P4 Oscillations In Flight - Cannot Find Cause

[Hands0n]

My P4 has been a superb performer since the day I bought it. It is very well taken care of, not crashed or abused. It has flown like its on rails, steady and reliably throughout, until recently.

Yet of late I've noticed that it oscillates in flight and have not been able to trace what is going wrong with it. The attached video hopefully illustrates what I am experiencing.

Ordinarly, with any other FC I'd be looking at the PID and winding back the P a little. In the DJI Go 4 app I have been experimenting with the Gains to try and reduce or better still completely eliminate the oscillations. But to no effect. They continue, although the P4 then becomes very sluggish to control inputs, even in GPS mode.....

Any thoughts, ideas or suggestions where I should be looking for the fault? I'm not usually stuck but this one is baffling me, which means its likely to be something bleeding obvious that I'm missing

Update 0: I'm now wondering if this is indeed a V02.00.0106 update issue. Hard to prove unless I can try a controlled test of downgrading to the previous V01.02.0602 version and back a few times.

Update 1: After a lot of research here and on the DJI forums I have successfully updated the P4 to V01.02.0602, although a full flight test is yet to be performed, the short back garden flights reveal none of the previous twitchiness, instability and over-reacting by the aircraft.

The exact procedure I followed, to the letter, was this:
(NB: I didn't have to re-link the controller to the P4)

1) Format and remove the SD card from the bird
2) Upgrade to the latest DJI Go app, if you haven't already
3) Upgrade the controller (with the bird off) to the latest version
4) Power off the controller when done and start up DJI Assistant 2 on your PC of choice
5) Connect the PC to the bird and power on the bird
6) Upgrade the bird and power off when finished.
7) If you have more batteries, insert them into the bird, power it back up and perform the much shorter upgrade again.
8) Repeat step 7 for as many batteries as you have
9) You may need to relink your controller to the bird ( I had to )
10) Record videos at less than 40K30 resolution to avoid the new camera rebooting issue.​

Update 2: Hold that thought. After all of that, when I took it out to the field for a "proper" flight the fault manifested again. The above is not a 100% success rate it seems ...

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[sar104]

I suggest that you extract a DAT file of a flight with that behavior from the internal SD card and post it for analysis.

 

[Hands0n]

Hi, thank you @sar104 - I have attached the last flight .DAT file. The other files of the day are too large to load on here ... Although I could put one on Google Drive and share the link on here if that'll help.

 

[Hands0n]

Actually ...here's the link that should allow to download from my Public area on Drive

FLY127.DAT

 

[BudWalker]

Actually ...here's the link that should allow to download from my Public area on Drive

FLY127.DAT

Really interesting, but I don't have a clue. I assume you're talking about this.

The only stick input was full elevator.

The flight had more than a few examples of this. I used DatCon to extract the interval 460 - 480 secs at the highest sampling rate possible. If anyone wants to take a look at this higher sampling rate the attached is a .zip containing the .csv that can be viewed by submitting to CsvView.

 

[sar104]

I'm not completely sure what I'm seeing in the video but I did hear you say "full forward". The DAT file that you sent is clearly for a longer flight, although that video might have been part of it. However, nowhere in the flight record is any period of steady full forward stick (elevator) - it's all over the place. Vertically the aircraft seems to be stable.

Horizontally you are seeing some strange pitch oscillations between 0.5 – 1 Hz that are not, apparently, caused by stick input. I don't see an immediately obvious cause - will need to look at that more closely. Below is a zoomed part of the dataset showing a lack of correlation between elevator input and pitch at the higher frequencies.

 

[sar104]

Really interesting, but I don't have a clue. I assume you're talking about this.
View attachment 87742
The only stick input was full elevator.

The flight had more than a few examples of this. I used DatCon to extract the interval 460 - 480 secs at the highest sampling rate possible. If anyone wants to take a look at this higher sampling rate the attached is a .zip containing the .csv that can be viewed by submitting to CsvView.

Yes - pretty sure that is what he meant.

 

[BudWalker]

I'm not completely sure what I'm seeing in the video but I did hear you say "full forward". The DAT file that you sent is clearly for a longer flight, although that video might have been part of it. However, nowhere in the flight record is any period of steady full forward stick (elevator) - it's all over the place. Vertically the aircraft seems to be stable.

View attachment 87744

Horizontally you are seeing some strange pitch oscillations between 0.5 – 1 Hz that are not, apparently, caused by stick input. I don't see an immediately obvious cause - will need to look at that more closely. Below is a zoomed part of the dataset showing a lack of correlation between elevator input and pitch at the higher frequencies.

View attachment 87745

I was about to say it was close enough to full elevator and took a look. It is close to full elevator, but it's also noisy. And at a freq I don't think is physically possible (52 Hz). I'm thinking maybe this noisy control input has something to do with the oscillations?

 

[BudWalker]

Scratch that. Here is an interval with full elevator input without the above noise. Still the same problem with pitch and roll.

 

[Hands0n]

Hi guys, thank you for the detailed study. The video segment is part of the FLY127.DAT flight. I got my colleague to perform a fly by with full forward pitch. It also behaves really crappy when descending vertically into its own propwash, so I'd expect a little oscillation but not as much as is apparent, it almost looks as if its about to flip over.

But vertical descents aside, it is the horizontal flight at large forward pitch that manifests this strange oscillation. My feeling is that it has arrived with .1000 firmware update, but I cannot be 100% certain of that. I updated last weekend, the P4 complained at the field of IMU No.1 issues and switched to No.2. I recalibrated the IMU in the field, and back at home, but do not recall this issue on that day. It feels like it only started today, and was consistent across three battery swaps.

