I was flying my Spark the past weekend and wanted to test its claimed maximum altitude that it can reach.

Everything went smoothly and the video feed was great, up until it reached around 1,100 feet which was around the cloud level, the Spark started to report really odd altitude readings, flickering between 0 feet to 900+ feet, then normalized until I went up to a bit shy of 1,300 feet, then the DJI Go 4 app showed "disconnected".

After more than 70 flys, I told myself not to panic, and just hit the RTH button since the RC still shows green. I heard the audible beep but waited for a few minutes, yet found no sign of my Spark. I started to think, what I could do to bring the Spark back as I was getting really worried. I disregarded my RC's RTH, and started dropping the stick to descend, at least for about 3 minutes, which felt eternity. Thank goodness I finally saw my Spark to show up, and the battery level has reached to extremely low level and must land. I manually landed my Spark and found a very little battery left. Good thing was, it did not crash and I managed to land it successfully. I looked at the Spark, and found some water spot all over probably due to entering and exiting the cloud and the moisture hit the body. I wonder if this could contribute the failure to read the altitude correctly, as well as causing the RTH to fail.

I further confirmed that the RTH failed multiple times when I looked at the flight replay from the DJI 4 go app. Unfortunately, it did not tell me why it failed.

Has anyone experienced this and how to tackle this problem? What would you do in my situation? Thanks.

James

Keep in mind that clouds, fog and sometimes water can throw the sensors of and spark goes a little nuts. Good on you talk your self out of panicking - best thing to do when ever something goes wrong.

The barometer is affected by movement of air around the drone, humidity, temperature changes and even voltage changes. Most of the time if you power back down and start up again it will reset. I suspect humidity and air pressure may have been the cause of incorrect altitude height.

With regards to rth because you were up so high wind could have been a factor in your spark’s slow return and when you pulled back throttle stick this cancelled RTH and you managed to land your spark safely.

I don’t think there was any malfunction, but it can be scary waiting to see your spark appear particularly if it moving very slowly. Alls well and ends well.

hallmark007 Posted at 2017-10-30 15:17
The barometer is affected by movement of air around the drone, humidity, temperature changes and even voltage changes. Most of the time if you power back down and start up again it will reset. I suspect humidity and air pressure may have been the cause of incorrect altitude height.

With regards to rth because you were up so high wind could have been a factor in your spark’s slow return and when you pulled back throttle stick this cancelled RTH and you managed to land your spark safely.

Capt. Hallmark007,

You're always responsive and helpful. Thank you for that!!

I can't tell you how much scare that I've run into in this incident but I learned a lesson of NOT flying into the cloud going forward, as you stated humidity & pressure could have caused this.

May I ask you this, God forbids if this happens again, do you think I should rely on the RTH button on the RC and let it do its job? The most difficult part was, I would have no idea of when I will get the visual of the Spark when the RTH is initiated. Would it fly back to me in Sports mode? or it will just take its time to return home?

Thank you!!

James

jcgoobee Posted at 2017-10-30 15:28
Capt. Hallmark007,

You're always responsive and helpful. Thank you for that!!

I think at that height and depending on distance Spark is always going to rth home slowly, if your aircraft is at 120 metres altitude it will stay at this altitude until it arrives over the homepoint and starts it’s descent. So yes my advice is if you have telemetry use small radar bottom left hand on your monitor for direction and height and distance and switch to and fly sport mode manually.

If there is no wind then outside 100metres spark will return at 10 m/s, but at 120 metres altitude there is nearly always wind so just depends if it’s with you or against you.

hallmark007 Posted at 2017-10-30 15:39
I think at that height and depending on distance Spark is always going to rth home slowly, if your aircraft is at 120 metres altitude it will stay at this altitude until it arrives over the homepoint and starts it’s descent. So yes my advice is if you have telemetry use small radar bottom left hand on your monitor for direction and height and distance and switch to and fly sport mode manually.

