The BIG GNSS (GPS) Issue -- The Constructive Thread

[GadgetGuy]

Yeah, they swapped GLONASS for Beidou. There are other variables that can be at play, but for now I'm going to try to keep track of the number of Beidou stats and see if it affects acquisition time.

Launching before GPS will not *cause* a fly away, but it is good practice. Also, if you're moving around, driving especially, a previous HP may be in memory so if something does go wrong while you're in the air and it hasn't updated, it can cause to go to the wrong home point. And, what if it captures the HP when you're above a tree, water, etc?

Perhaps you didn't understand that flying straight up over the launch point, and waiting for the Home Point to be set from 96 feet above it, results in the Home Point being set in the exact same location as you launched from, which is all that happens on the ground, when the Home Point is set there. Moving away from the Home point won't change the Home Point unless you reset it. Launching straight up can't be done under a tree. If you are launching straight up over water, that's on you! The Home point will still be over water, no matter what elevation it was set at.

 

[Meta4]

if you're moving around, driving especially, a previous HP may be in memory so if something does go wrong while you're in the air and it hasn't updated, it can cause to go to the wrong home point.

Homepoints only stay in memory until you power off the drone.
When you power on again, there is no homepoint until the drone gets good GPS and records a new one.

And, what if it captures the HP when you're above a tree, water, etc?

That's on you to manage.
You can either keep the drone above the area where you want the homepoint to be, or you can reset the homepoint to the controller's location.

 

[mobilehomer]

Homepoints only stay in memory until you power off the drone.
When you power on again, there is no homepoint until the drone gets good GPS and records a new one.

That's on you to manage.
You can either keep the drone above the area where you want the homepoint to be, or you can reset the homepoint to the controller's location.

Home point also resets if you land and motors off. Once you become airborne it acquires a new HP.

 

[GadgetGuy]

Home point also resets if you land and motors off. Once you become airborne it acquires a new HP.

Actually, it is rearming of the motors that sets the new HP and resets the elevation to zero. It's one way to bypass the 500m altitude limit, if you can still maintain RC control, by serially landing, disarming the motors, and then rearming them, so you can mountain climb, in 500m intervals of elevation.

 

[Paradigm]

Perhaps you didn't understand that flying straight up over the launch point, and waiting for the Home Point to be set from 96 feet above it, results in the Home Point being set in the exact same location as you launched from, which is all that happens on the ground, when the Home Point is set there. Moving away from the Home point won't change the Home Point unless you reset it. Launching straight up can't be done under a tree. If you are launching straight up over water, that's on you! The Home point will still be over water, no matter what elevation it was set at.

With the M2P, I've acquired a HP, flown the battery, then landed. I've driven to a completely new location. I've taken off without establishing a new HP, and on my map, the old homepoint, over a mile away, is still indicated as my
homepoint.

It has done this if I've swapped the battery or not.

When I get enough satellites, the location of the drone while it's in the air, is established as the new HP.

Now, maybe it's not a "real HP" as I've never tried RTH, but on my map, it shows the HP as being the previous locked location.

There are also various videos on YT with people experiencing this same phenomenon. And also the drone acquiring an inaccurate HP, which is a different topic, I admit.

I would be curious to find out if DJI has an officially recommend protocol as to taking off before a HP is established...

 

[GadgetGuy]

With the M2P, I've acquired a HP, flown the battery, then landed. I've driven to a completely new location. I've taken off without establishing a new HP, and on my map, the old homepoint, over a mile away, is still indicated as my
homepoint.

It has done this if I've swapped the battery or not.

When I get enough satellites, the location of the drone while it's in the air, is established as the new HP.

Now, maybe it's not a "real HP" as I've never tried RTH, but on my map, it shows the HP as being the previous locked location.

There are also various videos on YT with people experiencing this same phenomenon. And also the drone acquiring an inaccurate HP, which is a different topic, I admit.

I would be curious to find out if DJI has an officially recommend protocol as to taking off before a HP is established...

Ascending straight up to acquire the Home Point setting at 96’ is far different than launching and flying away without a new home point being set, which what you are describing doing! Definitely not advised! Your home point is being set a different location than where you launched from, unlike my recommendation.

 

[Meta4]

With the M2P, I've acquired a HP, flown the battery, then landed. I've driven to a completely new location. I've taken off without establishing a new HP, and on my map, the old homepoint, over a mile away, is still indicated as my
homepoint.

I'm not sure what you've seen, but your drone doesn't store any location data after you've powered it off.
The homepoint is lost and when you power up next time, the data fields for the home point are blank until it acquires GPS again.
If you can remember a flight where you've seen this, and post the flight data, I can show you confirmation of this.
There is no way that your drone can store a home point so that it's still there when you power up again.

There are also various videos on YT with people experiencing this same phenomenon. And also the drone acquiring an inaccurate HP,

Youtube is full of misinformation spread by people with little understanding of what they are doing.

I would be curious to find out if DJI has an officially recommend protocol as to taking off before a HP is established...

DJI don't have anything like that.
But if you want to launch before recording a home point (and there are situations where this has to be done), if you want a home point close to you so that a failsafe RTH would bring the drone back, either:

• Hover above the launch point until you get GPS and a homepoint, or
• Fly out and after the drone records a homepoint where it is, reset the homepoint to where you are.

 

[sar104]

With the M2P, I've acquired a HP, flown the battery, then landed. I've driven to a completely new location. I've taken off without establishing a new HP, and on my map, the old homepoint, over a mile away, is still indicated as my
homepoint.

