Beta Testers requested to fix Compass Issue

[N017RW]

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.

Done, and Done.

 

[Xrover]

here's my settings and info

Mounted hardware:
Boscam TS832 Video TX with Blue beam ultra Antenna, iosd mini. Zenmuse H3-3D Gimbal W/Plastic Spacer Plate and using Grey Absorber Balls, DJI Stock Props, Aftermarket Tall Landing Skid.
Gains Settings:
Basic: 125% 125% 140% 160%
Atti: 260% 260%
MOD Values:
Checked with P2 on floor center of room.
Gyro= fluctuates from 0.1- 0.6
Accel= 1.00
Compass (raw)= fluctuates 1468-1481
Gimbal Config:
TCG=5
HL=0
VL=-3
PA Software ver. 3.0
HW/FW info:
Main Controller Loader 29.0.2.1 Hardware ID 1000726528 Firmware 3.5
GPS Loader ---- Hardware ID 6.1.1.2 Firmware 6.0.1.4
Receiver Phantom Loader 15.0.0.6 Hardware ID 0400025859 Firmware 1.0.0.0
P330CB Loader 26.0.0.0 Hardware ID 02F0069444 Firmware 1.0.1.30
Gimbal CMU Loader 1.0.0.0 Hardware ID 300020309 Firmware 1.0.0.0
Battery Loader 1.5.0.0 Hardware ID 000002901 Firmware 2.0.0.0
H3-3D IMU Loader 0.0.0.8 Hardware ID 300020309 Firmware 1.0.0.6

That should be enough info to Hack into my Sparrow

 

[ianwood]

Xrover, that's perfect. Others, please follow xrover's format. I don't have my P2 with me but will post up Friday. Need all beta testers to post on this.

Also, any beta testers who haven't posted recently should also add any comments on the observations from the past few pages of observations especially what is consistent/inconsistent with your experience. We need to keep it consistent.

 

[John Shaw]

I am not sure where to get all that "hacking information". But here is my best shot...
with a stock P2V I may not have all that gimbal stuff.

Mounted hardware:
Stock P2V Nothing added except a few stickers.

Gains Settings:
Basic: 120% 120% 120% 120%
Atti: 200% 200%

MOD Values:
Checked with P2 on floor center of room.
Gyro= fluctuates from 0.1- 0.5
Accel= 1.00
Compass (raw)= fluctuates 1375-1385

DJI FC200 Camera V1.1.9
5.8GHz Remote Controller V1.0.2.22
5.8GHz Receiver V1.0.1.3
Android App V1.0.40
Phantom 2 Vision Autopilot System V3.05
Phantom 2 Vision Central Board V1.0.1.27
PC Assistant Software V3.0

I certainly agree that we have to pay attention to gains and added features/capabilities/weight.
Gains are there to adjust the Phantom's reaction to control inputs so they certainly will have a significant effect on things like sluggishness, over reaction, stability, ability to hold attitude etc.
Additional weight increases propeller downwash which may effect the baro sensor. It will also effect altitude response.

 

[Xrover]

I shot a video yesterday for a job.
This is what i encountered using 3.05 with everything i stated above.
I was flying in GPS mode. There was 10 mph variable winds and Atti mode wasn't going to be steady enough.

I was getting an approximate 10 degree uncommanded yaw to the left. Both in forward flight and a hover. I would attempt to correct and found that it was difficult to command a slow yaw. Even with a very soft stick movement, the yaw corrective action was very fast and difficult to control a slow smooth movement. Just adding the slightest amount of yaw stick input the P2 would speed up, slow down or stop yawing. My stick input would remain the same.

Flying straight was good but it would yaw without changing the initial heading.

The P2 was and has been on prior test flights, very twitchy in both a hover and slow flying. It seemed to be over reacting to it's corrections.

My video footage has various degrees of shakiness. I am assuming it is from the twitchiness during flight. It looks like it was a chug chug chug. Chug feeling.

 

[wkf94025]

Dang. That's not very encouraging. Based on your description, sounds like you would've been better off with v3.04 and a rotated compass.

