, the course set at takeoff.
When you see that caption I am toggling the left switch from the top to middle position about five times, this causes the Phantom to reset the locked course but not the home point.
The last three flights use the course locked while in flight. The motivation is that the initial course was wrong due to the declination problem but by the time I reset the course in flight the Phantom has figured that out. So the last two runs don't have a course error. I think this is the clearest demonstration of the declination problem.
If I had wanted to reset the home position in flight I would have had to toggle the left switch from the middle to the lower position about 5 times, but I didn't need that for this test.
Hope that helps.

 

[sar104]

I downloaded an app that reads Magnetic interference and gives me a (uT) Magnitude reading at the current position where I am holding the phone.
What number would be a threshold to indicate that it can start influencing the Compass on the P2 (non Vision)

Thanks

I use a similar method to check for magnetic interference, and what I look for is inconsistent or varying results in the vicinity. Typically you can zero out the reading and then display differential values.

 

[wkf94025]

John,

Thanks for clarifying. Your methodology is what I was assuming.

Kelly

 

[ianwood]

I have stock Phantom 2 Vision and have not seen any uncommanded yaw or yaw instability - none that I noticed. I did a hover test specifically to look for it - about 17 minutes just hovering in light gusts - no yaw at all. I planned to do more with movements but the rains set in.

This is consistent with my suspicion that the un-commanded yaw we are are seeing is related to the H3-3D and not specific to 3.05.

Your video wins the award for best demonstration so far. Using the shadow is a great idea. Your results are very consistent with mine. Your course lock is also consistent. Much better but still off. I noticed that it was hard to get lined up perfectly with the line on the ground. I think that is mostly result of judging when it's square as opposed to the compass causing a lateral drift. I struggle with this when filming.

Sar104, what are your thoughts? Your results still seem different. Perception or is something acting differently?

I use a similar method to check for magnetic interference, and what I look for is inconsistent or varying results in the vicinity. Typically you can zero out the reading and then display differential values.

I do the same. I use an app that shows a histogram for the three compass axes. Use it to move around slowly and watch how it changes. Move it closer and further away to suspect areas and look for changes.

 

[John Shaw]

Ian, I totally agree with your comment of how hard it is to get lined up exactly. I also feel the slight misalignment shown in the later flights is more likely due to not getting the Phantom perfectly aligned with the field. Most of the longer pauses before starting a run are as I struggle with that.

 

[Xrover]

Any update on the 3.05 Beta 2 firmware yet?

 

[sar104]

John's results are approximately what I expected to see in my own tests, based on the early beta test results. What is the declination at that location? Mine, clearly, were quite different and much worse, but I can't see what I did differently - they were basically the same tests. While I agree about the problems of lining up the aircraft, especially once airborne, I think we are talking about a couple of degrees as illustrated in John's video, not the 15 - 20 degrees that I saw.

I'm really quite baffled as to why we are seeing such diverse results in such nominally similar tests. Perhaps we do all need to list the gains that we are using and explore whether there are any other uncontrolled, significant variables (e.g. startup orientation). I did not get to test last weekend due to high winds, and I'm on travel until Friday, but this weekend I should have more time and will do another set of test flights, including installing 3.05 on my other P2 and adding a GPS logger to it to generate accurate tracks.

 

[ianwood]

Any update on the 3.05 Beta 2 firmware yet?

None yet but we only found out about it two business days ago so I would be surprised if it hatched that fast.

Sar104, I agree we should share gains. In fact we should all share any settings we think could influence results:

Mounted hardware (e.g. FatShark 250, OSDII, etc.):
Gains:
Gimbal settings (if applicable):
Mod values:
All firmware version numbers:
Phantom Assistant software version number:

My P2 didn't travel with me this week so will need to wait to Friday to post mine.

 

[Dadcat]

John's results are approximately what I expected to see in my own tests, based on the early beta test results. What is the declination at that location? Mine, clearly, were quite different and much worse, but I can't see what I did differently - they were basically the same tests.

How about everybody also put their general location in their profile. I can't remember where everyone is which makes it harder to follow what's going on. Also we won't have to ask what people's declination is all the time. If ever there were threads where knowing location is important to know, these are it. Only takes about 10 seconds.

 

[John Shaw]

John's results are approximately what I expected to see in my own tests, based on the early beta test results. What is the declination at that location? Mine, clearly, were quite different and much worse, but I can't see what I did differently - they were basically the same tests. While I agree about the problems of lining up the aircraft, especially once airborne, I think we are talking about a couple of degrees as illustrated in John's video, not the 15 - 20 degrees that I saw.

I'm really quite baffled as to why we are seeing such diverse results in such nominally similar tests. Perhaps we do all need to list the gains that we are using and explore whether there are any other uncontrolled, significant variables (e.g. startup orientation). I did not get to test last weekend due to high winds, and I'm on travel until Friday, but this weekend I should have more time and will do another set of test flights, including installing 3.05 on my other P2 and adding a GPS logger to it to generate accurate tracks.

My Declination is 17 degrees (US Seattle Area)
For the tests in the video you can see I just start up, takeoff and go right into testing. I did that because we know it learns and I wanted to get a look right from the start and watch it through the entire learning process. The entire video is one take, n0 cuts, so you see the whole thing - all I did was add the captions. I was impressed that there was a little movement but no TBE on takeoff, on the first runs return the JHook was gone. Course Lock has a significant error until reset in flight and that is an annoying shortcoming.

As previously posted...
I have never updated my gains. (at least nothing I am aware of)
Stock P2V
...............Pitch....Roll....Yaw....Vertical
Basic........120......120.....120....120
Attitude....200......200

I have never seen anything that motivated me to want to change them and I am happy with the current response.
I don't think I've ever felt there was a noticeable change except for the slower descent rate which has obviously changed.