It has done this if I've swapped the battery or not.

When I get enough satellites, the location of the drone while it's in the air, is established as the new HP.

Now, maybe it's not a "real HP" as I've never tried RTH, but on my map, it shows the HP as being the previous locked location.

There are also various videos on YT with people experiencing this same phenomenon. And also the drone acquiring an inaccurate HP, which is a different topic, I admit.

I would be curious to find out if DJI has an officially recommend protocol as to taking off before a HP is established...

As @Meta4 pointed out, the previous home point is not stored. Take a look at the flight log for that second flight, and you will be able to see the home point data.

 

[Paradigm]

I'm not sure what you've seen, but your drone doesn't store any location data after you've powered it off.
The homepoint is lost and when you power up next time, the data fields for the home point are blank until it acquires GPS again.
If you can remember a flight where you've seen this, and post the flight data, I can show you confirmation of this.
There is no way that your drone can store a home point so that it's still there when you power up again.

Youtube is full of misinformation spread by people with little understanding of what they are doing.

DJI don't have anything like that.
But if you want to launch before recording a home point (and there are situations where this has to be done), if you want a home point close to you so that a failsafe RTH would bring the drone back, either:

• Hover above the launch point until you get GPS and a homepoint, or
• Fly out and after the drone records a homepoint where it is, reset the homepoint to where you are.

I will try to recreate this, perhaps today. I'll report back.

 

[Meta4]

I will try to recreate this, perhaps today. I'll report back.

Just post the data from the flight where you made this observation and that will show what happened.

 

[ianwood]

We're not there yet. Today, it took 14 minutes on the new firmware to get a home point from a cold start! Granted it was purposefully a difficult spot (a few tall buildings) but an Inspire 2 doesn't take 14 minutes there. Even a Phantom 2 could get a home point in less time!

Warm start in other spots around town were nearly instant. So at least that is fixed.

 

[ZoneAerial]

We're not there yet. Today, it took 14 minutes on the new firmware to get a home point from a cold start! Granted it was purposefully a difficult spot (a few tall buildings) but an Inspire 2 doesn't take 14 minutes there. Even a Phantom 2 could get a home point in less time!

Warm start in other spots around town were nearly instant. So at least that is fixed.

Hopefully you report that to DJI. Their M3 has worse satellite acquisition than a Phantom 2. AFTER the “fix”.

 

[Meta4]

We're not there yet. Today, it took 14 minutes on the new firmware to get a home point from a cold start! Granted it was purposefully a difficult spot (a few tall buildings) but an Inspire 2 doesn't take 14 minutes there. Even a Phantom 2 could get a home point in less time!

Warm start in other spots around town were nearly instant. So at least that is fixed.

Poor time in an obstructed location isn't conclusive proof.
Depending on the amount of sky that's blocked and the number and distribution of sats that were still visible, maybe even 14 minutes might not indicate a problem with the drone.
Testing in the open is what's needed to be certain.

How much sky was unobstructed at that location?

 

[ZoneAerial]

Poor time in an obstructed location isn't conclusive proof.
Depending on the amount of sky that's blocked and the number and distribution of sats that were still visible, maybe even 14 minutes might not indicate a problem with the drone.
Testing in the open is what's needed to be certain.

How much sky was unobstructed at that location?

The thing is, I bet, an Air2S or even a Mini 2 would get GPS lock in that “partially” obstructed area in like 30 seconds. But their most expensive flagship consumer drone? Not so much…

 

[Meta4]

The thing is, I bet, an Air2S or even a Mini 2 would get GPS lock in that “partially” obstructed area in like 30 seconds. But their most expensive flagship consumer drone? Not so much…

Maybe it would and maybe it wouldn't.
If enough of the skyview is blocked, any GPS receiver will have difficulty.
I was just pointing out that we can't tell much from a single trial in an obstructed location with no information to understand how much the obstruction might have been affecting GPS.

 

[ZoneAerial]

Maybe it would and maybe it wouldn't.
If enough of the skyview is blocked, any GPS receiver will have difficulty.
I was just pointing out that we can't tell much from a single trial in an obstructed location with no information to understand how much the obstruction might have been affecting GPS.

Trust me- even after yesterday’s failed promised fix, my Air2S acquires satellites many times faster in the same locations. The Mavic 3 is a lemon.

 

[ianwood]

The thing is, I bet, an Air2S or even a Mini 2 would get GPS lock in that “partially” obstructed area in like 30 seconds. But their most expensive flagship consumer drone? Not so much…

This. I have tons of experience flying in areas where the sky is partially obstructed. This would have been maybe 1-2 minutes longer than normal on any other DJI craft. I will be doing more flights tomorrow.

 

[ZoneAerial]

This. I have tons of experience flying in areas where the sky is partially obstructed. This would have been maybe 1-2 minutes longer than normal on any other DJI craft. I will be doing more flights tomorrow.

This campaign needs you now

DJI Needs to Respond to Mavic 3 Owners who still have serious GPS issues

chng.it

 

[mavic53]

Now locking on to satellites in seconds as per Mavic 2

 

[ianwood]

Further data points. Definitely some lingering around the 9-12 sat point still but it could be yesterday's test was an outlier. Today's cold start in a more hospitable location (only small portions of the sky blocked) was about 2 minutes. Did a warm start on a street with plenty of tall buildings all around and it did admirably. About 1 min.

All in, if the 15 minutes I had yesterday was an outlier, I'm good. Stability is still phenomenal.