Spontaneous 40' ascend with no input

Have you tried adjusting the horizon by holding the C2 button on and rolling the right photo knob?

My horizon is perfect, but my launch surface is always 100% level and true, verified by my pocket level.
Although, I am curious about this C2 and photo wheel idea? You can adjust horizon?

Also, my altitude "bounce" is not huge, but varies a few feet. I too took a video which I will upload later as I finally caught it on film! Err, digital video.. haha anyway, does DJI have ANY stance on the altitude bounce issue?

@John Locke I just received my Brand New 2nd Phantom 4 on Thursday. It was sent to me by Drone World to replace the first Phantom 4 and my initial complaint of "horizon tilt". I calibrated all the IMU's, compass, controller, gimbal,, I think I even calibrated my index fingers with my thumbs...... Took it out on Friday and the new unit immediately began to display the "Altitude Yo-Yo" Something you guys may or may not have noticed, though, is that the DJI Go app is not correctly recording the altitude of the Phantom 4's flight. It has serious errors that will probably not reflect the true variations in altitude on any given flight. Luckily,,,, I made a video yesterday!! I present,, "The DJI Phantom 4 Altitude Yo-Yo Ballet" ....
To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Wow, just watched your video, holy smokes! Mine literally only yo yo'so a couple feet, like 2 or 3 at most! That is seriously u acceptable! DJI needs to address this as it's not an isolated issue.
 
Last edited:
So, I will go back and read, but I hadn't had a chance to before posting my findings. So forgive me in advance, but has there been a solution?
I tested mine out today, and while mine does NOT go 40' like yours, it does adjust anywhere from 2' - 5', consistently. It happened as follows:
Fly in P mode (doesn't matter where also, at my parents property which is nothing but fields in the country, or my house in town, same result), after launch, hovers solid as a rock, no drift, no "yo yo"
FLy straight out and up a little, never getting above 20ish feet from the ground at a max, make a few "laps" in race car oval track fashion, when headed back towards me after not stopping, I gently let go of the sticks. Immediately after it settles and stops, it will drop down a few feet.

Second attempt, same flight, Sport mode. Exact same path and particulars, same exact result.
Both modes, with ALL sensors off for OA I could find (under the sensor menu), same result.

Then, I noticed something, mostly because I was paying attention a LOT closer. while flying away from me, in a straight line and not touching the vertical / throttle stick, the P4 would slightly rise up as it was moving forward, and only go up 2 - 5ish feet... as I flew it maintained whatever height it ascended to on it's own, but as I slowed down, and then immediately stopped, it descended to the height at which I started, which of course was lower than it's current position. I tested this a few more times and it did this each and every time! So to me, and at least in my issue, it appears that the P4 is having a GPS holding issue, and simply cannot maintain properly set altitude as it flies forward, but then makes up for it when it stops and realizes "oh yeah, I am supposed to her here, not there".
I am pretty sure I calibrated IMU, but I will do so again just to be certain and try again, as I was able to consistently reproduce my results.
 
So, I will go back and read, but I hadn't had a chance to before posting my findings. So forgive me in advance, but has there been a solution?
I tested mine out today, and while mine does NOT go 40' like yours, it does adjust anywhere from 2' - 5', consistently. It happened as follows:
Fly in P mode (doesn't matter where also, at my parents property which is nothing but fields in the country, or my house in town, same result), after launch, hovers solid as a rock, no drift, no "yo yo"
FLy straight out and up a little, never getting above 20ish feet from the ground at a max, make a few "laps" in race car oval track fashion, when headed back towards me after not stopping, I gently let go of the sticks. Immediately after it settles and stops, it will drop down a few feet.

Second attempt, same flight, Sport mode. Exact same path and particulars, same exact result.
Both modes, with ALL sensors off for OA I could find (under the sensor menu), same result.

Then, I noticed something, mostly because I was paying attention a LOT closer. while flying away from me, in a straight line and not touching the vertical / throttle stick, the P4 would slightly rise up as it was moving forward, and only go up 2 - 5ish feet... as I flew it maintained whatever height it ascended to on it's own, but as I slowed down, and then immediately stopped, it descended to the height at which I started, which of course was lower than it's current position. I tested this a few more times and it did this each and every time! So to me, and at least in my issue, it appears that the P4 is having a GPS holding issue, and simply cannot maintain properly set altitude as it flies forward, but then makes up for it when it stops and realizes "oh yeah, I am supposed to her here, not there".
I am pretty sure I calibrated IMU, but I will do so again just to be certain and try again, as I was able to consistently reproduce my results.

