Hello,
I have a Mavic 2 Pro with 3 batteries. Today, I was out for flying at a completely normal day without wind.
I've had a nice flight with my first battery with no issues.
Then, on my second battery, I was planning a hyperlapse that would take 8 minutes. I started and I was on minute 4 when I realized that it came to the ending point way sooner than it reported. (I was checking something else on my phone when hyperlapse was in progress) I've switched to DJI GO and checked the hyperlapse, it was full of wiggles (I know hyperlapses aren't super smooth, I did them a lot before, but this one was like I've almost had several waypoints in a zig-zag pattern). Then I've immediately seen the drone switch to ATTI mode, losing all GPS signals. It then started to jump between several GPS satellite lock counts (e.g. becoming 0, next second 5, then 8, then again 0, then 3 etc. but never enough number to lock) while it started to drift on air. It wasn't uncontrollable, but it was fully out of GPS. I've landed safely. The second I landed, I've got full GPS lock with 15 satellites. I did an IMU calibration just in case something went wrong, restarted my drone, and had super GPS lock again immedately with no issues. I've armed and took off, and the moment I take off, I lose all GPS satellites again. I've tried this several times, when it landed it locked perfectly immediately (not taking even a second), and when I took off, it lost the whole GPS (falling from 15 to 0 immediately).

The first thing that I thought is whether a jammer was nearby (thought it wasn't near any government/military or a suspicable building/area or anything). I took off and flew a few hundred meters away (weather was clear and there was no wind so I was comfortable with attitude mode), but still no GPS. I gave up and landed. I also don't think a jammer would not be affecting GPS at all before takeoff, and kill the whole reception the moment I take off. In case of a jammer, it should just not lock at all before takeoff too. In an hour at a different location I've decided to check if the problem still persisted. I went to an open area, put the third battery, and voila, locked to 15 GPS satellites perfectly, flew about 20 minutes with no issues at all, then landed. I've also checked the previous two batteries (there was a bit charge left) and again, flying perfectly.

What might have happened?

Can you post log.  Here’s link just follow instructions.  http://www.phantomhelp.com/LogViewer/Upload/

Loss of GPS
Loss of GPS will lead to SUA drifting with the wind. Inertial Monitoring Unit (IMU) will maintain aircraft stability. Return-to-Home function will be lost. GPS signal can be lost when operating in mountainous areas, built-up areas, forest areas and any other areas where GPS signals are blocked. Vision positioning systems may not work if heights in excess of 5m are maintained.

Actions
• Maintain lateral position using right stick (mode 2) control inputs
• Land as soon as practical

I wish they didn't remove the secondary compass.  I've never seen issues like these with the older Mavic drones with redundancy compass/IMU.     Poor decision.

Well, I've just installed Assistant, but in this screen the spinner stays this way forever. I don't think it should be taking more than 5 minutes to load.

canp Posted at 12-12 13:37
[view_image]

Well, I've just installed Assistant, but in this screen the spinner stays this way forever. I don't think it should be taking more than 5 minutes to load.

You don’t need Assistant to upload your logs.

Okay I've just connected to my Windows virtual machine and it worked. Probably an issue with the Mac version of the Assistant. I'll be uploading flight logs in an hour.

Here is the link to the relevant flights' logs from today:
https://www.dropbox.com/s/lav265 ... 3_00-47-39.DAT?dl=0

sounds like a cold/cracked solder joint on the gps antenna.. vibration from the drone in flight is keeping it from working properly?  just a guess... the original mavic also has issues with gps wires cracking off the gps antenna board

Not sure if that's what you want. I've also uploaded the logs from my DJI GO app to the PhantomHelp, hope it helps:
http://www.phantomhelp.com/LogViewer/SJUKOC26NUOAN2NX8RBO/

DeuceDriv3r Posted at 12-12 14:42
sounds like a cold/cracked solder joint on the gps antenna.. vibration from the drone in flight is keeping it from working properly?  just a guess... the original mavic also has issues with gps wires cracking off the gps antenna board

