P4 with latest firmware: spontaneous altitude change and poor altimeter precision after upgrade

Joined
Jul 18, 2016
Messages
27
Reaction score
5
Age
41
Hi guys,

I have P4 for a month and did not face any problems prior to installing new firmware (the one which supposed to fix yaw yaw),

now I have:

- spontaneousl ascends and descends at full speed by 20-30 meters, SCARY
- landing with 23 meters of altitude or negative altitude

PLEASE those who expierenced the same glitches, post details here!

Thank you!
 
A copy of report of my friend, another P4 user:

"
report from today.

So the new thing is upgraded last version of Dji Go app. The behaviour of the bird is unchanged. I dont know what could they have done with once so well working sensor but now it's worth crap.
-Bird shot up the air to 3'-40 meters (how the **** do i know the alt if the altimeter is wrong by 20m error?), controls idle, it descends VERY RAPIDLY. Actually, maybe in fear my eyes got bigger, but it looked like it was descending even faster than when i command him to descend. Anyway, it showed 30+ meters, still descending, but i was afraid to let it eventually stop by itself and used UP command. And it didnt look 30m to me, it was by my estimation 15m or 20 tops, one cant say with shitty altimeter. I also lowered it hovering to 1.5 m while altimeter showed ******* 20+ meters.

Then i dared to do the take off in sport mode. Fast wroooom up to unknown altitude (id say 40-50m), left the commands idle, and it started to descend fast again. So, essentially the same behaviour as in normal mode.

I did reord a video, but, well, it's ****. Nothing can be seen or heard clearly on it. I even had help, one phone recorded the RC, the other one the bird, but, well, nothing can be seen reliably. I.e. one can see the bird ascending up, disappearing in the skies, and after some time it appears again (because it spontaneously descended by half of it's altitude), but what does it mean to see a blinking spot in the blue sky around? I have a video with RC recorded but my voice was not recorded well, cant be heard. I'll try to do it better next time. In the meantime, we cant do much, you said they didnt care anyway. So, till the next FW update..

However, i have to say that, once i take controls in my hands, it works nice and stabe. No yaw yaw, the altitude is stable, both on sensor reading and my eye estimation. However, having that inaccuracy in mind, i cant let it fly anywhere where i cant see him physically, so not to high and not too far since i have NO IDEA at which real altitude S-1 is! Imagine that. Instrument may show 30m and it can be at 2m. So i always need to see the bird physically which implies limitations of course. Not to repeat again, but as i said i landed with altitude indicating 26m on the instruments.....
"
 
I updated and my height display is way off.
Yesterday it was showing 18ft when it was in reality around 140ft up.
Had this a few times now.
 
I've been asked this question sooooo many times, that I started to believe that if my **** would fall off, DJI would also have assumed that the proper IMU calibration would help!!

If you had an IMU in your body and one of it's main jobs was to keep your member attached, when your member fell off.... would you not think your IMU might be the issue?

To put this another way, the barometer is part of the IMU and the barometer is how the P4 judges it's altitude after take off. Firmware can affect all aspects of the IMU and it's something easy to calibrate. So you first look at the thing that mostly affects what could be the issue and is easiest to check. This is the nature of the beast. When your car is not idling correctly do you hate it when people first tell you to check the engine?

Personally, I also recalibrate the IMU after there has been a large temp change from the time I last performed the calibration (around twice a year). I also do the calibration with the Phantom only slightly cooler than the temp outside.
 
Well, you have made yourself clear,

Sincerely, I did not know that barometer is a part of Inertia Measurement Unit ;)

Whatsoever, thus glitch came with the latest upgrade. Not only the barometer sucks, but it also commands bird by itself move up or down, and that sucks squared!

Yes, I did the IMU calibration (level controlled!), as soon as I saw the glitch after the upgrade. But NOT before I discovered the glitch, so you won't say that the reason of the glitch could be some "improper calibration".

Since then I have done dozens of all sorts of calibrations, and none helped a bit.
 
Inertial measurement unit - Wikipedia, the free encyclopedia

Just tell me how Inertaia Measurement Unit is connected with barometric sensor in your Imagination.... and i will recommend a good psycologist to fix this )))))

You take a wiki page which talks about them in general.... but, to use your own information:
"Recently, more and more manufacturers also include three magnetometers in IMUs. This allows better performance for dynamic orientation calculation in Attitude and heading reference systems which base on IMUs"

Did you read the own page you linked to?

