2nd flight on 3.04 CRASH-VRS

[Kelso Kubat]

pretty pissed off right now. literally went from this forum reading and commenting about ******* VRS to having it happen about 8 minutes in. got it on video. i powered forward nothing.
dropped the throttle to like 20% - nothing.
luckily got it in the grass.

WHAT THE **** IS GOING ON WITH THE PHANTOM'S
http://youtu.be/poKYDziRO3c

 

[damoncooper]

You're holding it wrong (just kidding).

I ditched one in the ocean and almost death-wobbled my 2nd today except for my recognition of the beginning of the signs and pulled out of it instantly. I hadn't loaded gimbal or anything on it, was on 3.0.4. I was not descending, had just applied both sticks on a millennium falcon-desired smooth "twist and turn and accelerate down and away" move away from 30ft hover and the wobble appeared briefly.

I recognized that SOB instantly and let go the sticks. It recovered but it was the same wobble that, had my heavy gear been attached, I'm convinced would have taken it down.

A recent update noted "additional stability with heavy loads" and I wonder if that's a work in progress.

One things for sure: with a full load, keeping it simple in terms of complicated maneuvers is probably prudent SOP or you risk destabilization. I realize this has nothing to do with VRS but it's probably just as fatal if not also managed.

Luckily I was only 30 feet up over a soft grassy field, practicing, but I am learning to apply one or two (max) axis of input at once or bad crap obviously is likely to happen.

YMMV

 

[Kelso Kubat]

video now posted......
locked up and slowly started coming down. this really makes me sad and discouraged to take this gear up in the air.
the uncertainty of a successful flight is starting to piss me off big F"ing time !!!!!!! :shock:

 

[sar104]

Just to clarify - were you trying to descend when it happened, or was the descent self-initiated?

 

[IrishSights]

Ah...where is this video then? I suggest Phantom bashing is pre judging the cause. Its seems that is all that goes on on PP these days. I'm not saying there are no Phantom problem machines, its the quickness to curse the machine straight off that is actually just a tad imature. IMO

Lets see the video, describe the wind situation etc. And 3.04 is probably not the cause.

Sent from my Galaxy Note 8

 

[IrishSights]

Ok I see the video now. This does not look like VRS to me, more like a failsafe descent. Kelso if you provide the community with all of the facts of the flight including battery state, it just might help others here to help you instead of your colourful bad language that only tells us your state of mind.

Sent from my Galaxy Note 8

 

[wkf94025]

I agree with Kelso and Damon. I am seriously starting to question v3.04 (and possibly versions recently prior to 3.04). I augered in this morning while flying along fairly normally. It was a two stage crash -- clear VRS that I nearly recovered from shortly before impact, but then at low altitude succumbed again, fortunately landing in a vineyard that had recently been tilled (soft soil). Only damage is one broken prop. I have the GoPro footage and a DVR with the iOSD overlay. Hope to get them online tomorrow. The pitch and roll numbers are all over the place. When VRS first kicked in, I recognized it quickly, let go of the throttle, and gave right stick upwind.

My confidence in this platform is plummeting.

I predict a growing increase in VRS fatalities as people move to the more recent firmware.

Kelly

 

[wkf94025]

This does not look like VRS to me, more like a failsafe descent.

Respectfully disagree with your assessment. I have watched Kelso's video as well. I think the only way to descend that fast in v3.04 (other than in Manual mode) is VRS.

Kelly

 

[IrishSights]

How come many including myself who have been early adopters of 3.04 and are not getting your described problem. 3.04 has been teste systematiclly, and I mean systematiclly by many experienced pilots here and it is generally come out as the best since 1.08. Its not perfect it still has no max speed descent override and is very slow forward speed in GPS mode. If it was a firmware fault we would all be experiencing it. Basic logic says there are other things influencing this as described here.

Sorry, because of lack of serious description of scenario, and flight status, I'm out of this thread.

Sent from my Galaxy Note 8

 

[wkf94025]

If it was a firmware fault we would all be experiencing it.

The world is riddled with HW/SW/FW bugs that require just the right (wrong) conditions to trigger them. The obvious ones that show up consistently are usually nailed early and easily by the manufacturer / designer. Rigorous QA is all about methodical code coverage and persistence in the face of a slowing rate of bug detection.

Detailed flight video with OSD overlay coming shortly.

Kelly

 

[4wd]

Logically, in what way could the software version be responsible?
The major changes were reducing the rate of descent and quite significant changes in the default gains.
Have the new defaults been applied to this phantom?
It may be that's more important to do than was realised. Easy to overlook too.

 

[SteveO]

I thought VRS was introduced by descending too fast. When I fly in 3.04 straight down, full speed with default gains it comes down painfully slow, smooth, and steady. I just downgraded to 1.08 and this thing descends crazy fast but I'm always able to recover it.

 

[IrishSights]

This does not look like VRS to me, more like a failsafe descent.

Respectfully disagree with your assessment. I have watched Kelso's video as well. I think the only way to descend that fast in v3.04 (other than in Manual mode) is VRS.

Kelly

How can you induce VRS at 2 m/s!

Sent from my Galaxy Note 8

 

[xgeek]

On first look it does look like a typical VRS issue (The noise of the motors and the stability) from 0:58 onwards. But that does not explain why the OP had no right stick control after it started to descend and he could not move out of it. Also as IrishSights points out at 2m/s speed it is hard to enter VRS in the first place.