Would the logs (not the small one but the huge 600MB+ one) have any data about GPS module? I've got the files but I have no idea how to view/examine the data inside it (just a huge DAT file with 4 flights inside)

canp Posted at 12-12 14:44
Not sure if that's what you want. I've also uploaded the logs from my DJI GO app to the PhantomHelp, hope it helps:
http://www.phantomhelp.com/LogViewer/SJUKOC26NUOAN2NX8RBO/

Looking at your log it just shows one warning at beginning of flight weak gps and for a split second it reports Atti mode, but immediately goes back into sports mode which is the mode you took off in , you said you had control over your aircraft so I presume you had gps,so unlikely it’s a problem with loose wire on the module, I haven’t ever heard of gps failing because of loose wire.

It may well have been that numbers of sats were not shown, but signal in your bar graph could have been shown gps signal.

I would try another area open space take off and keep altitude low until your happy you are getting signal, the best gauge of gps signal is on your bar graph. Also take off and fly for awhile in P mode.

hallmark007 Posted at 12-12 15:05
Looking at your log it just shows one warning at beginning of flight weak gps and for a split second it reports Atti mode, but immediately goes back into sports mode which is the mode you took off in , you said you had control over your aircraft so I presume you had gps,so unlikely it’s a problem with loose wire on the module, I haven’t ever heard of gps failing because of loose wire.

It may well have been that numbers of sats were not shown, but signal in your bar graph could have been shown gps signal.

Is it normal that number of satellites going to zero just after takeoff? It's a fairly open area and it was a clear weather. I really doubt losing 12 satellites in less than a split second is normal.

Here, I've also uploaded the previous flight (where the problem first started out of nowhere while flying normally). Seek into ~8m25s:

http://www.phantomhelp.com/LogViewer/1W5BYQEXW8Z1XFYOI5MF/

canp Posted at 12-12 15:18
Is it normal that number of satellites going to zero just after takeoff? It's a fairly open area and it was a clear weather. I really doubt losing 12 satellites in less than a split second is normal.

Here, I've also uploaded the previous flight (where the problem first started out of nowhere while flying normally). Seek into ~8m25s:

No it’s not normal, but I see you flew this flight in the same location, in order to rule in or out that area I would move somewhere completely different and try again , if the problem still exists then you could have a problem.
I noticed that during this flight you always had control and aircraft was in P GPS, do you know if signal bars had gone from green to grey or stayed green.

hallmark007 Posted at 12-12 15:30
No it’s not normal, but I see you flew this flight in the same location, in order to rule in or out that area I would move somewhere completely different and try again , if the problem still exists then you could have a problem.
I noticed that during this flight you always had control and aircraft was in P GPS, do you know if signal bars had gone from green to grey or stayed green.

I was preparing for a hyperlapse (checking some photo/video settings) and all of a sudden I remember falling back to ATTI mode. Honestly I can't remember the color of the bar.

canp Posted at 12-12 14:46
Would the logs (not the small one but the huge 600MB+ one) have any data about GPS module? I've got the files but I have no idea how to view/examine the data inside it (just a huge DAT file with 4 flights inside)

the logs will have gps satellite data like number in view, number locked etc..

if the antenna connection were loose I don't think the logs would be able to show that .. the gps receiver just would not be receiving signals and would have no idea why..

can't come up with another reasonable solution to why you would have gps lock on the ground then loose it immediately after throttle up.. then get it back the minute you were back on the ground

there was a guy here not long ago that found on the spark, the gps antenna is in contact with the top shell and in his case his antenna ceramic patch was cracked .. probably from a crash where the drone go hit on top.  he replaced the gps antenna and his bird was back in the air
edit...

actually I CAN think of when it would drop to zero SATS all of a sudden for no good reason.. any time the aircraft goes into ATTI mode.. the FC drops all sats while it resets ...

so as it has been mentioned.. try a different location.. not just for satilite reception but to ensure no mag interferecne issues.. the mavic 2 only has 1 compass not 2 like the original..

looking at some pics I have dug up on a mavic 2 teardown.. the GPS is co-located with the FC benieth the front cover unlike the original mavic that had the antenna on a seperate board up in the tail connected to the FC by a cable so unless the antenna is physically disconnected from the board, or damaged its likely not a physical problem..