The aircraft is stable vertically, I'd also say that there is no particular evidence (in flight) of the Roll being affected like this. I'm going by eyesight of course, so could be wrong, the charts above suggest that I am [wrong]. It is literally only the Pitch (I'd say forward as well as back). That disparity that @sar104 has observed above is what I see in flight, and in that video clip.

Would it be worth running the IMU Recalibration again or do you fellas think that I have a physical fault with the P4 (I hope not, its only just out of warranty!!!).

Or could it be the TX itself and not the P4 at all??

 

[BudWalker]

.....

The aircraft is stable vertically, I'd also say that there is no particular evidence (in flight) of the Roll being affected like this. I'm going by eyesight of course, so could be wrong, the charts above suggest that I am [wrong]. It is literally only the Pitch (I'd say forward as well as back). ......

From the .DAT it's clear that pitch and roll are both affected. Pitch has a bias of about -18 because of the elevator input. Given that bias the oscillations are between pitch up, roll right and pitch down, roll left. When I look at the video I can't see the roll component but I think that it would be harder to see from the side.

Have you checked the props? Maybe try replacing the props?

 

[sar104]

From the .DAT it's clear that pitch and roll are both affected. Pitch has a bias of about -18 because of the elevator input. Given that bias the oscillations are between pitch up, roll right and pitch down, roll left. When I look at the video I can't see the roll component but I think that it would be harder to see from the side.

Have you checked the props? Maybe try replacing the props?

I was also wondering about the props. The motor data look fairly normal, so I don't think it is an ESC or motor problem, but it could be a bad prop. It could also be an IMU issue, and all the IMU data demonstrate the oscillations, but it's difficult to separate cause and effect. I'll take a closer look tomorrow.

The noisy control data are also strange. I might be useful to see the mobile device flight log for comparison to the DAT file?

 

[Hands0n]

I was also wondering about the props. The motor data look fairly normal, so I don't think it is an ESC or motor problem, but it could be a bad prop. It could also be an IMU issue, and all the IMU data demonstrate the oscillations, but it's difficult to separate cause and effect. I'll take a closer look tomorrow.

The noisy control data are also strange. I might be useful to see the mobile device flight log for comparison to the DAT file?

I've attached the file from the mobile device. The video clip in the OP was taken near the end of the flight, within the last 10 minutes or so, a direct line left to right across the field, it does stand out in the flight log.

The propellers look to be in top condition, there is no visible physical damage, they've never been crashed or experienced a strike. They're also the original props with 12hr flight total time and 330,483ft travel distance across 101 flights. I do have a spare that that I will try , hopefully today if I can get out again ...

If both IMU display he same oscillations wouldn't that make IMU issues less likely?

I've also uploaded the flight log to here -- DJI Flight Log Viewer - PhantomHelp.com

 

[Hands0n]

I'm now wondering if this is indeed a V02.00.0106 update issue. Someone replied to my YouTube post linking the DJI forum that has many people complaining about the same behaviour and matching video examples!

Aircraft totally unstable and aggressive after recent firmware up...

 

[Laar]

I had that same problem that suddenly popped up during previous firmware. As soon as i updated to the current firmware, it totally disappeared and is now more stable than it was out if the box.

 

[RémyHeli]

Once I started to read 1st post I immediatly thought was probably related to last firmware update causing issue to some of us since 2 month now. At the same time DJI removed the option to rollback to previous firmware so many pilots were stuck. Some smart guys worked on this to provide an unofficial tool to rollback which worked successfully for many pilots, or nearly same as before.

 

[Hands0n]

OK, so here's the latest from me. I have downgraded the P4's firmware from v02.00.0106 to the previous version that it was on, namely v01.02.0602. Unfortunately I will not be in a position to fly it unti later in the week or this Saturday! Work gets in the way of being a big kid.

 

[rcflyingaddict]

Once I started to read 1st post I immediatly thought was probably related to last firmware update causing issue to some of us since 2 month now. At the same time DJI removed the option to rollback to previous firmware so many pilots were stuck. Some smart guys worked on this to provide an unofficial tool to rollback which worked successfully for many pilots, or nearly same as before.

Yep same here... Spent an hour or so last evening (time that I'll never get back - lol) on the dji forums reading about the issues folks are having with their P4's post the latest update. I unboxed my P4P yesterday and after a half dozen flights one of the first things I noticed was the lack of 'rock solid' stability in hover. My only other dji is a MP and perhaps I'm too used to how it locks in. My P4P is not horrible and hopefully not as bad as folks have it in the dji thread they dont feel safe to fly but I guess I always assumed the P4P would actually be more stable due to it's size.

 

[hunterxero]

@Hands0n I have exactly this same behavior. And it started after the V02.00.0106 update. I remember the update in particular, because after I put it on, the "landed-idle" motor noise was much lower.

Since then in fast forward flight, my P4 makes exactly the same noise as in your video. If I fly it sport mode and full throttle, the oscillation is a little more pronounced and frequent.

I have tried the following remedies:
1. Tightened down the screws for the prop mounts (minor improvement)
2. re-calibrated compass (no change)
3. re-calibrated IMUs 3x (no change)
4. Replaced the starboard motors (1 CW and 1 CCW, no change, although they "seem" quieter)
5. re-calibrated compass and IMUs (no change)
6. I tried, but was unable to downgrade the firmware (no change)

I'm going to replace the port motors. I'll update you when that change is complete.

 

[RémyHeli]

What do you mean by "was unable to downgrade the firmware"? Can you provide more details please.