DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Slow GPS lock - is it Firmware or Satellite Constellations

I also noticed that the specs on every other model I checked so far (4) say:

  • Satellite Systems​

  • GPS+GLONASS

Where Mavic 3 says:

  • GNSS​

  • GPS + Galileo + BeiDou
One difference is lack of Glonass, but I would not expect that to be root cause here. Just a data point.
That's basically all I was speculating on for thread discussion... the lack of GLONASS and the Addition of BeiDou.
 
Yes, the RC Pro ALSO has GPS and does use GLONASS. With a standard controller, the phone GPS is used. That is the controller location in the radar in the lower left.
To gain more understanding...
Great points, I'm used to a standard controller that use Tablet GPS.

On this discussion, I hadn't looked at the RC Pro as Android device & GPS; My attention was on Craft. Appreciate your note!

Being that said, the RC Controller has GPS for it's location. I understand it's not assisting the Craft to GPS lock... but provides it's location as HomePoint to craft for mission references and a coordinate for RTH if disconnected.

Are there additional considerations?
Does the RCPro use BeiDou too? (I haven't looked).
 
Sorry for phone autocorrect... that's not the issue.
If you notice in 1st post it's spelled correctly
Sorry but it wasn't spelled correctly at first. I came along and caught it before a moderator edited the title and your first post. Go back and check at the end of the first post and you'll see it was edited by a moderator.

edit: Nevermind, I see you've edited post #1 since then.
 
Sorry but it wasn't spelled correctly at first. I came along and caught it before a moderator edited the title and your first post. Go back and check at the end of the first post and you'll see it was edited by a moderator.
I don't see a moderator "edit" note. I'm on a phone screen & browser.
I went back and corrected 1 miss-spelling within the context... not title.
If I misspelled the title, I don't recall.
If a Moderator corrected for me... thanks, but is there a note that indicates edit.
 
but is there a note that indicates edit.
There was. At the bottom where it now says 'last edited at such-and-such time' it said 'last edited by a moderator'. Then when you came along and edited it again it changed again to reflect your edit time.

Don't worry about it please, just get back to your theory.

I'm on a phone too...aggravating ain't it?
 
There was. At the bottom where it now says last edited at such-and-such time it said last edited by a moderator. Then when you came along and edited it again it changed again to reflect your edit time.

Don't worry about it please, just get back to your theory.

I'm on a phone too...aggravating ain't it?
Lol... all good! Sm screen, Old eyes, in thought writting, auto correct... equates to misspellings at times. Normally try to reread to correct but often pass over misspelling. I'm not without errors for sure! Humorously I noticed you followed suit.. so you're probably correct.

Just tring to stimulate discussion and alternative thought paths.
Interesting how common any deviation is quickly opposed when it's normally the 1st way to Dx an issue.

I hadn't considered it... but after writing I pondered if the anti-DJI communications would crop up. No direct association, other than BeiDou system using a 2way communication vs beacon.

It is interesting that the 1st craft that losses GLONASS and applies BeiDou has a GPS lock performance problem. Somewhere in the mix, I'd look at that change as a possibility too.
 
Lol... all good! Sm screen, Old eyes, in thought writting, auto correct... equates to misspellings at times. Normally try to reread to correct but often pass over misspelling. I'm not without errors for sure! Humorously I noticed you followed suit.. so you're probably correct.

Just tring to stimulate discussion and alternative thought paths.
Interesting how common any deviation is quickly opposed when it's normally the 1st way to Dx an issue.

I hadn't considered it... but after writing I pondered if the anti-DJI communications would crop up. No direct association, other than BeiDou system using a 2way communication vs beacon.

It is interesting that the 1st craft that losses GLONASS and applies BeiDou has a GPS lock performance problem. Somewhere in the mix, I'd look at that change as a possibility too.
BeiDou can be operated in 2-way mode, but that's not how simple GNSS receivers use it - that's receive only.

Another possibility is that DJI doesn't have a license for its equipment to receive BeiDou signals in the US, and has turned off BeiDou reception in the ublox M8030 chip until it gets one. That would leave just GPS and Galileo, which would reduce the number of available satellites and slow down the solution. Or maybe they tried to switch it back to GPS + Galileo + GLONASS and messed up the firmware.
 
  • Like
Reactions: db4476
BeiDou can be operated in 2-way mode, but that's not how simple GNSS receivers use it - that's receive only.