Jammer seems unlikey, the gps uses several bands at 1.2 and 1.5 mhz. The commercially available drone jammers work at 2.4 and 5.8. I dont know why you would want to just jam gps signals. There are some jammers that multiband and jam from 1.5 to 30 mhz but I suspect you would have seen your control and video break up when the jammer swept those bands. I would rule out a jammer if I were you.

Hi canp,

Had a look at your 2 flights; in both flights GPS count and reception suddenly lowered!
Resulting in flying in OPTI and ATTI.

In the short flight you was in control beacuse your drone was in OPTI mode flying, Not P-GPS (no reception, no sats)

In your first flight there is missing 175 seconds of data, a gap between 2m 21s <> 5m 16 s.

Not normal behaviour, seen the place your were flying one would not expect to loose GPS reception, so if you have not ask DJI for help i advice you to do so.

IMO i suspect your drone has a hardware problem in the GPS module, but if it this GPS drop did not happen the other flights... just be carefull the next flights!Nothing worse that fly far away and loose GPS....(ATTI and no RTH function!)


cheers
JJB

JJBspark Posted at 12-13 08:25
Hi canp,

Had a look at your 2 flights; in both flights GPS count and reception suddenly lowered!

I think you will find while Aircraft shows to be in Atti mode, it never moved an inch, it was over 100ft altitude and 158 ft from home so if we are supposed to think Aircraft was in Atti mode how do we equate for the fact that it held perfectly still
Opti vision won’t work at alt of 100ft so it’s most likely that gps was holding aircraft in position but showing low count for satellite, so this might have more to do with bad data than a hardware problem with gps.
If aircraft was truly in Atti mode we would have seen if drifting, it didn’t and OP said he had control, opti mode was of no use at this altitude, so for me gps was working, but bad data was showing for some reason.

hallmark007 Posted at 12-13 12:43
I think you will find while Aircraft shows to be in Atti mode, it never moved an inch, it was over 100ft altitude and 158 ft from home so if we are supposed to think Aircraft was in Atti mode how do we equate for the fact that it held perfectly still
Opti vision won’t work at alt of 100ft so it’s most likely that gps was holding aircraft in position but showing low count for satellite, so this might have more to do with bad data than a hardware problem with gps.
If aircraft was truly in Atti mode we would have seen if drifting, it didn’t and OP said he had control, opti mode was of no use at this altitude, so for me gps was working, but bad data was showing for some reason.

HI hallmark,

Nice to see you back on this forum! I like your pictures of Sweden and Ireland.

About his short flight ; Flightlog showes that it was flying in OPTI, phanthom help does not show that.
OPTI is selected with GPS is not availiable and when there is enough visual reference, if not ATTI is selected.

Distance away form HP has nothing to do with flying in OPTI or not, as long as there is vis ref OPTI remains OPTI.
Height not important, only the visual reference is trigger for yes/no OPTI. * (and ofcourse 100 ft flying with only blue sky....no OPTI)

* mayby i am wrong, but never read in a DJI manual that there is a height limit for OPTI, so if you have that reference do share it on here.

And as we all know, in OPTI the drone works stabilized like in P-GPS mode.

This is why i always fly with a screen recorder active! more data to analyze, and on the app OPTI is seen but not directly in the flightlogs.

cheers
JJB

canp, unfortunately where you live the government is sensitive to drones so they implemented this a few years ago Drone Jammers . So best of luck but this maybe your problem.

JJBspark Posted at 12-14 00:13
HI hallmark,

Nice to see you back on this forum! I like your pictures of Sweden and Ireland.