I really doubt that the aircraft relies on GPS positioning to maintain altitude, I'm pretty sure that's done by barometer.
 
I really doubt that the aircraft relies on GPS positioning to maintain altitude, I'm pretty sure that's done by barometer.

It doesn't do it in Atti mode.. so that is why I came to that conclusion... Otherwise I would have thought the same.
 
GPS only locks lateral motion.

Barometer is altitude.

See above as to the thought process, I do know how these things work, but it's the only thing NOT on in Atti, hence my deduction....
 
See above as to the thought process, I do know how these things work, but it's the only thing NOT on in Atti, hence my deduction....

Agreed - I follow your reasoning. But just to clarify, it's not that GPS is off in Atti, it's just that it doesn't use it for positioning. Atti just maintains attitude and altitude, and so it would be odd if the aircraft used entirely different methods to maintain altitude in different modes. But possible, I guess.
 
Agreed - I follow your reasoning. But just to clarify, it's not that GPS is off in Atti, it's just that it doesn't use it for positioning. Atti just maintains attitude and altitude, and so it would be odd if the aircraft used entirely different methods to maintain altitude in different modes. But possible, I guess.

I totally get it, and I am stumped as well. my thought s maybe something, while in GPS modes is somehwow interfering with the barometer possibly? I don't know, it's just odd altogether. It is a DJI product after all! ;)

But, in all seriousness, this sounds like an issue that more than 1 or 2 people are experiencing... I hope we don't get left out in the cold by DJI, and a fix is "under way"
Granted, mine isn't 40', and thank goodness... I could even live with mine, but will it get worse? Those are my questions.. I knew I should have just waited! haha
 
I totally get it, and I am stumped as well. my thought s maybe something, while in GPS modes is somehwow interfering with the barometer possibly? I don't know, it's just odd altogether. It is a DJI product after all! ;)

But, in all seriousness, this sounds like an issue that more than 1 or 2 people are experiencing... I hope we don't get left out in the cold by DJI, and a fix is "under way"
Granted, mine isn't 40', and thank goodness... I could even live with mine, but will it get worse? Those are my questions.. I knew I should have just waited! haha

I've seen a few feet of descent on stopping, but I haven't investigated it carefully, even though it did result in a collision with the top branches of a small tree on one occasion when I was bringing it in to land. Remarkably the aircraft backed out and recovered even though the props hit the branches. Now I give it much more clearance.
 
  • Like
Reactions: arcaine25
I also have noticed issues like this as well. In particular moving forward and the drown rises. Stop moving forward it skinks to where it should be.

I also have noticed that when stopping it sinks a few feet then rises. Odd stuff
 
  • Like
Reactions: arcaine25
I also have noticed issues like this as well. In particular moving forward and the drown rises. Stop moving forward it skinks to where it should be.

I also have noticed that when stopping it sinks a few feet then rises. Odd stuff

agreed, it's seriously odd, and needs fixed! Phantom 3 doesn't do it, at least mine anyway.
 
  • Like
Reactions: Romonaga
My P3P never exhibited any hint of this settling and raising back up, and nobody in this forum ever reported any hint of this with P3P/P3A either, to my knowledge. I'm sure DJI knows they have a problem with P4's YOYOing 4 to 8'. Thousands of craft are in the field, and I think a majority of the craft have this issue to some extent. Mine would settle 4 to 8' from the day I got it. Over time I accepted this behavior as a standard personality feature (begrudgingly) because so many others were reporting the same thing. It's only after it grew to a 40' to 60' YOYO effect that I said that's enough, she's going to the shop.