Kelly

 

[John Shaw]

Xrover could you explain what you meant by"Flying straight was good but it would yaw without changing the initial heading." I don't know how to picture that as a yaw requires a change of heading by definition. I am not trying to be picky, just making sure I can picture what you are describing.

 

[ianwood]

3.04 and 3.02 with the H3-3D has similar issues with un-commanded yaw. Hard to tell if they are worse in 3.05 or the same. John with 3.05 and the P2V sees no un-commanded yaw issues.

3.05 yaw input does seems less consistent. I noticed this even more with the dual rate which would do nothing, do very little, do very little, and then whooomp, big yaw. That could bea gain settings issue but I don't know for sure.

What are the stock gains for 3.04, 3.02 and 2.0? Does anyone know?

As for your twitching and shaking video, I don't have those issues. Gimbal set up maybe a contributing factor. Do you have any twitching between gimbal and P2? Maybe try a flight without the gimbal and see if it is still the same.

Interesting that John's gains (P2V) are different than yours (P2Z).

 

[Xrover]

Xrover could you explain what you meant by"Flying straight was good but it would yaw without changing the initial heading." I don't know how to picture that as a yaw requires a change of heading by definition. I am not trying to be picky, just making sure I can picture what you are describing.

Good question John.

I mixed up my terminology. I should have said Course not Heading.

Example: I begin to fly straight towards a distant tree or building. I line up the center V indicator on my iosd mini fpv display. I try to maintain a straight course keeping the object centered. What happens is the nose of the P2 will start to Yaw to the left which I can see in the FPV. The P2 though will continue on the original course without changing it's original course even though the nose yaws to the left.

Make sense?

 

[John Shaw]

Xrover could you explain what you meant by"Flying straight was good but it would yaw without changing the initial heading." I don't know how to picture that as a yaw requires a change of heading by definition. I am not trying to be picky, just making sure I can picture what you are describing.

Good question John.

I mixed up my terminology. I should have said Course not Heading.

Example: I begin to fly straight towards a distant tree or building. I line up the center V indicator on my iosd mini fpv display. I try to maintain a straight course keeping the object centered. What happens is the nose of the P2 will start to Yaw to the left which I can see in the FPV. The P2 though will continue on the original course without changing it's original course even though the nose yaws to the left.

Make sense?

Xrover, that does make sense and I can picture it. Thanks
I gather since you say "I try to maintain a straight course" that you are making a control input to hold course. Is that a roll input? Or are you trying to cancel out the yaw? Or is their no control input at that point?

 

[Xrover]

I will be giving fwd. Stick only. Moving about 8 mph. Then after a few feet the nose will start to yaw to the left. I will then add a bit of right yaw stick and the P2 nose will yaw back right and as soon as i have the nose pointed straight again, i will ease up ont the yaw stick input and it will yaw back left after a second or Two. This could have a lot to do with the H3-3D GImbal trying to adjust it's yaw control. This might or might not be related to 3.05

 

[shartlza]

Xrover could you explain what you meant by"Flying straight was good but it would yaw without changing the initial heading." I don't know how to picture that as a yaw requires a change of heading by definition. I am not trying to be picky, just making sure I can picture what you are describing.

Good question John.

I mixed up my terminology. I should have said Course not Heading.

Example: I begin to fly straight towards a distant tree or building. I line up the center V indicator on my iosd mini fpv display. I try to maintain a straight course keeping the object centered. What happens is the nose of the P2 will start to Yaw to the left which I can see in the FPV. The P2 though will continue on the original course without changing it's original course even though the nose yaws to the left.

Make sense?

I have experienced that on 2.0 through 3.04 with the H3-2D and H3-3D, but thought it was the declination issue. It is like you are flying in CL and you slightly yaw it to the left, but the craft still stays the same course. Man, all of this stuff needs to get fixed so they can release a new firmware to fix all this!!!

 

[d4ddyo]

I scanned through the few recent posts. back to the beta test compass issue fix.