I don’t doubt that aircraft reported opti mode, but while you say home distance doesn’t matter, it shows aircraft being held steady at 158ft from home, now we know that barometer will hold altitude but to hold position you need vision sensors working but will not work at 100ft so there should be movement either closer to homepoint or further away.
In opti mode in order to have vision sensors working, you must be in P mode, not sport, you must also be in range of the ground, so maybe you can explain how aircraft managed without gps and bottom vision sensors to hold its position ?

hallmark007 Posted at 12-13 12:43
I think you will find while Aircraft shows to be in Atti mode, it never moved an inch, it was over 100ft altitude and 158 ft from home so if we are supposed to think Aircraft was in Atti mode how do we equate for the fact that it held perfectly still
Opti vision won’t work at alt of 100ft so it’s most likely that gps was holding aircraft in position but showing low count for satellite, so this might have more to do with bad data than a hardware problem with gps.
If aircraft was truly in Atti mode we would have seen if drifting, it didn’t and OP said he had control, opti mode was of no use at this altitude, so for me gps was working, but bad data was showing for some reason.

I think the "never moved an inch" is not indicative of it actually not moving, but actually a symptom of the GPS problems that were occurring at the time.  We can see in the log that once the GpsLevel drops from 4 bars down to 0 bars, the GPS Lat/Lon values stopped updating.  Before that, even with no stick input, the Lat/Lon values changed slightly with each line in the log.  This is what I would expect to see in a "normal" situation.

The aircraft may not have been moving much, but if you look at the HSPEED values, the accelerometers definitely detected some movement that was not being picked up via the GpsSpeed values nor the actual GPS Lat/Lon location.  

Based on looking at the location via Google Earth and Google Maps, it certainly seems like there should be a reasonable GPS signal.  I would be very suspicious of some sort of a potential hardware problem.  If it were me, I might try a different location that is safe, flat, and wide open to see if the same behavior persists.  If it does, I would get in touch with DJI and start the support process.

Everyone suggests flying in a different location, but that's exactly what the OP reported doing in his first post. A hardware problem would most likely reoccur at the new location, unless perhaps the temperature was significantly different an hour later.

A jammer operating only in the airspace above the ground also seems unlikely unless ground zero was in radio shadow due to a building or other landscape feature.

I'd def keep an eye on the bird to see if problems continue, and fly only in safe, open spaces until you feel comfortable it's not an intermittent hardware issue. I'm also uncomfortable switching apps while flying (though I've done so on occasion w/o having any issues), but that's just me.

Good luck!

JJBspark Posted at 12-14 00:13
HI hallmark,

Nice to see you back on this forum! I like your pictures of Sweden and Ireland.

keep in mind that the height spec for opti are not height above take off position but height above the surface (objects) under the aircraft at the time if its sufficient for the sensors to lock onto patterns...

so you could be 200 ft in the air but 30 ft above a rooftop and it could be opti stabilized

JJBspark Posted at 12-14 00:13
HI hallmark,

Nice to see you back on this forum! I like your pictures of Sweden and Ireland.

mavic pro 2 sensor specs per website

Downward        Precision Measurement Range: 0.5 - 11 m
Detectable Range: 11 - 22 m

also mavic 2 'might' use other sensors not just its down sensors for opti mode position keeping .. don't know for sure but they give specs for all the sensors.. and their 'precision'

https://www.dji.com/mavic-2/info
on the mavic pro 1  I dont have the 2 Pro manual but suspect similar verbage but better capability based on specs above:

from the manual

Keep sensors clean at all times. Dirt or other debris may adversely affect their effectiveness. The Downward Vision System is only effective when the aircraft is at altitudes of 0.3 to 13 meters.Forward and Downward Vision Systems may not function properly when the aircraft is yingover water.
The Forward and Downward Vision System may not be able to recognize patterns on the ground in low light (less than 100lux).
Do not use other ultrasonic devices with a frequency of 40KHz when Forward and DownwardVision Systems are in operation.