If there is no wind then outside 100metres spark will return at 10 m/s, but at 120 metres altitude there is nearly always wind so just depends if it’s with you or against you.

Thanks Capt,

I do have one follow up question though. The small radar in the lower left corner can be toggled with a map. Yes I can read the map just fine but I don't quite understand of how the radar works. I looked at the DJI documentation and found no relevant information about the radar. Is there a simple way to interpret the radar layout? This will help me a lot when I navigate. Oh yes, I guess the main reason I was able to descend the Spark near me, was that the small map still works, so I know where approximately it was at.

Thank you.

James

jcgoobee Posted at 2017-10-30 15:47
Thanks Capt,

I do have one follow up question though. The small radar in the lower left corner can be toggled with a map. Yes I can read the map just fine but I don't quite understand of how the radar works. I looked at the DJI documentation and found no relevant information about the radar. Is there a simple way to interpret the radar layout? This will help me a lot when I navigate. Oh yes, I guess the main reason I was able to descend the Spark near me, was that the small map still works, so I know where approximately it was at.

Small little icon on right hand upper corner of map , click on it, it will bring up radar your homepoint is at centre of radar just turn aircraft icon to point towards centre of radar and fly directly to homepoint , it is very useful tool and easier to fly with than map .

Was your GPS still working at the time? I didn't see you mention it failing or anything like that.
If it was, why should Spark need the barometer or vision sensors to tell its altitude? GPS can do that much more reliably and accurately, and it would overcome the issue of changes of perceived air pressure or the presence of reflected light from clouds, water etc.
Of course, if it were programmed into its firmware... If it isn't, it should be.

djiuser_Kf4iPA3 Posted at 2017-10-30 16:09
Was your GPS still working at the time? I didn't see you mention it failing or anything like that.
If it was, why should Spark need the barometer or vision sensors to tell its altitude? GPS can do that much more reliably and accurately, and it would overcome the issue of changes of perceived air pressure or the presence of reflected light from clouds, water etc.
Of course, if it were programmed into its firmware... If it isn't, it should be.

Gps takes its reference only from homepoint .
The mapping system used does not have altitude information for the gps to reference. The gps and barometer give altitude from the take off point, but would need the altitude information for the ground your flying over to give you accurate Altitude from gps,
What would AC do if you have gps dropout ,and there are many more practical reasons why barometric pressure is used in aviation to measure altitude.

djiuser_Kf4iPA3 Posted at 2017-10-30 16:09
Was your GPS still working at the time? I didn't see you mention it failing or anything like that.
If it was, why should Spark need the barometer or vision sensors to tell its altitude? GPS can do that much more reliably and accurately, and it would overcome the issue of changes of perceived air pressure or the presence of reflected light from clouds, water etc.
Of course, if it were programmed into its firmware... If it isn't, it should be.

The GPS isnt knowing if you have flown the spark just over a plattform which the vision system is seeing a cloud sometimes.

There have been many accounts of the spark seeing a cloud or fog as the ground... You were above a cloud and the spark thought that the cloud was the ground.  RTH didn't work because the spark was trying to land on what it thought should be firm ground and could not.   You are lucky that the stick down brought the spark back down. Others have found that they could not descend through a cloud.  This has nothing to do with GPS or barometer... it has everything to do with the downward vision sensor misinterpreting clouds as ground, that is why your altitude said zero even up so high.

LouisP Posted at 2017-10-30 17:08
There have been many accounts of the spark seeing a cloud or fog as the ground... You were above a cloud and the spark thought that the cloud was the ground.  RTH didn't work because the spark was trying to land on what it thought should be firm ground and could not.   You are lucky that the stick down brought the spark back down. Others have found that they could not descend through a cloud.  This has nothing to do with GPS or barometer... it has everything to do with the downward vision sensor misinterpreting clouds as ground, that is why your altitude said zero even up so high.

You may need to go back and read OP.

