GPS new Firmware

[Crashdummytest]

I have to chuckle about everyone's impatience with GPS acquisition times! Travel back with me to the year 1989. The average time to acquire enough satellites to calculate a 2D position (3 satellites) or a 3D position (4 satellites) was 10 to 15 minutes! Almost every single time!! I remember, as a young sailor in the us navy, standing in a Westwood marine supply store in Long Beach, CA in December 1989, as a Garmin representative was setting out a table with a brand new item that just came to market, a handheld GPS marine plotting device. It was a Garmin GPS model 45, and it could tell you your location anywhere in the world. That was truly an amazing concept back then. The price was approximately $500, pretty steep for a device with unknown capabilities. Well I went ahead and took the plunge and bought one. My ship, a Guided Missile Frigate was departing for a 6 month tour of the Western Pacific later that week and I wanted to see my ship's location without having to always go find the quartermaster in plot. Despite it not having a moving map display, it was amazingly accurate, even with the commercial Selective Availability (SA) (intentional degradation) of public GPS signals (implemented for national security reasons), applied. The other issue we had to contend with back then was the Dilution of precision (DOP), or geometric dilution of precision (GDOP) as there were only 24 satellites orbiting earth, with only 12, at any one time, viewable per hemisphere. But even then, my accuracy was always less than 50 meters, sometimes less than 30 meters. My commanding officer saw me out on the bridge wing one morning using my Garmin, he was very curious as to its capabilities, as he had never seen one before. My GPS 45 had impressed him enough that after we got back from deployment 7 months later, the Skipper insisted we get GPS navigation for the ship! Long story short, my ship, the USS Reuben James FFG-57, was the first frigate to receive the new technology, and we were featured in GPS World magazine for that accomplishment. As I was also the ship's Intelligence photographer, some of my photographs were included in that issue of the magazine, so I also got some by-lines. I guess it's all about perspective! A minute or two is no big deal!!View attachment 142641

But still, it takes just seconds with my other drones. Only the M3 is of 1989 standard .

 

[Deleted member 154956]

A few years later, Garmin came out with an aviation handheld, the GPSMAP 195, that I could mount to my airplane yoke, and fly around with unheard of precision and situational awareness. Satellite acquisition times were vastly improved by then (1995ish), but I still had to place an external antenna out on the cockpit dashboard by the windscreen to get reliable signals.

 

[surfbird]

But still, it takes just seconds with my other drones. Only the M3 is of 1989 standard .

It almost feels like the A-GPS info isn't transfered from the phone. I don't have a mavic3 (yet), if someone has the knowledge, it could be a good idea to check the AGPS status in the drone.
If it's not updated, a gps-fix will take a long time, as it did in 1989

 

[ZoneAerial]

It almost feels like the A-GPS info isn't transfered from the phone. I don't have a mavic3 (yet), if someone has the knowledge, it could be a good idea to check the AGPS status in the drone.
If it's not updated, a gps-fix will take a long time, as it did in 1989

It is bad with the RC pro as well, so it isn’t just an issue with the phone’s a-gps.

 

[surfbird]

It is bad with the RC pro as well, so it isn’t just an issue with the phone’s a-gps.

It works exactly the same way, the agps isn't really the standard agps-format.
It's an byte array containing timestamp and lat/lon. At least last time I checked. If faulty info gets transfered it can really mess up the gps fix.

public static synchronized DataFlycSendAgpsData getInstance() {
DataFlycSendAgpsData dataFlycSendAgpsData;
synchronized (DataFlycSendAgpsData.class) {
if (f15076a == null) {
f15076a = new DataFlycSendAgpsData();
}
dataFlycSendAgpsData = f15076a;
}
return dataFlycSendAgpsData;
}

public void a(byte[] bArr) {
this.f15078c = bArr;
}

public void a(short s, byte b2, byte b3) {
System.arraycopy(c.b(s), 0, this.f15077b, 0, 2);
this.f15077b[2] = (byte) ((b3 << 7) | (b2 & Byte.MAX_VALUE));
}