JJBspark Posted at 12-13 08:25
Hi canp,

Had a look at your 2 flights; in both flights GPS count and reception suddenly lowered!

can you tell based on your log analyzer if going into opti/atti based on a compas issue where the FC drops the sats on purpose  vs the GPS receiver having issues?

DeuceDriv3r Posted at 12-14 06:45
keep in mind that the height spec for opti are not height above take off position but height above the surface (objects) under the aircraft at the time if its sufficient for the sensors to lock onto patterns...

so you could be 200 ft in the air but 30 ft above a rooftop and it could be opti stabilized

Hi,

true, but in this case ground was more or less flat with lot of reference. (google earth)

In the logs are 2 parameters about vision usage, one for the vision height and one for the vision position.

i think that at certain height level the vision height is not oke for showing good height indication, but may be working for helping the vision system in total for good positioning.
Vision postion becomes TRUE if GPS data is not there or not used  and there is enough reference.
In all the files i have seen TRUE means OPTI is active. (but on phanthom help conversion GPS mode is indicated).

Height 125 feet when at OPTI.  When there is a IMU error etc GPS reception drops from 5 to 0 in 0.1 sec ; GPS data not used. In this flight GPS sats number and reception does not show that behaviour. Why ?  that`s the big question....


Top line GPS #, bottom line GPS reception

cheers
JJB

hallmark007 Posted at 12-14 03:01
I don’t doubt that aircraft reported opti mode, but while you say home distance doesn’t matter, it shows aircraft being held steady at 158ft from home, now we know that barometer will hold altitude but to hold position you need vision sensors working but will not work at 100ft so there should be movement either closer to homepoint or further away.
In opti mode in order to have vision sensors working, you must be in P mode, not sport, you must also be in range of the ground, so maybe you can explain how aircraft managed without gps and bottom vision sensors to hold its position ?

158 feet from home has nothing to do with holding postion in OPTI mode.

OPTI mode becomes ACTIVE if no GPS etc and there is enhough visual reference.
So IMO has nothing to do with P or Sport mode. ( i will test this out again, to start flying without GPS and in Sport mode )

OPTI you cannot select yourself, it will happen as described above.


See#28 ; Vision system = vision height and vision position.

Cheers
JJB

Hi canp,

I have a question about your flight you uploaded (16-35-56)

After 1m 36s you changed your P-GPS mode into ? guess its one of the automatic flight modes, in the log its logged as "OTHER"   What was it ??



cheers
JJB

JJBspark Posted at 12-14 08:58
158 feet from home has nothing to do with holding postion in OPTI mode.

OPTI mode becomes ACTIVE if no GPS etc and there is enhough visual reference.

Opti mode won’t hold Aircraft in place, unless you are within 8 metres of the ground, barometer will hold altitude, but if vision sensors are out of range then they cannot hold position.

I think you will remember having this discussion regarding why limited height was in forced on aircraft this is for reasons that downward  vision sensors are limited to certain height, it’s less for spark more for P4Pro , opti vision or give it its real name optical vision will only work correctly when vision sensors are in reach of something to lock on to, so forward sensors coming within 15 metres of obstacle can trigger break on aircraft.
In sport mode optical sensors don’t work.

DeuceDriv3r Posted at 12-14 06:55
can you tell based on your log analyzer if going into opti/atti based on a compas issue where the FC drops the sats on purpose  vs the GPS receiver having issues?

Hiya,

I cannot find any error on IMU, Compass or Yaw.
Wish OP had a screen recorder active during his flights!

cheers
JJB

JJBspark Posted at 12-14 08:58
158 feet from home has nothing to do with holding postion in OPTI mode.

OPTI mode becomes ACTIVE if no GPS etc and there is enhough visual reference.

I believe that the mavic uses opti even IF there is still GPS because if you look at the specs on GPS only position accuracy vs how solid the mavic holds its position in a hover near the ground.. clearly its not holding position using GPS alone

I think it reports OPTI when the only AID to position is optical i.e. GPS down.. before it goes into freewheeling ATTI mode which is just altitude stabilization.