Another possibility is that DJI doesn't have a license for its equipment to receive BeiDou signals in the US, and has turned off BeiDou reception in the ublox M8030 chip until it gets one. That would leave just GPS and Galileo, which would reduce the number of available satellites and slow down the solution. Or maybe they tried to switch it back to GPS + Galileo + GLONASS and messed up the firmware.
All Great Points! I wasn't attempting to indicate DJI was utilizing 2way... just a difference in systems.

Your suggestions have great merit to consider.
 
Regarding the SLOW GPS lock.
Not a expert on GPS, just read into the issues a bit to help understand why such a long delay to lock GPS.

From my rather limited understanding it's not a Mavic 3 FW related correction, but more of a satellite constellation group decision. In past DJI models (and other brands) have included the GLONASS and Galileo constellations as primary, along with the GPS GNSS constellation for a long time. The GLONASS system was removed from the Mavic 3 Firmware and substituted with the BeiDou constellation.

The GLONASS is a Russian Satellite system, Galileo is EU, and the BeiDuo is a Chinese satellite system... along with the GPS USA system. In the long scheme, the BeiDuo is suppose be much more accurate for aircraft under altitude 2,000Km and able to see 50% more satellites compared to other GPS systems.

The Craft's Firmware programming and the embedded circuits determine what satellite systems to examine & accept. Currently the long delay is the BeiDui Satellite system taking a much longer time communicating.

This may be associated to the BeiDou Constellation design... GPS, GLONASS, Galileo are basically "beacons" pickup by GPS GNSS Devices. The BeiDuo is a 2-way communication system, providing the ability to also transmit back to the Satilites: text msg, and various low level communications.

On the Mavic 3, my assumption is the quick 4-6 Satilites are likely GPS system, the extended time to obtain the remaining 6 satellites are likely BeiDuo system.

If that's accurate assumption, then a more efficient BeiDuo communication protocol would be needed to be programmed or the need to utilize GLONASS again to speed up the GPS lock. Until this is addressed, Mavic 3 will take time to lock onto 12+ Satilites.

A few interesting articles:



"not a Mavic 3 FW related correction, but more of a satellite constellation group decision."

Constellation choice is not an issue. The GNSS chipset used by DJI sets the default constellations to GPS+GLONASS, to change that would require software.

" my assumption is the quick 4-6 Satilites are likely GPS system, the extended time to obtain the remaining 6 satellites are likely BeiDuo system."

Possibly, but we do not know which constellations are in use, do we?

" a more efficient BeiDuo communication protocol would be needed to be programmed or the need to utilize GLONASS again to speed up the GPS lock. Until this is addressed, Mavic 3 will take time to lock onto 12+ Satilites"

The communication you refer to is very basic, think twitter @ 140 characters and is only available to China's military not the public. The communication capability has nothing to do with GNSS positioning at all. As for adding GLONASS back to speed up positional lock, the GNSS chip already has the capability to use GLONASS since it is part of the default settings. We are back to software again.

To sum up, it is not a constellation problem, since dropping GLONASS and adding BeiDou is done by software at the chip level. The issue started after a firmware update in December, which leads me to the conclusion that it is a software issue.

Cheers!
 
BeiDou can be operated in 2-way mode, but that's not how simple GNSS receivers use it - that's receive only.

Another possibility is that DJI doesn't have a license for its equipment to receive BeiDou signals in the US, and has turned off BeiDou reception in the ublox M8030 chip until it gets one. That would leave just GPS and Galileo, which would reduce the number of available satellites and slow down the solution. Or maybe they tried to switch it back to GPS + Galileo + GLONASS and messed up the firmware.
That is a very real possibility.

Cheers!
 
"not a Mavic 3 FW related correction, but more of a satellite constellation group decision."

Constellation choice is not an issue. The GNSS chipset used by DJI sets the default constellations to GPS+GLONASS, to change that would require software.

" my assumption is the quick 4-6 Satilites are likely GPS system, the extended time to obtain the remaining 6 satellites are likely BeiDuo system."

Possibly, but we do not know which constellations are in use, do we?

" a more efficient BeiDuo communication protocol would be needed to be programmed or the need to utilize GLONASS again to speed up the GPS lock. Until this is addressed, Mavic 3 will take time to lock onto 12+ Satilites"

The communication you refer to is very basic, think twitter @ 140 characters and is only available to China's military not the public. The communication capability has nothing to do with GNSS positioning at all. As for adding GLONASS back to speed up positional lock, the GNSS chip already has the capability to use GLONASS since it is part of the default settings. We are back to software again.