* op said that he was getting incorrect altitude measurements ie barometer no mention of VPS.
* op said he hit RTH and heard beeping so RTH was working, it may have been working slow because of altitude at 120 metres, he decided to cancel RTH by moving right stick throttle and lowering AC which is exactly what happened.
What you are talking about is a very different situation where some users could not lower AC because VPs taking a reading from clouds and thinking it was close to the ground, this did not happen in this situation.
Everything worked correctly in this situation there was no malfunction just one incorrect measurement from barometer, everything else worked as it should.

hallmark007 Posted at 2017-10-30 17:26
You may need to go back and read OP.

* op said that he was getting incorrect altitude measurements ie barometer no mention of VPS.

That is exactly what happened.. Barometer gives altitute when the drone is more than several feet off of the ground (out of sensor range).. close to the ground the down facing sensors take over because they are much more accurate than the barometer at telling how close the ground is.  
When others have flown over fog or cloud, the altitude display showed zero or just a couple of feet because the down facing sensors thought that was how high the spark was.
OP said that his altitude flipped  between zero and over a thousand feet, that sounds exactly like the down facing sensor saying 'there is the ground right under me' and then not seeing the cloud switching back to the barometer.
RTH was trying to work.. beeping and descending to what it THOUGHT was the ground but we must assume that spark is smart enough to know that even though the down facing sensor says 'you have landed' it does not stop the props because it knows that it is still flying.. I would like to know exactly what the RTH error said.
I think we are trying to say the same thing, but maybe using two different ways of saying it.

LouisP Posted at 2017-10-30 18:31
That is exactly what happened.. Barometer gives altitute when the drone is more than several feet off of the ground (out of sensor range).. close to the ground the down facing sensors take over because they are much more accurate than the barometer at telling how close the ground is.  
When others have flown over fog or cloud, the altitude display showed zero or just a couple of feet because the down facing sensors thought that was how high the spark was.
OP said that his altitude flipped  between zero and over a thousand feet, that sounds exactly like the down facing sensor saying 'there is the ground right under me' and then not seeing the cloud switching back to the barometer.

There was no RTH error when you press RTH at 120 metres first thing AC will do is fly to homepoint not descend.
Op cancelled RTH by pulling down on throttle, once he did this AC descended so in other words AC worked as it should.

AC height went from 900 to 0 when op was climbing not descending.

hallmark007 Posted at 2017-10-30 18:54
There was no RTH error when you press RTH at 120 metres first thing AC will do is fly to homepoint not descend.
Op cancelled RTH by pulling down on throttle, once he did this AC descended so in other words AC worked as it should.

OP said 'I further confirmed that the RTH failed multiple times when I looked at the flight replay from the DJI 4 go app.'
I assume that it showed a message box that said it failed.

Please note that flying your drone through cloud can cause crash and your drone may find it hard to land. I would not recommend you do the test and reach the cloud again.

hallmark007 Posted at 2017-10-30 17:26
You may need to go back and read OP.

* op said that he was getting incorrect altitude measurements ie barometer no mention of VPS.

-flickering between 0 feet to 900+-

And same probably, better assumingly has happened in disconnection mode. Dont you think?
Otherway round, 1300 feet, 3 minutes down, almost "in time", methinks

hallmark007 Posted at 2017-10-30 15:53
Small little icon on right hand upper corner of map , click on it, it will bring up radar your homepoint is at centre of radar just turn aircraft icon to point towards centre of radar and fly directly to homepoint , it is very useful tool and easier to fly with than map .

Hi Capt,

I thought the default icon (the one which looks like a Google map) does show my homeport as I used it to find my way home should the video feed fails. When I click on the upper right corner of the small icon, a blueish map shows up. I thought it shows the transmission power between Spark and the RC. I just didn't know how to interpret the graphical representation of it.

James

djiuser_Kf4iPA3 Posted at 2017-10-30 16:09
Was your GPS still working at the time? I didn't see you mention it failing or anything like that.
If it was, why should Spark need the barometer or vision sensors to tell its altitude? GPS can do that much more reliably and accurately, and it would overcome the issue of changes of perceived air pressure or the presence of reflected light from clouds, water etc.
Of course, if it were programmed into its firmware... If it isn't, it should be.

