Pursuing a refund - Flyaway not found...
- Thread starter Leon001
- Start date
ianwood
Taco Wrangler
When the compass is off, the Phantom will fly to the wrong point. It will then recompute a new heading to the correct location, try to fly to it, miss it, rinse, lather, repeat. The pattern forms a circle. Hence TBE.
How does that not define exactly what happened? It moved in a near perfect growing arch in the same direction..
move, i missed, move further, i missed, move further.
If the compass was feeding random, yet erroneous data then I could see somewhat of a circle. But if that compass data is consistent but erroneous it would lead to a line. Ive been working with autonomous robots since I was 11. Ive seen it happen and could reproduce it, yet on a much smaller scale.
ianwood
Taco Wrangler
Because there is no circular pattern. It's always a circle. Compass based fly aways start off with small circles and build up to giant ever expanding circles.
EDIT:
I'm not saying it is categorically not a compass issue but it is not characteristic of typical compass issues.
The arc of this path is too broad to not have thrown a compass error. It is almost straight implying nearly a 180 degrees of error.
EDIT:
I'm not saying it is categorically not a compass issue but it is not characteristic of typical compass issues.
The arc of this path is too broad to not have thrown a compass error. It is almost straight implying nearly a 180 degrees of error.
Last edited:
Which is exactly what THIS path is. It just grew quickly. The arch created in THIS flight is near geometrically perfect.
I don't see why a flyaway claim would not be paid in your situation with dji. Something I feel is in your favor you pressed RTH at two different points and each time the P3 did not respond.
And the #1 reason is the Speed. Your quad was flying really fast, and a point in which you had full throttle it had no real effect on the speed. That alone should be all you need for proof.
Btw, your quad reached a speed of 65.5mph!
And the #1 reason is the Speed. Your quad was flying really fast, and a point in which you had full throttle it had no real effect on the speed. That alone should be all you need for proof.
Btw, your quad reached a speed of 65.5mph!
ianwood
Taco Wrangler
Like I said, that would require a huge compass error close to 180 degrees which would have been reported by the app. Frank, do you get get raw mag X, Y, Z in the data? I can confirm with the raw data.
Again, I disagree. This could be easily somewhere between 45 and 90 degrees. The initial radius is fairly small. Add to that the increase in speed and it would certianly increase the arc. A higher compass offset would have started at a more random appearing movement.What is the frequency the P3 compares its location vs its expected location?
ianwood
Taco Wrangler
Again, this is not typical of a compass issue. Too big, too fast. Without raw data, there is no point debating it further. If someone wants to send me raw X, Y, Z mag data, it shouldn't be hard to figure out if it played a role or not.
That speed would suggest very high tail winds. I have never gotten over 45mph with a tailwind.
Funny how over time your story changes.. First is was NOT a compass issue. Then it could be only if the compass was close to 180 out. Now its not typical. My point is that we don't know what DID cause the fly away beyond the obvious fact that it culminated from the power-up sequence preceding the flight. Validating its potential as a compass issue supports the presumption that switching to atti mode could have saved the bird. If it can be documented that, without a doubt, a transition to atti mode would save most of these flights then that can be voiced as a recovery step. But first we HAVE to determine all the potential causes. If we don't then we are only flipping switches and frantically slamming levers around to HOPE for a result. All while wasting much needed time! Your not being able to tell me the frequency that the NAZA samples and compares location only further shows that you dont KNOW, you simply disagree. Which is absolutely fine. I can accept that! But call it what it is!
Im NOT the best at decyphering log data, but does this not point to the compass failing to respond? And then alternating 90 degrees?
ianwood
Taco Wrangler
No.. looking back I found something that supported YOU. Do you have anything useful to provide? Or should I add you to my ignore list. All you can say is that I don't know what I'm talking about but you cant support anything YOU say. Seems the one with no comprehension is you.
Either assist, jsutify your thoughts, or go away.
Either assist, jsutify your thoughts, or go away.
ianwood
Taco Wrangler
If it doesn't fall into an easily identifiable pattern, I don't jump to conclusions.
I see a few things so far:
I see a few things so far:
- RTH button press was recognized but the Phantom did not appear to respond to it normally. (unclear if this is logged locally or bounced back from the p3).
- A throttle input was responded to with a corresponding increase in altitude.
- Horizontal speed, linear increase throughout.
- Compass, gradual rotation throughout. No raw X, Y, Z values to evaluate.
Ian is correct. You can view your own file for a comparison.
First, I didnt jump to anything. I just asked a question. Thus my NOT knowing why it displayed that way. Yes, mine is similar until the point where it appears I took off. Seeing that my logs look similar does disprove that bit of evidence. A simple answer would have done everyone well. Much more so than a pathetic attitude! No-one can explain why these fly-aways are happening. Yet it seems, at least in my experience, that when someone actually puts forth the efforts to perhaps solve the problem they will be met by banter and BS. Its no wonder you (the community) cant solve problem!