Is this much bigger with co-related issues than just a compass declination issue with everyone posting about gains and 3d gimbal issues and such?

 

[landonkk]

I scanned through the few recent posts. back to the beta test compass issue fix.

Is this much bigger with co-related issues than just a compass declination issue with everyone posting about gains and 3d gimbal issues and such?

I think that the j-hook issue is resolved and now people are starting to notice other issues that have been present in other versions of the released firmware that are not directly related. Because of that, this testing risks the chance of turning into a 'solve-all', where we try to fine tune the thing to perfection and end up in the testing stage for months and months... In my opinion.

 

[d4ddyo]

I scanned through the few recent posts. back to the beta test compass issue fix.

Is this much bigger with co-related issues than just a compass declination issue with everyone posting about gains and 3d gimbal issues and such?

I think that the j-hook issue is resolved and now people are starting to notice other issues that have been present in other versions of the released firmware that are not directly related. Because of that, this testing risks the chance of turning into a 'solve-all', where we try to fine tune the thing to perfection and end up in the testing stage for months and months... In my opinion.

makes sense... this is ridiculous.

 

[onfourblades]

UPDATE: Recieved from dji

Thanks for your patience, the beta firmware need to be tested for a little more time. and I will advise you when it is ready.

Best regards,

Zach
DJI Innovations

 

[d4ddyo]

UPDATE: Recieved from dji

Thanks for your patience, the beta firmware need to be tested for a little more time. and I will advise you when it is ready.

Best regards,

Zach
DJI Innovations

Ok. There's a light. A faint one but a light.

 

[ianwood]

I have experienced that on 2.0 through 3.04 with the H3-2D and H3-3D, but thought it was the declination issue. It is like you are flying in CL and you slightly yaw it to the left, but the craft still stays the same course. Man, all of this stuff needs to get fixed so they can release a new firmware to fix all this!!!

Agreed. And I've seen it on other firmwares too. And I think one or two others have as well. But, I don't think it's related to declination issue and it's not new to the beta.

I think that the j-hook issue is resolved and now people are starting to notice other issues that have been present in other versions of the released firmware that are not directly related. Because of that, this testing risks the chance of turning into a 'solve-all', where we try to fine tune the thing to perfection and end up in the testing stage for months and months... In my opinion.

No, it's mostly fixed, about 80%, and we still don't have an explanation for why it doesn't work for sar104. There are other oddities we are ruling out. It's a process. If you want to help, you're more than welcome to provide some analysis. There's plenty of video to look at.

If anyone wants to be critical of the process, this is not the thread. There's a different thread for that: viewtopic.php?f=19&t=18159

As I have said a couple days ago, DJI is working on the next version of the beta which they expect should get us better results than the current beta. I would expect we have it by the weekend but no guarantees.

 

[d4ddyo]

No, it's mostly fixed, about 80%, and we still don't have an explanation for why it doesn't work for sar104. There are other oddities we are ruling out. It's a process. If you want to help, you're more than welcome to provide some analysis. There's plenty of video to look at.

If anyone wants to be critical of the process, this is not the thread. There's a different thread for that: viewtopic.php?f=19&t=18159

As I have said a couple days ago, DJI is working on the next version of the beta which they expect should get us better results than the current beta. I would expect we have it by the weekend but no guarantees.

Yeah we get it. Some of us do know the process and what a beta is. If this where a automobile we would have recalls by now.

 

[d4ddyo]

No, it's mostly fixed, about 80%, and we still don't have an explanation for why it doesn't work for sar104. There are other oddities we are ruling out. It's a process. If you want to help, you're more than welcome to provide some analysis. There's plenty of video to look at.

If anyone wants to be critical of the process, this is not the thread. There's a different thread for that: viewtopic.php?f=19&t=18159

As I have said a couple days ago, DJI is working on the next version of the beta which they expect should get us better results than the current beta. I would expect we have it by the weekend but no guarantees.

Yeah we get it. Some of us do know the process and what a beta is. If this where a automobile we would have recalls by now.

But the communication is way better at least.