I have a few theories to explain this behavior. I have a hunch one of them is the culprit.
1. The downward thrust of high speed air over the arms create a sucking vortex at the vents on the bottom of each arm. The total sum of the suction at all 4 arms collectively create a negative atmospheric pressure inside the body, depending on the prop speed. This suction plays havoc with the barometer sensor, preventing the sensor from measuring reality outside of the body. This could explain why it does this only in a hover, because horizontal movement disrupts the air around to vents that prevent the vortex suction from occurring. This could be tested by putting tape over the vents for a quick test, making sure not to run it very long so it doesn't overheat. I'm not sure if anyone is willing to test this. Mine is at DJI getting fixed.
2. The fan located inside the body plays havoc with the barometer sensor, potentially blowing on the sensor (I'm unsure of the sensor location, it's just a possibility I'm thinking). I'm wondering if anyone is willing to test this theory by disconnecting the fan and take a short spin in a cold environment (to prevent overheating) to see if the YOYO effect stops.
3. DJI has a bad barometer sensor vendor, they bought poor quality sensors with high tolerances. Some people in this forum claim they absolutely don't have any sinking or rise of the craft after stopping and giving no stick input, so some barometer sensors are working correctly. However too many are complaining about the same symptom, it's not an isolated problem.
4. Bad firmware. I truly doubt bad firmware is the cause because DJI would have fixed it by now, I would think. However it may be a contributing factor. They must have had hundreds of complaints by now. I was hoping this would have been fixed in one of their first firmware updates, but no joy.

With my 10X increase in YOYO effect, I'm 90% expecting this to be caused by a bad barometer. I doubt DJI will tell me the cause when I get it back, I expect no explanation, minimizing their risk of fueling rumors.
 
Last edited:
  • Like
Reactions: arcaine25
I have no yo-yo problems with my P4 in hover or after stopping from flight.
 
My P3P never exhibited any hint of this settling and raising back up, and nobody in this forum ever reported any hint of this with P3P/P3A either, to my knowledge. I'm sure DJI knows they have a problem with P4's YOYOing 4 to 8'. Thousands of craft are in the field, and I think a majority of the craft have this issue to some extent. Mine would settle 4 to 8' from the day I got it. Over time I accepted this behavior as a standard personality feature (begrudgingly) because so many others were reporting the same thing. It's only after it grew to a 40' to 60' YOYO effect that I said that's enough, she's going to the shop.

I have a few theories to explain this behavior. I have a hunch one of them is the culprit.
1. The downward thrust of high speed air over the arms create a sucking vortex at the vents on the bottom of each arm. The total sum of the suction at all 4 arms collectively create a negative atmospheric pressure inside the body, depending on the prop speed. This suction plays havoc with the barometer sensor, preventing the sensor from measuring reality outside of the body. This could explain why it does this only in a hover, because horizontal movement disrupts the air around to vents that prevent the vortex suction from occurring. This could be tested by putting tape over the vents for a quick test, making sure not to run it very long so it doesn't overheat. I'm not sure if anyone is willing to test this. Mine is at DJI getting fixed.
2. The fan located inside the body plays havoc with the barometer sensor, potentially blowing on the sensor (I'm unsure of the sensor location, it's just a possibility I'm thinking). I'm wondering if anyone is willing to test this theory by disconnecting the fan and take a short spin in a cold environment (to prevent overheating) to see if the YOYO effect stops.
3. DJI has a bad barometer sensor vendor, they bought poor quality sensors with high tolerances. Some people in this forum claim they absolutely don't have any sinking or rise of the craft after stopping and giving no stick input, so some barometer sensors are working correctly. However too many are complaining about the same symptom, it's not an isolated problem.
4. Bad firmware. I truly doubt bad firmware is the cause because DJI would have fixed it by now, I would think. However it may be a contributing factor. They must have had hundreds of complaints by now. I was hoping this would have been fixed in one of their first firmware updates, but no joy.

With my 10X increase in YOYO effect, I'm 90% expecting this to be caused by a bad barometer. I doubt DJI will tell me the cause when I get it back, I expect no explanation, minimizing their risk of fueling rumors.

I am willing to try the tape theory over the vents, I mean why not. I can 100% replicate the issue every time, so if it's not raining today, I will give it a go when I get home from work.

I too lean on the bad barometer issue, but I am hoping, and probably pipe dreaming, that it's a firmware thing, maybe bad code not communicating with the barometer properly?
Not sure, I think I am just leaning on the hopes that it's an easy fix as opposed to one that requires me sending it in. i've never had to send one in, and don't want to if i can avoid it.
 