 

[wkf94025]

Here is the video of my VRS crash this morning:

http://youtu.be/cgXqsZTB3GM

Several things to note as you watch both the DVR with OSD and the footage off the GoPro starting at 02:28:
1) plenty of battery; power was not a factor; 67% left when the fun begins, 62% on impact.
2) flying along steadily in ATTI at 100m altitude, ~13 m/sec horizontal speed, starting to climb a bit, when VRS hits
3) random dots on the screen are apparently an iOSD Mini bug
4) this H3-3D gimbal spasms often on any flight that isn't very slow and steady
5) listen to the motors in the GoPro footage in the last segment. Hard to say if they/NAZA are doing the right thing.
6) Notice I pulled out of the Death Elevator with about 12 meters to spare, with the fun having started at 100 meters altitude
7) Notice the clearly bogus iOSD numbers on pitch, roll, and up/down velocity. Shortly after recovery at 9 meters you see a moment when NAZA+iOSD think the P2 is climbing and descending at appreciable rates, when it's clearly not based on the image and the absolute altitude.
8) If you watch the pitch number at about 0:28 into the video, just before the **** hits the fan, it appears I let off on the forward right stick a bit, and/or gave some up throttle, as pitch is dropping, upward velocity increasing, though still keeping about 10m/sec of forward speed. Winds were moderate and were blowing right to left as the p2 was flying along just before first VRS.
9) Just before final crash at 0:50 seconds, notice the pitch number and ascent number. I suspect that pattern somewhat matches the pattern at 100 meters before the fun began. Notice I was moving along in level flight at ~7m/sec when the final crash took place. Rest assured I was simply trying to get back home after the last minute recovery from a 90 meter ride in the VRS death elevator.
10) Notice the ascent/descent arrows are often on at the same time. Pretty sure that's not supposed to happen.

CORRECTION: this P2, which was new in January, is running v3.02 firmware, not v3.04. I have a newer P2 from April that is running v3.04.

Welcome your insights.

Kelly

 

[xgeek]

That does not look like a typical VRS. You was moving forward when it went crazy. To me it looks like you had another issue rather than VRS from descending too fast. Also what was the beep at 0:51 just before it flipped over?

2) flying along steadily in ATTI at 100m altitude, ~13 m/sec horizontal speed, starting to climb a bit, when VRS hits

Just with that info shows it could not have been VRS from descending. Something else went wrong.

 

[TheFlyingSentinel]

Guys ... I read a lot about the 3.04 Version ... I did about 50 flights with 3.04 and it is just fine ... no issue at all ...

BUT ... the issue that the my Phantom decents very rapidly out of the blue sky happens to me in the past as well. In my case it was related to a "bad" battery. With this particular battery the Phantom has dropped from sky allmost like a brick when the battery lavel was somewhere between 50 - 60% ...

No warning ... no Errors ... just rapidly decents without being able to control the decent path that much.

I send the battery for replacement ... got the replacement and the issue is gone ...

So maybe the cause is not really the Firmware ... maybe you have a faulty battery like I had ...

cheers,

Nik

 

[IrishSights]

Here is the video of my VRS crash this morning:

http://youtu.be/cgXqsZTB3GM

Several things to note as you watch both the DVR with OSD and the footage off the GoPro starting at 02:28:
1) plenty of battery; power was not a factor; 67% left when the fun begins, 62% on impact.
2) flying along steadily in ATTI at 100m altitude, ~13 m/sec horizontal speed, starting to climb a bit, when VRS hits
3) random dots on the screen are apparently an iOSD Mini bug
4) this H3-3D gimbal spasms often on any flight that isn't very slow and steady
5) listen to the motors in the GoPro footage in the last segment. Hard to say if they/NAZA are doing the right thing.
6) Notice I pulled out of the Death Elevator with about 12 meters to spare, with the fun having started at 100 meters altitude
7) Notice the clearly bogus iOSD numbers on pitch, roll, and up/down velocity. Shortly after recovery at 9 meters you see a moment when NAZA+iOSD think the P2 is climbing and descending at appreciable rates, when it's clearly not based on the image and the absolute altitude.
8) If you watch the pitch number at about 0:28 into the video, just before the **** hits the fan, it appears I let off on the forward right stick a bit, and/or gave some up throttle, as pitch is dropping, upward velocity increasing, though still keeping about 10m/sec of forward speed. Winds were moderate and were blowing right to left as the p2 was flying along just before first VRS.
9) Just before final crash at 0:50 seconds, notice the pitch number and ascent number. I suspect that pattern somewhat matches the pattern at 100 meters before the fun began. Notice I was moving along in level flight at ~7m/sec when the final crash took place. Rest assured I was simply trying to get back home after the last minute recovery from a 90 meter ride in the VRS death elevator.
10) Notice the ascent/descent arrows are often on at the same time. Pretty sure that's not supposed to happen.

CORRECTION: this P2, which was new in January, is running v3.02 firmware, not v3.04. I have a newer P2 from April that is running v3.04.

Welcome your insights.

Kelly

I know i said i was out of this thread but...Kelly, I gotta really commend you on an excellent report describing your situation. If all problem flights provided this much info the helpers would be in much better place to offer suggestions or give further insight.

Well done!

Sent from my Galaxy Note 8

 

[tom3holer]

Kelly,

I watched your video and agree it does not look like VRS.

I think its another battery issue. At one point for a brief sec the battery indicator symbol shows almost zero battery. There are also several frames where you see the battery voltage, percent, BUT the battery symbol is missing that points to a battery/battery sensing issue to me in my humble opinion.
The other clue is the prop sounds, they just die down or shut off when it acts up.

I really think its a power failure problem and you have very good evidence for a case with DJI or your dealer.

 

[damoncooper]

Interesting. I guess the good news is that if it was a battery issue that caused my lost P2 at sea, that battery is now out of my battery rotation