Hi,

When I watched a recorded flight in the DJI GO app I suddenly saw that the spark registered a negative altitiude. It's not the first time I've seen this so I uploaded the last time to PhantomHelp:

https://tinyurl.com/yccc3nw9

If you go to time 6m 1.6s you'll see the IMU go to 0 while the VPS is showing 12.5 ft and after that the IMU shows negative numbers until I start to rise again. So the IMU and VPS are completely out of sync. The app never showed me negative numbers during the actual flight so I'm not to concerned but it is a strange choice to show the IMU altitude in the app during playback.

My drone is on firmware 0900 and the app on build 4.1.22

Up to approx 30 ft your correct height is reflected through VPS, above this height your Aircraft will now use barometer to measure height you will see when VPS rises above this height it will read 0 and your correct height will be measured by barometer, barometer is very accurate so no need to worry, and when lowering your altitude below 30ft VPs will again kick in.
You may also notice when flying low over buildings or trees VPs kicking in and you may get incorrect values for altitude, other aircraft like P4Pro have capability to give readings in telemetry.

hallmark007 Posted at 2017-12-26 11:39
Up to approx 30 ft your correct height is reflected through VPS, above this height your Aircraft will now use barometer to measure height you will see when VPS rises above this height it will read 0 and your correct height will be measured by barometer, barometer is very accurate so no need to worry, and when lowering your altitude below 30ft VPs will again kick in.
You may also notice when flying low over buildings or trees VPs kicking in and you may get incorrect values for altitude, other aircraft like P4Pro have capability to give readings in telemetry.

It's not the vps that is showing the negative numbers, it's the IMU. I understand how vps works, I showed my son at home how it works with a polka dot blanket on the ground with the spark hovering above it. When he gently pulled the blanket the spark would follow it. My guess is that the IMU gyros are not precise enough to estimate the altitude but the app uses the recorded IMU data in the app during playback.

Raploc Posted at 2017-12-26 11:58
It's not the vps that is showing the negative numbers, it's the IMU. I understand how vps works, I showed my son at home how it works with a polka dot blanket on the ground with the spark hovering above it. When he gently pulled the blanket the spark would follow it. My guess is that the IMU gyros are not precise enough to estimate the altitude but the app uses the recorded IMU data in the app during playback.

I didn’t say it was VPS I’ll explain again, VPS is used when taking off and climbing to 30ft and landing or flying below 30 ft this is your correct altitude above this your correct altitude is measured by barometer.
IMU as used in phantomhelp is barometer measurement.

So the barometer is not reliable in my case or it accumulated a lot of error readings. It was windy but that was above the tree line not at 4 meters.

OR... the other possibility is that there was a change in atmospheric pressure during your flight.  Even a change as small as 0.01 psi can result in almost 20 ft change in calculated altitude.  In any case your takeoff altitude is always 0.  I don't think the IMU does any recalibration during a flight, but probably  only reports a moving average of the last few data points.  If your pressure sensor is flaky it may produce an erroneous average, but I don't think, in that case, you're readings would be as consistent.  My vote is pressure change.   Butas Hallmark commented,  VPS is much more accurate when within range and over a relatively solid surface and is insensitive to atmospheric pressure (mostly :-) )

Griffith Posted at 2017-12-26 13:04
OR... the other possibility is that there was a change in atmospheric pressure during your flight.  Even a change as small as 0.01 psi can result in almost 20 ft change in calculated altitude.  In any case your takeoff altitude is always 0.  I don't think the IMU does any recalibration during a flight, but probably  only reports a moving average of the last few data points.  If your pressure sensor is flaky it may produce an erroneous average, but I don't think, in that case, you're readings would be as consistent.  My vote is pressure change.   Butas Hallmark commented,  VPS is much more accurate when within range and over a relatively solid surface and is insensitive to atmospheric pressure (mostly :-) )

I think you are right, a drop of 1 hPa would be about 8 meters (24 ft). That would explain the difference in height. Too bad the spark doesn't calibrate the barometer with the vps system during the flight and that it is used as the height reference in the app during  playback.

And there is one more element to take into consideration if I may add it: If the take off point is higher than the following flight area, there will be also discrepancies in altitude measurements (negative values), or am I wrong?

I mean a rise in pressure not a drop, the IMU would then show the spark to be at altitude when it landed not below the reference point it took during take off.

Wachtberger Posted at 2017-12-26 13:27
And there is one more element to take into consideration if I may add it: If the take off point is higher than the following flight area, there will be also discrepancies in altitude measurements (negative values), or am I wrong?

True, but in my case that wasn't the issue.

Raploc Posted at 2017-12-26 13:31
True, but in my case that wasn't the issue.

Because the Netherlands are so flat? Please accept my apologies for this little joke, but I just couldn't resist ;-).

Wachtberger Posted at 2017-12-26 13:39
Because the Netherlands are so flat?  Please accept my apologies for this little joke, but I just couldn't resist ;-).

Yep, the ground here is pretty flat  
.

Raploc Posted at 2017-12-26 13:50
Yep, the ground here is pretty flat  
.

Being your "neighbour" I know that quite well. Good that you took it as meant :-)

Raploc Posted at 2017-12-26 11:58
It's not the vps that is showing the negative numbers, it's the IMU. I understand how vps works, I showed my son at home how it works with a polka dot blanket on the ground with the spark hovering above it. When he gently pulled the blanket the spark would follow it. My guess is that the IMU gyros are not precise enough to estimate the altitude but the app uses the recorded IMU data in the app during playback.

Hi, Raploc. Did it happen frequently? I would recommend you cool down the AC and take off your drone after a few minutes if you encounter this situation again. It may be caused the change of the atmospheric pressure of the environment.

DJI Elektra Posted at 2017-12-26 23:51
Hi, Raploc. Did it happen frequently? I would recommend you cool down the AC and take off your drone after a few minutes if you encounter this situation again. It may be caused the change of the atmospheric pressure of the environment.

Hi Elektra, I've checked other flight records and it seems to happen after 4-5 minutes of flight. The ambient temperature was low on all occasions. Somewhere in the range of 4 to 10 degrees Celsius.

Edit: Is this normal behaviour at these ambient temperatures? Does anyone else recognise this with his or her Spark?

DJI Elektra Posted at 2017-12-26 23:51
Hi, Raploc. Did it happen frequently? I would recommend you cool down the AC and take off your drone after a few minutes if you encounter this situation again. It may be caused the change of the atmospheric pressure of the environment.

Hi Elektra, see my reaction/question, is this normal behaviour for the Spark with these low ambient temperatures?

Wachtberger Posted at 2017-12-26 13:27
And there is one more element to take into consideration if I may add it: If the take off point is higher than the following flight area, there will be also discrepancies in altitude measurements (negative values), or am I wrong?

Correct.  A common problem for those flying down slope.

Shameless bump. Any admins out there who can provide an answer to my question regarding the "overheating" at low ambient temperatures?

Raploc Posted at 2017-12-27 22:14
Shameless bump. Any admins out there who can provide an answer to my question regarding the "overheating" at low ambient temperatures?

Sorry for the late response. Please provide me your DJI Account so we can look into your flight records for further analysis. Or you can PM me about it.

DJI Elektra Posted at 2017-12-29 04:43
Sorry for the late response. Please provide me your DJI Account so we can look into your flight records for further analysis. Or you can PM me about it.

Send via PM

Raploc Posted at 2017-12-29 05:43
Send via PM

Got your PM. Please wait patiently.