But.... if you wan to know what is in DJI's IMU you can refer to DJI's own wiki page:

IMU - DJI Wiki

"IMU (Inertial Measurement Unit) includes a 3-axis accelerometer, a 3-axis angular velocity and a barometric altimeter. It is used to recognize attitude."

It's not easy for anyone to know this. But this is why it's always recommend, because it controls pretty much everything that holds the Phantom stationary in the air.
 
I opened up the IMU on my P2 to locate the barometer, small thing with tiny holes on each end. I used light air pressure on it after being in water for approx 15 min 8' water, but I think none had gotten into it. It worked fine after. I don't know about my P4 yet but sure it's in or near the IMU section, wanting to take it apart bad, but still under warranty.


Sent from my iPad using PhantomPilots
 
I had a scare last night and I run the latest firmware and just did the app update. I also calibrated the IMUs and compass after the app update. I actually calibrate the compass before each flight if I change take-off location at all. I was descending full speed from fairly high altitude but when I released the control stick it just kept dropping--fast. I pushed up and no response--still dropping fast. I released it and tried again and it stopped about 15 feet from slamming down at a high rate of speed right in front of me. Had not had this happen before. Something seems now to be wrong. I think I will tip-toe when it comes to descending to land now.
 
You take a wiki page which talks about them in general.... but, to use your own information:
"Recently, more and more manufacturers also include three magnetometers in IMUs. This allows better performance for dynamic orientation calculation in Attitude and heading reference systems which base on IMUs"

Did you read the own page you linked to?

But.... if you wan to know what is in DJI's IMU you can refer to DJI's own wiki page:

IMU - DJI Wiki

"IMU (Inertial Measurement Unit) includes a 3-axis accelerometer, a 3-axis angular velocity and a barometric altimeter. It is used to recognize attitude."

It's not easy for anyone to know this. But this is why it's always recommend, because it controls pretty much everything that holds the Phantom stationary in the air.


Well, that means that after the firmware upgrade, the interpretation of the barometric sensor data, while calibrated ideally, does not work normally.

The glitch.

Worse off all the glitch commands the craft to ascend or descend on its own. And not slowly drifting, but full speed up or down by 20 - 30 meters!
 
I had a scare last night and I run the latest firmware and just did the app update. I also calibrated the IMUs and compass after the app update. I actually calibrate the compass before each flight if I change take-off location at all. I was descending full speed from fairly high altitude but when I released the control stick it just kept dropping--fast. I pushed up and no response--still dropping fast. I released it and tried again and it stopped about 15 feet from slamming down at a high rate of speed right in front of me. Had not had this happen before. Something seems now to be wrong. I think I will tip-toe when it comes to descending to land now.

it is NOT a physical state, this is purely a glitch! And they dont want even to recognize it!
 
I also have false altitude readings, off by about 40 to 50', which limits my ability to trust Litchi flight paths.

I also have a twitch in the yaw precision, something that's not easily reproduced but occurs about ever third or fourth flight as I make slow yaw turns in a hover, sometimes while moving too. The craft just turns about 10 degrees by itself, spontaneously.

Maybe DJI will sort these P4 anomalies out by the time the P5 arrives next year. ;-)
 
Well, that means that after the firmware upgrade, the interpretation of the barometric sensor data, while calibrated ideally, does not work normally.

The glitch.

Worse off all the glitch commands the craft to ascend or descend on its own. And not slowly drifting, but full speed up or down by 20 - 30 meters!

DJI recommends an IMU calibration after a firmware update. If firmware changes these things and part of the update is to recalibrate, it's not a glitch... it's part of the process.

It's not something that happens every time, but it can happen. I can't say why it happens some times and not another.

You mention that it dropped that distance when you were at full speed. So the change only affected the angle ever so slightly. It's the full throttle that made it pronounced.

Try the IMU calibration or don't. It's your choice. It may not solve the problem but it's a good place to start. You can complain about the issue but it's certainly not going to work toward a solution.
 
I also have false altitude readings, off by about 40 to 50', which limits my ability to trust Litchi flight paths.

I also have a twitch in the yaw precision, something that's not easily reproduced but occurs about ever third or fourth flight as I make slow yaw turns in a hover, sometimes while moving too. The craft just turns about 10 degrees by itself, spontaneously.