Hi djiuser_Kf4iPA3,

from what I recall, the GPS was working UNTIL the DJI Go 4 app shows that the Spark was disconnected. However, when I relaunched the app, the GPS map returned but no video feed until it hit around 700 feet (give or take). At that time, I felt a bit relax as both the GPS map and video feed came back alive, but I was in total panic mode when I was unable to see what was going on between 700 - 1300 feet and the RTH failed. As "Hallmark007" mentioned, it might take a bit of time for the RTH to work, but when I didn't have the visual of the craft coming back home, I was really worried.

James

LouisP Posted at 2017-10-30 17:08
There have been many accounts of the spark seeing a cloud or fog as the ground... You were above a cloud and the spark thought that the cloud was the ground.  RTH didn't work because the spark was trying to land on what it thought should be firm ground and could not.   You are lucky that the stick down brought the spark back down. Others have found that they could not descend through a cloud.  This has nothing to do with GPS or barometer... it has everything to do with the downward vision sensor misinterpreting clouds as ground, that is why your altitude said zero even up so high.

Hey LouisP,

This makes perfect sense now. Thank you!!! The last video feed I saw, was when the craft ascending above the cloud and I could see the bright sunshine (yes, it was a magical moment!). You know, I've seen so many Youtube videos that folks flying their drones (and some Sparks too) above the cloud and I wanted to experience that. This is how I got myself into trouble. But LouisP, in this case, are you saying that I should not fly above the cloud or I might risk of losing my Spark since the sensor could misinterpret clouds as ground like you said? I wonder how others managed to land safely when they did this stunt.

James

DJI Elektra Posted at 2017-10-30 19:17
Please note that flying your drone through cloud can cause crash and your drone may find it hard to land. I would not recommend you do the test and reach the cloud again.

DJI Elektra,

Lesson learned. I've done this so foolishly and I will never make the attempt anymore.

James

Folks, is there anyway that I can look closer to the flight log by downloading the blackbox via "DJI Assistant" ? I may not be able to understand everything the report says but at least I can share the details of my experience with the folks who participated in this thread. Now the scary experience is over, I'm very interested in the actual cause of the RTH failure in detail. I'm not trying to proof who's right and who's wrong when replying to my original message, but learn a bit more about drone aviation. Thank you all, especially to Captain Hallmark007 and LouisP. You guys are awesome!! Thanks for educating me.

James

I was able to export the flight log. Please see attached link to my Google drive. Not sure how I can read it though, but perhaps some of the experienced pilot can read it? Thank you.
Flight log

Rawsome Posted at 2017-10-30 15:06
Keep in mind that clouds, fog and sometimes water can throw the sensors of and spark goes a little nuts. Good on you talk your self out of panicking - best thing to do when ever something goes wrong.

Yes Rawsome,

By reading the feedback from others, I've learned my precious (and could be expensive) lesson. I should be more cautious and realize that this drone contains a lot of sensitive electronics within. whew... it was a close call.

James

jcgoobee Posted at 2017-10-30 21:24
Yes Rawsome,

By reading the feedback from others, I've learned my precious (and could be expensive) lesson. I should be more cautious and realize that this drone contains a lot of sensitive electronics within. whew... it was a close call.

Hi,

As tempting as this is, this is the very reason I don’t fly my drone out of line of sight not to mention that you’re not supposed legally anyway in most countries.

Glad to hear you got her home safely.

hallmark007 Posted at 2017-10-30 16:34
Gps takes its reference only from homepoint .
The mapping system used does not have altitude information for the gps to reference. The gps and barometer give altitude from the take off point, but would need the altitude information for the ground your flying over to give you accurate Altitude from gps,
What would AC do if you have gps dropout ,and there are many more practical reasons why barometric pressure is used in aviation to measure altitude.