To sum up, it is not a constellation problem, since dropping GLONASS and adding BeiDou is done by software at the chip level. The issue started after a firmware update in December, which leads me to the conclusion that it is a software issue.

Cheers!
All good info... although in my opinion it started out of the box new... 2 M3 Cine units, both slow GPS out of box. After 2 FW updates, I haven't noticed any difference one way or other... continues extreme slowness. Have appreciated the FW updates for addition features & enhancements, but no change in GPS from day one across 4 Midwest States.

You're correct on the 2way communication limitations, that was sited multiple times in the links provided in initial post. That wasn't the concern... other than indicating BeiDou is based a different system design.

Playing on the 2way communication... not concerned but just playing on capabilites possible. (Don't interpret as suggestion or concern). As a Vet exposed to MI, Elec SigWarefare and experienced in machine language.... a 140 "characters" is a lot space if methodology & construct is planned out for communication. I could send a ton of info in a few Morse lines or Elec Teletype transmission if established protocols on both ends.

Great discussion and input... it'll be interesting to discover what the cause & correction ends up producing.
 
All good info... although in my opinion it started out of the box new... 2 M3 Cine units, both slow GPS out of box. After 2 FW updates, I haven't noticed any difference one way or other... continues extreme slowness. Have appreciated the FW updates for addition features & enhancements, but no change in GPS from day one across 4 Midwest States.

You're correct on the 2way communication limitations, that was sited multiple times in the links provided in initial post. That wasn't the concern... other than indicating BeiDou is based a different system design.

Playing on the 2way communication... not concerned but just playing on capabilites possible. (Don't interpret as suggestion or concern). As a Vet exposed to MI, Elec SigWarefare and experienced in machine language.... a 140 "characters" is a lot space if methodology & construct is planned out for communication. I could send a ton of info in a few Morse lines or Elec Teletype transmission if established protocols on both ends.

Great discussion and input... it'll be interesting to discover what the cause & correction ends up producing.
If you never had reasonable GPS out of the box, it is possible that both aircraft that you received might have already been upgraded at the factory to one of the updates, depending upon when you received them. It may also be that all the Cine units were slow to begin with. My M3 non-Cine unit was initially fine, until I installed the first update. It got even slower with the second update. Looking forward to to the next update, which promises a fix.
 
  • Like
Reactions: Dougcjohn
If you never had reasonable GPS out of the box, it is possible that both aircraft that you received might have already been upgraded at the factory to one of the updates, depending upon when you received them. It may also be that all the Cine units were slow to begin with. My M3 non-Cine unit was initially fine, until I installed the first update. It got even slower with the second update. Looking forward to to the next update, which promises a fix.
I purchased 1st, extremely slow.. that was early in the M3 and hadn't been too much chatter on GPS. It also behaved oddly.. connected & disconnected with RC Pro multiple times on first 3 flights in attempting a hover in parking lot. Then appeared to roll & buck on a level flight. No FW updates yet. I returned it and ordered a 2nd Cine. When it arrived, it performed identical on GPS but more normal on the other issues.

A few weeks later, the Dec FW was released, followed by Jan release. Performed both via DJI Asst on both RC Pro & M3.

Improvements & new features but similar slow GPS.

I too am looking forward to GPS correction and SDK.
 
I purchased 1st, extremely slow.. that was early in the M3 and hadn't been too much chatter on GPS. It also behaved oddly.. connected & disconnected with RC Pro multiple times on first 3 flights in attempting a hover in parking lot. Then appeared to roll & buck on a level flight. No FW updates yet. I returned it and ordered a 2nd Cine. When it arrived, it performed identical on GPS but more normal on the other issues.

A few weeks later, the Dec FW was released, followed by Jan release. Performed both via DJI Asst on both RC Pro & M3.

Improvements & new features but similar slow GPS.

I too am looking forward to GPS correction and SDK.
My initial flights with the M3 non-Cine version on the original FW were with the RC-N1. After getting the RC-Pro, I had to update everything. Maybe those with good satellite reception on the initial FW were limited to those with the RC-N1. Hmmm...
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
130,592
Messages
1,554,181
Members
159,596
Latest member
da4o98