I believe that the ground sensors are uses any time its close enough to a surface that is lit well enough and has a pattern that it can decern movement over.. its different than what it uses for height.. that is sonar, the movement is an algo with the cameras...

JJBspark Posted at 12-14 09:09
Hiya,

I cannot find any error on IMU, Compass or Yaw.

looking at your traces its clear that ATTI and the loss of GPS signals are about the same time.. but its the chicken and egg which one caused it.

loss of GPS can put the bird into opti/atti

atti due to compass or some other FC issue will drop all sat bars and sat track to 0 while it resets... but you don't see compass errors in the log like a typical yaw/compass error/IMU working exception atti scenario .. strange..

JJBspark Posted at 12-14 09:03
Hi canp,

I have a question about your flight you uploaded (16-35-56)

Hi JJBspark,

I was trying to setup some creative shots and was actually trying to find a way to take a hyperlapse-pano of the scene (couldn't find a way, any recommendations are welcome in PM as it would go off topic), so I was playing with hyperlapse and other quick shot modes. That "other" is most probably hyperlapse, as I was looking through the menus of waypoints and course lock modes to find a suitable mode.

Boffin Posted at 12-14 01:27
canp, unfortunately where you live the government is sensitive to drones so they implemented this a few years ago Drone Jammers . So best of luck but this maybe your problem.

Hello,

While it might be the case, I find it unlikely as there wasn't anything sensitive around. It wasn't near a military base, government building, there wasn't a case of some "important person" coming or anything, and it was just a regular day with no special occasion. Yet, again, I was flying perfectly before it for 20 minutes before the incident began. Also, losing GPS right at the moment I take off and locking again right the moment I land really doubts me about a jammer. Of course, even with all these, it is possible, but I'd say unlikely.

KlooGee Posted at 12-14 05:55
I think the "never moved an inch" is not indicative of it actually not moving, but actually a symptom of the GPS problems that were occurring at the time.  We can see in the log that once the GpsLevel drops from 4 bars down to 0 bars, the GPS Lat/Lon values stopped updating.  Before that, even with no stick input, the Lat/Lon values changed slightly with each line in the log.  This is what I would expect to see in a "normal" situation.

The aircraft may not have been moving much, but if you look at the HSPEED values, the accelerometers definitely detected some movement that was not being picked up via the GpsSpeed values nor the actual GPS Lat/Lon location.  

There are many strange things in the log, I believe this could be caused more likely through inefficient gps , if there is a problem with gps unit being bad it is more likely to show on the ground, there was a problem with some GPS receivers with old P4 , these units were defective and showed no satellite detection on the ground which is what I would expect, not intermittent gps.

It can easily be solved by OP going to different location and trying another flight, although I would recommend a very conservative and safe location, if problem continues then is most likely craft goes back , if it flys ok then something else caused the problem.

Ex Machina Posted at 12-14 06:30
Everyone suggests flying in a different location, but that's exactly what the OP reported doing in his first post. A hardware problem would most likely reoccur at the new location, unless perhaps the temperature was significantly different an hour later.

A jammer operating only in the airspace above the ground also seems unlikely unless ground zero was in radio shadow due to a building or other landscape feature.

On the two logs posted by OP, both locations were identical and flights were very close together. Unless there is another log.

Hi, we are really sorry you’re experiencing this problem. May we know if there is any error message on the app? Is the firmware updated? If this issue persists, the GPS module may be malfunctioning. You may record a video and send it to us for diagnosis, or send the flight records to us. We appreciate your understanding.

DJI Wanda Posted at 12-14 13:42
Hi, we are really sorry you’re experiencing this problem. May we know if there is any error message on the app? Is the firmware updated? If this issue persists, the GPS module may be malfunctioning. You may record a video and send it to us for diagnosis, or send the flight records to us. We appreciate your understanding.

Hello Wanda,

Up in few messages above, I've posted the flight logs (both the small TXT ones and the huge DAT ones) from the problematic flights, they might help investigating. You may forward those logs to the technical team for examination.

Best,
Can.