Eric13 Posted at 2018-6-25 14:02
"The great mystery is:  Why did DJI add this?"

It may not be a mystery after all. I contacted DJI support Facebook/USA with that question.

That's someone at DJI support  (not the best place for accurate information) making a guess as everyone does on the forum.
It's not explaining why it was programmed that way.

It's obvious that within 20 metres you should never need to use RTH but why make the Phantom land where it is if someone (usually a timid new flyer) tries it?

Eric13 Posted at 2018-6-25 12:05
Is GPS kind of inaccurate or is it because the US who run the GPS system are adding a factor that gives GPS receivers a deviating reading?
That's what I heard - the US military itself can pinpoint with GPS pretty exact since they know the deviation factor (which changes constantly)

The USA when they initially started up the second generation GPS system created an offset to the public available system called SA. Selective Availability.
Note - none of this applies to the parallel Military frequency which is unavailable to public.

It was in the mistaken belief that throwing in random errors to the Time Shift / Doppler and data link of the sats would reduce their effectiveness in Terrorists use.
It was a useless gesture as electronic firms along with Govt's such as UK developed what was called Differential GPS receivers.

During SA - the ground stations monitoring the sats would throw up random errors into the command coding. You the public would then receive wandering GPS positional results. BUT if you had a Differential Unit - this would also be receiving signals from Ground based stations dotted around various countries that would 'correct' the Sat data received. This then negated the SA that was imposed by USA.

Later Presidential decree instructed USAF (who were primary operators of the US GPS system) to turn off SA and keep it on standby in case of need in future.

That Decree meant that GPS would be more reliable and expect to give results averaging round the 2 - 3m figure ...
It is incorrect to assume that stopping SA increased accuracy of GPS. It stopped the deliberate induced errors, leaving the GPS signal as it always was in its pure state.

Differential GPS has resulted in a legacy though where it can be still used by facilities to AUGMENT the sat signals to provide as accurate as possible results. In monitoring Sats as they pass .. it can send correcting info to the Rx ... it can in ideal conditions bring results within that 2m limit.

I have a number of GPS units ... my principle Yacht GPS has built in Augmentation software to utilise any Differential signals still available. Plus the ability to be User adjusted. But I paid good money for it ... unlike the $10 chips that are in your P3.

Last point - people go on about having Glonass as well. The only advantage here is the increased number of Sats available for the Unit to read .. that Glonass does not have SA (now a mute point anyway) ... is run completely different to US system.

nigel

Labroides Posted at 2018-6-25 15:34
Is GPS kind of inaccurate or is it because the US who run the GPS system are adding a factor that gives GPS receivers a deviating reading?
That's what I heard - the US military itself can pinpoint with GPS pretty exact since they know the deviation factor (which changes constantly)
What you are describing was known as Selective Accuracy.

Sorry L ....

Consumer available GPS is not as accurate as Military GPS.

They have different frequencies, encryption and data correction from ground tracking.

It is true that Military GPS cannot thread a missile into an air-duct - but it can get darned close enough for other 'guidance' to finish the job.

I may be a Civvy - but have been and will in future be involved in 'Ops' where it will be important.

Nigel

Since we're talking about satellites, I have a question I've never seen asked or discussed here in any detail.  

How, like exactly, does the GPS/Glonass system work on these DJI ACs?  Does anybody actually know?  I know in the older generation, we needed at least six GPS satellites to work.  But I know that other devices such as a tablet need only four to work.  Presumably DJI requires six to promote accuracy.

But how does that transfer to the GPS/Glonass system?  My assumption is that the GPS/Glonass system requires at least four GPS satellites and at least four Glonass satellites to work, but that would be considered a very weak signal.  And 5/5 would be better but still considered weak.  Same with 6/4.  6/6 or 7/5 is a good signal.  This is only my personal theory.  Comments please?  Thanks

solentlife Posted at 2018-6-26 01:24
Sorry L ....

Consumer available GPS is not as accurate as Military GPS.

I chose my words carefully and stand by what I said.
Publicly available GPS is as accurate as military GPS.
Go and see what any surveyors are using.
It's accurate and it's publicly available for civilian use.
It's not a $200 consumer grade device but I didn't use the word consumer there.