@override // dji.midware.data.manager.P3.p
protected void doPack() {
int length = this.f15077b.length + this.f15078c.length;
if (this._sendData == null || this._sendData.length != length) {
this._sendData = new byte[length];
}
System.arraycopy(this.f15077b, 0, this._sendData, 0, this.f15077b.length);
System.arraycopy(this.f15078c, 0, this._sendData, this.f15077b.length, this.f15078c.length);
}

 

[ZoneAerial]

It works exactly the same way, the agps isn't really the standard agps-format.
It's an byte array containing timestamp and lat/lon. At least last time I checked. If faulty info gets transfered it can really mess up the gps fix.

public static synchronized DataFlycSendAgpsData getInstance() {
DataFlycSendAgpsData dataFlycSendAgpsData;
synchronized (DataFlycSendAgpsData.class) {
if (f15076a == null) {
f15076a = new DataFlycSendAgpsData();
}
dataFlycSendAgpsData = f15076a;
}
return dataFlycSendAgpsData;
}

public void a(byte[] bArr) {
this.f15078c = bArr;
}

public void a(short s, byte b2, byte b3) {
System.arraycopy(c.b(s), 0, this.f15077b, 0, 2);
this.f15077b[2] = (byte) ((b3 << 7) | (b2 & Byte.MAX_VALUE));
}

@override // dji.midware.data.manager.P3.p
protected void doPack() {
int length = this.f15077b.length + this.f15078c.length;
if (this._sendData == null || this._sendData.length != length) {
this._sendData = new byte[length];
}
System.arraycopy(this.f15077b, 0, this._sendData, 0, this.f15077b.length);
System.arraycopy(this.f15078c, 0, this._sendData, this.f15077b.length, this.f15078c.length);
}

Well, here’s to hoping there will be a hotfix if enough people complain- (text deleted by moderator)

 

[NCDroneEdits]

Mine is still the same as before the update - takes a while.

 

[Paul2660]

I have to chuckle about everyone's impatience with GPS acquisition times! Travel back with me to the year 1989. The average time to acquire enough satellites to calculate a 2D position (3 satellites) or a 3D position (4 satellites) was 10 to 15 minutes! Almost every single time!! I remember, as a young sailor in the us navy, standing in a Westwood marine supply store in Long Beach, CA in December 1989, as a Garmin representative was setting out a table with a brand new item that just came to market, a handheld GPS marine plotting device. It was a Garmin GPS model 45, and it could tell you your location anywhere in the world. That was truly an amazing concept back then. The price was approximately $500, pretty steep for a device with unknown capabilities. Well I went ahead and took the plunge and bought one. My ship, a Guided Missile Frigate was departing for a 6 month tour of the Western Pacific later that week and I wanted to see my ship's location without having to always go find the quartermaster in plot. Despite it not having a moving map display, it was amazingly accurate, even with the commercial Selective Availability (SA) (intentional degradation) of public GPS signals (implemented for national security reasons), applied. The other issue we had to contend with back then was the Dilution of precision (DOP), or geometric dilution of precision (GDOP) as there were only 24 satellites orbiting earth, with only 12, at any one time, viewable per hemisphere. But even then, my accuracy was always less than 50 meters, sometimes less than 30 meters. My commanding officer saw me out on the bridge wing one morning using my Garmin, he was very curious as to its capabilities, as he had never seen one before. My GPS 45 had impressed him enough that after we got back from deployment 7 months later, the Skipper insisted we get GPS navigation for the ship! Long story short, my ship, the USS Reuben James FFG-57, was the first frigate to receive the new technology, and we were featured in GPS World magazine for that accomplishment. As I was also the ship's Intelligence photographer, some of my photographs were included in that issue of the magazine, so I also got some by-lines. I guess it's all about perspective! A minute or two is no big deal!!View attachment 142641

I agree the GPS-45 was an amazing device. Game changer. Put Garmin in first place and they are still there. Interesting story also about your captain. Thanks for sharing.