Maybe DJI will sort these P4 anomalies out by the time the P5 arrives next year. ;-)

Dont you think that a product with a price point like this, sold with 1000% profit, should be as reliable as a car is, they are selling us 50 bucks toys for 15 hundreds and yet they do not bother of good beta testing!

Shame.
 
I have posted this problem a few times. I have the same problems with altitude and just received my craft back from DJI service. Altitude is always off by 15-30 feet in all directions. If you full throttle straight up it will go up a little and then start to drop. It's like it knows where you left off on the stick but so much force in going up then brings it back down to where it's suppose to be. Kinda weird, and trust me Dji knows the problems. They just do not admit it, I sent mine out for tilted horizon, minor yoyo effect, altitude off and that was before the last update. The update came when it was at service and they replaced the main board, installed the updates, tested and sent back. It still has the alt problem, horizon I do believe will never be fixed but it is better. At least the yoyo is gone. I do believe like I said the craft keeps going and then starts to come down after full throttle because there was a spot that when you let go of the sticks.... It tries to get back to that point. In all my records the altitude was always off, just go back and review from the time you had your P4. One of the forums I am in the guys said that the altitude will always be off because of the barometer. If that's true then I guess we have to live with that. Let me know your thoughts...


Sent from my iPhone using PhantomPilots mobile app
 
Last edited:
DJI recommends an IMU calibration after a firmware update. If firmware changes these things and part of the update is to recalibrate, it's not a glitch... it's part of the process.

It's not something that happens every time, but it can happen. I can't say why it happens some times and not another.

You mention that it dropped that distance when you were at full speed. So the change only affected the angle ever so slightly. It's the full throttle that made it pronounced.

Try the IMU calibration or don't. It's your choice. It may not solve the problem but it's a good place to start. You can complain about the issue but it's certainly not going to work toward a solution.

No no no! I didn't mention I had any horizontal speed, I even did not touch the stick, I full throttle UP, gain some 50 meters altitude and released the stick and it started to descend, on its own, really quickly and that was scary! It descended by more than 20 meters when it stopped!

There was a user reported that he badly hit ground due to this glitch, do not tell us "it is part of the process" unless we recognize that the process is "fooling us enthusiasts and luring out our money for a 100 bucks toy! " Full of bugs!
 
This situation is standard corporative attitude to designing products. While engineers do their best, the marketing department pushes to the limit, set deadlines, cuts the design time, shorten testing and not providing enough resources and time for beta testing, it is all for this new bird appeared in March and not in June t.ex, because for the guys only money matter.

They screwed with barometric sensor, so it is too much affected by the bird speed and immediate movement history. Instead of designing 4 sensors bridge configuration which would correct any offset error but requires some design time they said to themselves probably:

"Okay, we compensate this later in software",

They knew pretty well what would likely to happen afterwards, but they did not care really, because they are corporative guys and they treat us as fools that waste own money for such a s... for which they would never do, so once they realized we are fools in their eyes, they treat us accordingly.

The whole situation is inappropriate.
 
I have posted this problem a few times. I have the same problems with altitude and just received my craft back from DJI service. Altitude is always off by 15-30 feet in all directions. If you full throttle straight up it will go up a little and then start to drop. It's like it knows where you left off on the stick but so much force in going up then brings it back down to where it's suppose to be. Kinda weird, and trust me Dji knows the problems. They just do not admit it, I sent mine out for tilted horizon, minor yoyo effect, altitude off and that was before the last update. The update came when it was at service and they replaced the main board, installed the updates, tested and sent back. It still has the alt problem, horizon I do believe will never be fixed but it is better. At least the yoyo is gone. I do believe like I said the craft keeps going and then starts to come down after full throttle because there was a spot that when you let go of the sticks.... It tries to get back to that point. In all my records the altitude was always off, just go back and review from the time you had your P4. One of the forums I am in the guys said that the altitude will always be off because of the barometer. If that's true then I guess we have to live with that. Let me know your thoughts...


Sent from my iPhone using PhantomPilots mobile app
If I would have made my thoughts about this issue public, the expession of my attitude towards these corporative guys would be qualified as "inappropriate content" on this forum ;)

So I keep being calm, but watch my lips carefully #^^#}|>>%##
 

Members online

No members online now.

Forum statistics

Threads
143,054
Messages
1,467,297
Members
104,919
Latest member
BobDan