I used the word consumer here:
Consumer GPS doesn't have pinpoint accuracy.
You need expensive survey grade GPS to get that.

And what I wrote was addressing the previous post which was about SA.
What I wrote was accurate.

Civilians have access to very accurate GPS equipment if they pay enough for the good gear.
The military does not have some sort of super-accurate secret military grade GPS that the rest of us don't.

Mark The Droner Posted at 2018-6-26 02:07
Since we're talking about satellites, I have a question I've never seen asked or discussed here in any detail.  

How, like exactly, does the GPS/Glonass system work on these DJI ACs?  Does anybody actually know?  I know in the older generation, we needed at least six GPS satellites to work.  But I know that other devices such as a tablet need only four to work.  Presumably DJI requires six to promote accuracy.

You need at least 4 sats to get a 3D fix with a handheld GPS unit but the quality of the position fix depends on the geometry of the satellite spread.
DJI settled for 6 to have a safety margin and minimise the chance of using a poor spread.

The Glonass system also requires a minimum of 4 sats to calculate a position and (I would guess) DJI also want 6.
Your Phantom doesn't mix and match GPS + Glonass.
You need a full set (at least 6) of either for the Phantom to work in P-GPS.
Three of each won't work.

Labroides Posted at 2018-6-26 02:31
You need at least 4 sats to get a 3D fix with a handheld GPS unit but the quality of the position fix depends on the geometry of the satellite spread.
DJI settled for 6 to have a safety margin and minimise the chance of using a poor spread.

Right - so 5/5 won't work at all?  But 6/4 will work, but it'll be weak.  Would you say that's correct?  Thanks

Mark The Droner Posted at 2018-6-26 03:22
Right - so 5/5 won't work at all?  But 6/4 will work, but it'll be weak.  Would you say that's correct?  Thanks

You need 6 of either of the other systems before the Phantom will accept their location data.
Six GPS sats should be fine because you only need 4 for the system to calculate a fix.
Add two more and the spread is sure to be good enough to provide a reliable fix.
Five of each won't allow you to fly in P-GPS but that's just a hypothetical point.

Up in the sky, you will always get more than 6 GPS sats.
You would have to be flying low down in a constrained area to get less.

Labroides Posted at 2018-6-26 03:31
You need 6 of either of the other systems before the Phantom will accept their location data.
Six GPS sats should be fine because you only need 4 for the system to calculate a fix.
Add two more and the spread is sure to be good enough to provide a reliable fix.

Thank-you.

Labroides Posted at 2018-6-26 02:17
I chose my words carefully and stand by what I said.
Publicly available GPS is as accurate as military GPS.
Go and see what any surveyors are using.

Do you think that the Ground Tracking Stations give same quality of error correction to Public available GPS as they do to the Military GPS ? No they do not.
I have access to both Consumer and Military ... and I can tell you straight - I know which is more consistent .. which is more accurate. And it 'aint the consumer GPS.

Survey Units are in fact far more sophisticated than just receiving GPS  signals. They are first referenced to a spot point before commencing the job. This gives the unit a reference error to start. It then during use further references by use of the Chain Boy and distance readings to the 'pole' ... ever onder why it has such a strange carnival style top to it ?

Without that initial reference - the Geo Survey machine has to rely on its high performance software to diagnose and sift through results to arrive at displayed. Its not much use to a fast moving machine ... because of that.
My property has borders that have been disputed a number of times and I have had to pay silly money to have Geo Survey work done ... I always insist on being present at all points and always interested in the gear they use.

I was also working on Seismic Vessel for a while - CGG of France ... who designed a system called Syledis. This used transponders placed on rigs / shore points etc. It gave accurate position to within 10cm ... but was only short range. We could tell when the rig 'skidded' its tower etc.
GPS was a back-up and because CGG was a company involved with French Military, the Seismic cable is an extra long version (3km's) of a Sub detection cable used by Navy (usually only 200 - 400m long) - we had both generally available GPS and a Military set. We were under strict orders to secure the military set. The GPS was more accurate on the military and far more consistent but it still did not come close to Syledis in any form.

Sorry ... bad subject to pick with me ...