That's great and all. Keep the barometer and everything. But as long as GPS is working, it can give you an altitude with the same accuracy and reliability it gives you your location in the other two dimensions. In fact it's much more accurate than barometric pressure which is more susceptible to environmental conditions like you mentioned.

djiuser_Kf4iPA3 Posted at 2017-10-30 21:44
That's great and all. Keep the barometer and everything. But as long as GPS is working, it can give you an altitude with the same accuracy and reliability it gives you your location in the other two dimensions. In fact it's much more accurate than barometric pressure which is more susceptible to environmental conditions like you mentioned.

You know, I always make it a habit to check the satellite reception, and ensure there are at least 15 before lifting off.  Like I mentioned earlier, I've learned a valuable lesson this time... haha.. Thanks djiuser_Kf4iPA3.

James

Found an app to record my flight record from DJI go 4, and uploaded to YouTube. Please take a look if you have a chance. Please mute the volume too so it wouldn't be too distracting. Thank you so much.

jcgoobee Posted at 2017-10-30 21:06
DJI Elektra,

Lesson learned. I've done this so foolishly and I will never make the attempt anymore.

Take it easy. I think you can have more nice flights in the future.

DJI Elektra Posted at 2017-10-30 22:13
Take it easy. I think you can have more nice flights in the future.

Thank you, Elektra.

jcgoobee Posted at 2017-10-30 22:25
Thank you, Elektra.

Since you are listed as located in the United States, you can call the FAA and tell them the problems you encountered at 1300 feet.  They keep records of such experiments.

jcgoobee Posted at 2017-10-30 22:06
Found an app to record my flight record from DJI go 4, and uploaded to YouTube. Please take a look if you have a chance. Please mute the volume too so it wouldn't be too distracting. Thank you so much.

Flight Event

Can you upload your flight log to link below just click on link and follow instructions come back here and post your link. This will help .

http://www.phantomhelp.com/LogViewer/Upload/

djiuser_Kf4iPA3 Posted at 2017-10-30 21:44
That's great and all. Keep the barometer and everything. But as long as GPS is working, it can give you an altitude with the same accuracy and reliability it gives you your location in the other two dimensions. In fact it's much more accurate than barometric pressure which is more susceptible to environmental conditions like you mentioned.

* The GPS concept measures altitude relative to the Earth's center; the math involved won't change with tides, weather, erosion, mountains, valleys, temperatures, magnets, construction, or anything else.
* The GPS concept in fact has no idea where the ground or the sea level might be at any given location on the surface of the Earth. It's completely ignorant of geography or the tides. The only geography it knows is a theoretical radius from the center of the Earth that it calls 0 MSL (0 units above mean sea level).

hallmark007 Posted at 2017-10-31 02:37
* The GPS concept measures altitude relative to the Earth's center; the math involved won't change with tides, weather, erosion, mountains, valleys, temperatures, magnets, construction, or anything else.
* The GPS concept in fact has no idea where the ground or the sea level might be at any given location on the surface of the Earth. It's completely ignorant of geography or the tides. The only geography it knows is a theoretical radius from the center of the Earth that it calls 0 MSL (0 units above mean sea level).

The barometer has no concept of the geography either. It has no advantage in that regard over GPS.
But the barometer has more margin of error affected by its surrounding environment, and is therefore less accurate than an altitude measurement by GPS.

I don't know whether Spark uses GPS to measure altitude at all; if it doesn't - it should.

djiuser_Kf4iPA3 Posted at 2017-10-31 03:00
The barometer has no concept of the geography either. It has no advantage in that regard over GPS.
But the barometer has more margin of error affected by its surrounding environment, and is therefore less accurate than an altitude measurement by GPS.