Paul

 

[GusTek]

That is the hope. I agree it appeared after December 10th update. The question I guess is why, and why are they sticking with it this way. One member had all kinds of conspiratorial thoughts on why. I'm not convinced. At one point he said "No other drone from DJI has ever dine this" - actually if you google DJI drone slow/poor satellite acquisition, you will see that almost EVERY model, particularly folding droned since the original Mavic have had this at one time or another.

This one has it pretty bad though. I was really hoping for a fix. This business about cold boot of the GPS Module makes sense. No one sees it on successive battery flights, just the first one. Also, any flights I do within an hour or so grab sats right away. It's as if the module keeps a list for quite a while, then eventually clears volatile memory and a full reboot is required. There's also systems that use assist from an internet connection to get a base list, but I can't claim to know what is going on with Mavic 3. We could go back to the firmware it launched with and get faster locks. Who's with me?
*crickets*

You just might be into something with the slow sar connections and the internet.. I’ve got the RC PRO and it takes a bit to get connected (I rarely connect it to via hotspot). I wonder if the DJI RC-N1 takes the same time?

 

[Meta4]

You just might be into something with the slow sar connections and the internet.. I’ve got the RC PRO and it takes a bit to get connected (I rarely connect it to via hotspot). I wonder if the DJI RC-N1 takes the same time?

What you are asking isn't clear, but if you are asking about the time the RC-N1 takes to get GPS ..... it doesn't have a GPS receiver, so it never gets GPS.
Your RC pro has a GPS receiver built-in to the android tablet.

 

[RocketBrew]

My understanding is that the aircraft does its own thing for GNSS. No GNSS data is sent to the aircraft from the controller or the from the device connected to the controller (if it has GNSS capabilities).

The app running on the device can use GNSS data from the device for certain functions, but I don't believe it is required and not communicated to the aircraft.

It is bad with the RC pro as well, so it isn’t just an issue with the phone’s a-gps.

 

[Crashdummytest]

My understanding is that the aircraft does its own thing for GNSS. No GNSS data is sent to the aircraft from the controller or the from the device connected to the controller (if it has GNSS capabilities).

The app running on the device can use GNSS data from the device for certain functions, but I don't believe it is required and not communicated to the aircraft.

My M3 has 112 flights. A few times I have rebooted just the aircraft due to impatience and it seems to help (on the lock issue).

 

[Point Zero]

I have to chuckle about everyone's impatience with GPS acquisition times! Travel back with me to the year 1989. The average time to acquire enough satellites to calculate a 2D position (3 satellites) or a 3D position (4 satellites) was 10 to 15 minutes! Almost every single time!! I remember, as a young sailor in the us navy, standing in a Westwood marine supply store in Long Beach, CA in December 1989, as a Garmin representative was setting out a table with a brand new item that just came to market, a handheld GPS marine plotting device. It was a Garmin GPS model 45, and it could tell you your location anywhere in the world. That was truly an amazing concept back then. The price was approximately $500, pretty steep for a device with unknown capabilities. Well I went ahead and took the plunge and bought one. My ship, a Guided Missile Frigate was departing for a 6 month tour of the Western Pacific later that week and I wanted to see my ship's location without having to always go find the quartermaster in plot. Despite it not having a moving map display, it was amazingly accurate, even with the commercial Selective Availability (SA) (intentional degradation) of public GPS signals (implemented for national security reasons), applied. The other issue we had to contend with back then was the Dilution of precision (DOP), or geometric dilution of precision (GDOP) as there were only 24 satellites orbiting earth, with only 12, at any one time, viewable per hemisphere. But even then, my accuracy was always less than 50 meters, sometimes less than 30 meters. My commanding officer saw me out on the bridge wing one morning using my Garmin, he was very curious as to its capabilities, as he had never seen one before. My GPS 45 had impressed him enough that after we got back from deployment 7 months later, the Skipper insisted we get GPS navigation for the ship! Long story short, my ship, the USS Reuben James FFG-57, was the first frigate to receive the new technology, and we were featured in GPS World magazine for that accomplishment. As I was also the ship's Intelligence photographer, some of my photographs were included in that issue of the magazine, so I also got some by-lines. I guess it's all about perspective! A minute or two is no big deal!!View attachment 142641