Nigel

I tested the RTH feature with the P4A+ today, ensuring it is set on the app to RTH in the event of RC disconnect. To test this i put the drone at 60 meters high and 180 meters away so I could see it, turned off the RC and waited 30 seconds to see if it kicks in and nothing the drone just stayed there hovering did not move so powered back on the RC and flew it back double checked the setting and there it was set to RTH if disconnect from RC. Seems that this feature failed, I had this numerous times on the Mavic Pro and it always worked. Like to test these things so that If this ever happens I can trust in my drone to return home.

Suren Posted at 2018-6-26 09:07
I tested the RTH feature with the P4A+ today, ensuring it is set on the app to RTH in the event of RC disconnect. To test this i put the drone at 60 meters high and 180 meters away so I could see it, turned off the RC and waited 30 seconds to see if it kicks in and nothing the drone just stayed there hovering did not move so powered back on the RC and flew it back double checked the setting and there it was set to RTH if disconnect from RC. Seems that this feature failed, I had this numerous times on the Mavic Pro and it always worked. Like to test these things so that If this ever happens I can trust in my drone to return home.

Please upload and post the link:  http://www.phantomhelp.com/LogViewer/Upload/

solentlife Posted at 2018-6-26 07:22
Do you think that the Ground Tracking Stations give same quality of error correction to Public available GPS as they do to the Military GPS ? No they do not.
I have access to both Consumer and Military ... and I can tell you straight - I know which is more consistent .. which is more accurate. And it 'aint the consumer GPS.

Sorry ... bad subject to pick with me ...
That's one thing I can agree with you on.
There's nothing anyone can tell you about GPS

But if I was talking to anyone else, I'd still stand by what I said.
Publicly available GPS is as accurate as military GPS.
Whether the military has something more sophisticated, is irrelevant.
It's not more accurate.

Mark The Droner Posted at 2018-6-26 11:03
Please upload and post the link:  http://www.phantomhelp.com/LogViewer/Upload/

Unfortunately his flight data would only show what happened before switching off the controller.
What happened after that would be the interesting part.

Suren Posted at 2018-6-26 09:07
I tested the RTH feature with the P4A+ today, ensuring it is set on the app to RTH in the event of RC disconnect. To test this i put the drone at 60 meters high and 180 meters away so I could see it, turned off the RC and waited 30 seconds to see if it kicks in and nothing the drone just stayed there hovering did not move so powered back on the RC and flew it back double checked the setting and there it was set to RTH if disconnect from RC. Seems that this feature failed, I had this numerous times on the Mavic Pro and it always worked. Like to test these things so that If this ever happens I can trust in my drone to return home.

That's sounds unusual.  The first time I've heard of it
Just the one test?
You definitely switched off the controller - not just the display?
What happens when you initiate RTH with the screen or controller button?

Labroides Posted at 2018-6-26 14:38
Unfortunately his flight data would only show what happened before switching off the controller.
What happened after that would be the interesting part.

I wanted to see where the homepoint was recorded

Mark The Droner Posted at 2018-6-26 15:29
I wanted to see where the homepoint was recorded

It's unlikely the Phantom could get 60m high and 180 metres out without getting GPS and recording a homepoint.
But if it did happen, I guess that could explain the failure to come home.

Labroides Posted at 2018-6-26 15:33
It's unlikely the Phantom could get 60m high and 180 metres out without getting GPS and recording a homepoint.
But if it did happen, I guess that could explain the failure to come home.

Once I had my P2V mark a homepoint behind me in the woods several hundred feet.  And I didn't notice it until several months later when I went over the log.  It only happened that once that I know of out of several hundred flights.  I have no explanation.  Also, sometimes when you're watching the AC and it's in RTH, it looks like it's hovering but actually it's slowly rising to min RTH height first, so you think it's not RTHing when actually it is.  But the log wouldn't help us there.  

Mark The Droner Posted at 2018-6-26 11:03
Please upload and post the link:  http://www.phantomhelp.com/LogViewer/Upload/

I looked at the log and it just ends when the RC disconnects does not say much about why it did not work

Labroides Posted at 2018-6-26 14:38
Unfortunately his flight data would only show what happened before switching off the controller.
What happened after that would be the interesting part.