My P3P never exhibited any hint of this settling and raising back up, and nobody in this forum ever reported any hint of this with P3P/P3A either, to my knowledge. I'm sure DJI knows they have a problem with P4's YOYOing 4 to 8'. Thousands of craft are in the field, and I think a majority of the craft have this issue to some extent. Mine would settle 4 to 8' from the day I got it. Over time I accepted this behavior as a standard personality feature (begrudgingly) because so many others were reporting the same thing. It's only after it grew to a 40' to 60' YOYO effect that I said that's enough, she's going to the shop.

I have a few theories to explain this behavior. I have a hunch one of them is the culprit.
1. The downward thrust of high speed air over the arms create a sucking vortex at the vents on the bottom of each arm. The total sum of the suction at all 4 arms collectively create a negative atmospheric pressure inside the body, depending on the prop speed. This suction plays havoc with the barometer sensor, preventing the sensor from measuring reality outside of the body. This could explain why it does this only in a hover, because horizontal movement disrupts the air around to vents that prevent the vortex suction from occurring. This could be tested by putting tape over the vents for a quick test, making sure not to run it very long so it doesn't overheat. I'm not sure if anyone is willing to test this. Mine is at DJI getting fixed.
2. The fan located inside the body plays havoc with the barometer sensor, potentially blowing on the sensor (I'm unsure of the sensor location, it's just a possibility I'm thinking). I'm wondering if anyone is willing to test this theory by disconnecting the fan and take a short spin in a cold environment (to prevent overheating) to see if the YOYO effect stops.
3. DJI has a bad barometer sensor vendor, they bought poor quality sensors with high tolerances. Some people in this forum claim they absolutely don't have any sinking or rise of the craft after stopping and giving no stick input, so some barometer sensors are working correctly. However too many are complaining about the same symptom, it's not an isolated problem.
4. Bad firmware. I truly doubt bad firmware is the cause because DJI would have fixed it by now, I would think. However it may be a contributing factor. They must have had hundreds of complaints by now. I was hoping this would have been fixed in one of their first firmware updates, but no joy.

With my 10X increase in YOYO effect, I'm 90% expecting this to be caused by a bad barometer. I doubt DJI will tell me the cause when I get it back, I expect no explanation, minimizing their risk of fueling rumors.

The problem with the simple "bad barometer" hypothesis is that it doesn't really explain why it would only drift on stopping. A bad barometer would be expected to give fluctuating readings at all times, leading to the kind of spontaneous variations that you reported in this thread.

I find it somewhat remarkable that they are even able to use barometric sensors to obtain vertical resolution of a few feet in a flying device. The placement of aircraft static pressure sensors is critical to measurement of ambient pressure, and even then is susceptible to variation due to dynamic pressure induced during deviation from straight, level flight. I would guess that DJI is doing a lot of signal post-processing to extract the altitude during flight, and that most of these errors will be introduced at that stage.
 
  • Like
Reactions: arcaine25
I would guess that DJI is doing a lot of signal post-processing to extract the altitude during flight, and that most of these errors will be introduced at that stage.

This right here ^^^^ is why I am hoping / thinking that firmware could be the answer!
 
This right here ^^^^ is why I am hoping / thinking that firmware could be the answer!

Agreed, except that it presumably cannot just be firmware, because then it would be a universal problem for the P4. There may be some subtle interplay between hardware variations (maybe different sensor vendors?) and firmware data reduction. That still doesn't mean that it cannot be fixed in the firmware, of course. This feels a lot like the original P2 compass issue.
 
Agreed, except that it presumably cannot just be firmware, because then it would be a universal problem for the P4. There may be some subtle interplay between hardware variations (maybe different sensor vendors?) and firmware data reduction. That still doesn't mean that it cannot be fixed in the firmware, of course. This feels a lot like the original P2 compass issue.

That is true, not everyone is experiencing it, but I wonder how many have the most recent firmware and who is still on "release" firmware? I only ask because the day I got it, I chose not to update before the first flight. now, mind you, I did almost immediately after a quick maiden, and I surely wasn't paying attention and trying to find issues, but I didn't notice any yo yo. But, again, it could have simply been the fact that I wasn't looking for it, so I didn't see it. But, thinking back now, I clearly recall a time while "chasing" my wife, and it wasn't happening. I don't know for sure, I am probably grasping at straws, but I know whatever the case, definitely needs addressed by DJI in some form or fashion.
 

Recent Posts

Members online

No members online now.

Forum statistics

Threads
143,066
Messages
1,467,352
Members
104,933
Latest member
mactechnic