I get what you are saying but for me its like this. My M2P would always (in three years of heavy use) acquire multiple satellite locks in around 30 seconds. I never once had to wait to take off because the signal strength was too weak. Now DJI have changed how their system works and the wait is ridiculous. Why advertise a drone with such a great battery life when you need to waste 10% of it or more waiting for a homepoint to be recorded when a drone three years older still has zero problems in this area? Nobody wants to pay this kind of money for a drone that performs much than its 3 year older predecessor. Its that simple

 

[VenomXts]

People used to use slide rules as well and personal computers never existed in the distant past. Your point is irrelevant although you are trying to be endearing.
For some people it takes a lot longer than a minute or 2. It isn’t acceptable.

missing a shot because of waiting for GPS is not acceptable if your used to having to wait x with every other drone,.

 

[UAVquadman]

I'm finding no difference in GPS acquisition time and I'm still getting the GPS SIGNAL WEAK issue....Curious as to what others think.

that's why I returned mine. I was in shock how long it takes.

 

[UAVquadman]

I get what you are saying but for me its like this. My M2P would always (in three years of heavy use) acquire multiple satellite locks in around 30 seconds. I never once had to wait to take off because the signal strength was too weak. Now DJI have changed how their system works and the wait is ridiculous. Why advertise a drone with such a great battery life when you need to waste 10% of it or more waiting for a homepoint to be recorded when a drone three years older still has zero problems in this area? Nobody wants to pay this kind of money for a drone that performs much than its 3 year older predecessor. Its that simple

the Mavic 3 Sucks!!! LOL.... I thought mine was defective..
.

 

[VenomXts]

the Mavic 3 Sucks!!! LOL.... I thought mine was defective..
.

As a owner, it doesn't suck by any means... its just in beta still. Camera, flight time etc... just doesn't support price for a fully baked drone.

 

[Crashdummytest]

I get what you are saying but for me its like this. My M2P would always (in three years of heavy use) acquire multiple satellite locks in around 30 seconds. I never once had to wait to take off because the signal strength was too weak. Now DJI have changed how their system works and the wait is ridiculous. Why advertise a drone with such a great battery life when you need to waste 10% of it or more waiting for a homepoint to be recorded when a drone three years older still has zero problems in this area? Nobody wants to pay this kind of money for a drone that performs much than its 3 year older predecessor. Its that simple

My thoughts exactly, this is ridiculous. I compare directly to my Air2S.

 

[Vanzeppelin]

Same issues with GPS here. It was not a problem before the 12/10 update, but now it takes a significantly longer time to get a lock. With the M2P I can be in the exact same area and have a lock in 30-45 seconds.

It's not a huge issue, but when a 3 year old product outperforms your recent purchase in any metric it's annoying. I often have to take off, reach the imposed 98 ft. limit, and wait for it to finally get a sufficient lock from there. I'm not sure how people that are saying it's normal GPS behavior can justify that when no DJI drone I've owned before has ever taken this long--maybe the Phantom 1 but I don't even remember how it indicated lock, as it was just a GoPro screwed into the bottom.

 

[MrEddieNunez]

I'm finding no difference in GPS acquisition time and I'm still getting the GPS SIGNAL WEAK issue....Curious as to what others think.

I live in NYC and prior to the .500 firmware, I was getting a GPS home point lock under 1 minute. Now it is taking 2-3 minutes.