That is correct, already looked at log, it should have intiated rth when rc disconnected

Labroides Posted at 2018-6-26 14:40
That's sounds unusual.  The first time I've heard of it
Just the one test?
You definitely switched off the controller - not just the display?

Put the rc completely off, when i manually press the rth it works though

Mark The Droner Posted at 2018-6-26 15:29
I wanted to see where the homepoint was recorded

Homepoint was recorded correctly because when rc was powered on again i hit the rth manually and it came home, as I said was only 60 meters high and 150 meters away to test this

Labroides Posted at 2018-6-26 15:33
It's unlikely the Phantom could get 60m high and 180 metres out without getting GPS and recording a homepoint.
But if it did happen, I guess that could explain the failure to come home.

full gps 12 sats and my little friend said "your home point has been recorded"

What is your min RTH height, what was your actual height?  Could you post the log?

Labroides Posted at 2018-6-26 14:32
Sorry ... bad subject to pick with me ...
That's one thing I can agree with you on.
There's nothing anyone can tell you about GPS

Your choice L ...

Having had to put my life on the reliability and accuracy of GPS on more than one occasion - I know the reality of Military vs Publicly available GPS.

Its not only Military Personnel that go into areas of extreme ... want to trust a Garmin or TomTom in that ???

For practical purposes and the algorithms used in modern consumer units - they perform excellently. The user is unaware of the variables it is discarding etc.

But never confuse that with the higher precision / better algorithm military units.

Nigel

Labroides Posted at 2018-6-26 14:38
Unfortunately his flight data would only show what happened before switching off the controller.
What happened after that would be the interesting part.

What about the on-board data file ?

Nigel

Suren Posted at 2018-6-26 19:24
Put the rc completely off, when i manually press the rth it works though

Something definitely not right now ... there should be no way the RTH button can work if Controller is powered OFF. Where would the signal be generated from ??
Now looking at your later post - I think you meant to say :

RTH did not work when Controller switched off ... but once you powered controller again - it worked ??

Nothing wrong with that - the AC records the RTH position - not the controller or tablet. Just strange that the AC did not RTH when you powered down ...

Have you re-checked all SMART and FAILSAFE settings .. ?? Made sure you really have RTH as the set option and not just Hover ?

Nigel

solentlife Posted at 2018-6-27 02:19
Something definitely not right now ... there should be no way the RTH button can work if Controller is powered OFF. Where would the signal be generated from ??
Now looking at your later post - I think you meant to say :

I think you need to read post #171 again, coupled with post #182.

Suren Posted at 2018-6-26 19:25
Homepoint was recorded correctly because when rc was powered on again i hit the rth manually and it came home, as I said was only 60 meters high and 150 meters away to test this

This is part of the problem.  You never said it was 150 meters away.  You said it was 180 meters away.  Now you say it was 150 meters away?  Was it 150 or 180?  Maybe it was actually only 120 meters away?  How do we know?  We have to take your word for it.  

Why not post the log?  It takes less than 30 secs.  http://www.phantomhelp.com/LogViewer/Upload/

solentlife Posted at 2018-6-27 02:15
Your choice L ...

Having had to put my life on the reliability and accuracy of GPS on more than one occasion - I know the reality of Military vs Publicly available GPS.

Yeah Nigel ... So that's compasses and GPS now
No-one else has ever used GPS before
You're the only person that knows anything about GPS
... as you showed in post #162 where you posted a whole essay to waffle about what had already been answered in two lines in post #160 hours earlier.
We bow to your amazing knowledge

Mark The Droner Posted at 2018-6-27 03:11
I think you need to read post #171 again, coupled with post #182.

No .. I have already .. and he posted 181 with the powered off but RTH working !! Which the three posts taken as posted do not add up.

Also it seems to indicate that he's surprised RTH was still available when he powered back up.

I just have a feeling that somethings not being admitted to ??

Just trying to think why the AC would hover and not RTH. Only instance I have seen that is when my P3 had the tablet shut down but RC was still on and connected. I let go sticks so AC would not fly off into unknown ... it hovered ... I hit RTH on RC and she came back ... while I restarted tablet.
If the RC had gone off - the AC would have RTH'd. Unless option to Hover had been selected instead.