* The Spark concept measures altitude from GPS at the takeoff point, and we call that 0 ATO (zero units above takeoff). From then on, everything in the software of the controller and the aircraft is measured relative to that 0 ATO reference point.
While gps could measure height what is the advantage of it if gps signals can be blocked lost , so your saying introduce a system that is more than likely to fail, again your using an analogy used in phones which is totally different as phones use google maps aerial photography government data surveys etc but this can be unreliable as we see hills bulldozed buildings knocked rivers dried up etc.
Again your trying to reinvent the wheel, gps is not all you crack it up to be, we have all been driving up the new road when we’ve been told to turn back because our satellite system doesn’t have the information about the road.

It is my understanding that you are not supposed to operate above 400 ft in the US.  Experiments like this put lives in jeopardy and cause additional regulations from the government that impact all of us that enjoy this hobby.  Here is a link to the FAA guidelines incase you would like to review them.

https://www.faa.gov/uas/faqs/

hallmark007 Posted at 2017-10-31 04:28
* The Spark concept measures altitude from GPS at the takeoff point, and we call that 0 ATO (zero units above takeoff). From then on, everything in the software of the controller and the aircraft is measured relative to that 0 ATO reference point.
While gps could measure height what is the advantage of it if gps signals can be blocked lost , so your saying introduce a system that is more than likely to fail, again your using an analogy used in phones which is totally different as phones use google maps aerial photography government data surveys etc but this can be unreliable as we see hills bulldozed buildings knocked rivers dried up etc.
Again your trying to reinvent the wheel, gps is not all you crack it up to be, we have all been driving up the new road when we’ve been told to turn back because our satellite system doesn’t have the information about the road.

Once again you put words in my mouth and then argue with yourself...
Did I say to remove the barometer? No, I didn't. Keep it for redundancy and use it when GPS loses satellite LOS. But as long as GPS is working and is locked on enough satellites, it will tell altitude more accurately than a barometer.
And your last remark - what does that have to do with anything? GPS doesn't control any road database, your navigation app does. GPS just provides a 3D coordinate, and a very accurate one. If there's a road missing in your navigation app's database, it has nothing to do with GPS performance.

hallmark007 Posted at 2017-10-31 02:35
Can you upload your flight log to link below just click on link and follow instructions come back here and post your link. This will help .

http://www.phantomhelp.com/LogViewer/Upload/

Got it... see if this works.

Flight Log

Kansas Boy Posted at 2017-10-31 04:56
It is my understanding that you are not supposed to operate above 400 ft in the US.  Experiments like this put lives in jeopardy and cause additional regulations from the government that impact all of us that enjoy this hobby.  Here is a link to the FAA guidelines incase you would like to review them.

https://www.faa.gov/uas/faqs/

Kansas Boy, thanks for the note and I will be more cautious in the future and won't go as high as the rule permits.

djiuser_Kf4iPA3 Posted at 2017-10-31 05:01
Once again you put words in my mouth and then argue with yourself...
Did I say to remove the barometer? No, I didn't. Keep it for redundancy and use it when GPS loses satellite LOS. But as long as GPS is working and is locked on enough satellites, it will tell altitude more accurately than a barometer.
And your last remark - what does that have to do with anything? GPS doesn't control any road database, your navigation app does. GPS just provides a 3D coordinate, and a very accurate one. If there's a road missing in your navigation app's database, it has nothing to do with GPS performance.

My last remark was directed at the fact that you are using the analogy of gps in your phone and you seem to think that if it works in your phone then it should work in a drone the same way but it doesn’t and it won’t.
There is no problem with barometer it doesn’t need redundancy when was the first or last crash from a drone you have seen from a fault in barometer or a bad reading , but we have seen many from gps loss, so you think barometer should be used as redundancy for loss of gps yet gps causes more problems. I think you’ve got that the wrong way around.
I have already explained to you why you won’t get correct readings from gps , I don’t need to explain again. You are just coming at this from a perspective of something that might work without any real testing .
What you need to do is explain how gps would save your craft from crashing getting lost and how it will be much more reliable than the job barometer that is there at the present, take into account of the reliability of barometer against gps and tell us all how it would be more likely to crash or lose your AC using barometer rather than gps.