Nigel

As we all know, these things turn out to be pilot error 98% of the time - or more.  Some solid info would be nice.  

Labroides Posted at 2018-6-27 03:17
Yeah Nigel ... So that's compasses and GPS now
No-one else has ever used GPS before
You're the only person that knows anything about GPS

Yes L ... sure ..... you know best.

Iran - Iraq Conflict
Gulf War 1
Afghanistan

Yes .. and a few other minor conflicts as well ... you'd be surprised how much a Civilian like me has to do in such ... my life is worth more than trusting to your interpretations. Sorry if blunt - but its fact.
I didn't see you on the front line xxxx xxxxx   xxxxx  from Kuwait ... xxxxx me at the small crossing point ..

Didn't see you setting up the fuel dumps for forward units...

Lives depended on it .... like I say - didn't see you there.

Nigel

Edited out confidential data

''as you showed in post #162 where you posted a whole essay to waffle about what had already been answered in two lines in post #160 hours earlier.''

Go back and he actually asked more than just SA .. he mentioned the Airport Augmentation as well ... my 'waffle' as you put it supplies the answer and explanation of why / what it was ... some people like to know more than just a piffling bit of a very interesting subject.

Nigel

Just to nail the lid on your coffin L :

''We bow to your amazing knowledge''

You can't even get the name right !!
Dunno where you get the term Selective Accuracy from because its always been Selective Availability .... because it was not only the degradation of info - but also the ability to switch it off .... Availability.

Please see :

https://www.gps.gov/systems/gps/modernization/sa/

There is one aspect of SA that is still there ... which is not mentioned in the Gov Doc for good reason - switch Off capbility ... plus degradation is still possible - you only have to ask users in vicinity of Military Exercises ... check out Marine Notice to Mariners etc. warning of GPS 'disruption' not only by local means but also via the actual sats themselves.

Bye L ... you astound me at times ...

Nigel

solentlife Posted at 2018-6-27 03:35
Yes L ... sure ..... you know best.

Iran - Iraq Conflict

You can't even get the name right !!
Dunno where you get the term Selective Accuracy from because its always been Selective Availability
Big deal ... it was 20 years ago and I made a minor slip up on the wording.
Enjoy it, you don't get that opportunity very often.

Yes .. and a few other minor conflicts as well ... you'd be surprised how much a Civilian like me has to do in such ... my life is worth more than trusting to your interpretations. Sorry if blunt - but its fact.
I didn't see you on the front line xxxx xxxxx   xxxxx  from Kuwait ... xxxxx me at the small crossing point ..

Didn't see you setting up the fuel dumps for forward units...
Lives depended on it .... like I say - didn't see you there.

I didn't see you anywhere that I was either
But now I understand why you're the only person that knows anything about GPS.

ps ... WTF does any of that have to do with anything on this forum?

Suren Posted at 2018-6-26 19:24
Put the rc completely off, when i manually press the rth it works though

Put the rc completely off, when i manually press the rth it works though .
Can you clarify the timing for this?
Is there any chance that you switched off and a couple of seconds later pressed RTH and saw the drone begin to RTH .... as it would normally a few seconds after switching off?

Have you tested this multiple times to see if it's an ongoing and repeatable issue  or just a one-off?

Lol 3 pages and a bit in 2½ year now soon the same on a day.
What does civil contra military precision have to do with RTH and the strange behavior about not performing a RTH within 20 meters in the first place?
If the aircraft know it has moved 20 meters away from the HP, for sure it should be able to fly 20 meters back whether i would take detour to RTH altitude or not on the way, just imo.

solentlife Posted at 2018-6-27 02:19
Something definitely not right now ... there should be no way the RTH button can work if Controller is powered OFF. Where would the signal be generated from ??
Now looking at your later post - I think you meant to say :

Hi Nigel, RTH did not work when Controller switched off ... but once you powered controller again - it worked ?? Yes you are correct guess my wording came out wrong

solentlife Posted at 2018-6-27 02:19
Something definitely not right now ... there should be no way the RTH button can work if Controller is powered OFF. Where would the signal be generated from ??
Now looking at your later post - I think you meant to say :

All failsafe settings have been checked and verfied as well