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

[Suren]

Who cares? if they can’t fix it then I’d just as well believe him as I would Mad Angler. And the only people who know the real issue (DJI) obviously aren’t telling what’s up.

If Dji were ever to become transparent about this issue it will say a lot for their credibility but as we know they will quietly put out FW Updates with "bug fixes" or "Optimised flying experience"

 

[ZoneAerial]

If Dji were ever to become transparent about this issue it will say a lot for their credibility but as we know they will quietly put out FW Updates with "bug fixes" or "Optimised flying experience"

They have some of the worst changelogs ever.

 

[Skywatcher2001]

Suren, I applaud you sir and thank the aerial GODS for our moderators. I just spent an hour in that forum and ZoneAerial is right. It is a zoo, inside of a circus next to a train wreck (and that's being polite so our mods don't thrash me)

You are somewhat of a DJI 'point man' in both forums and they obviously trust you. Reading thru that dumpster fire it seems this may (and I stress 'may') be a hardware issue . . . and that is discouraging knowing the history of DJI and owning a significant hardware issue.

So . . . now what? Another waiting game?

 

[Suren]

Suren, I applaud you sir and thank the aerial GODS for our moderators. I just spent an hour in that forum and ZoneAerial is right. It is a zoo, inside of a circus next to a train wreck (and that's being polite so our mods don't thrash me)

You are somewhat of a DJI 'point man' in both forums and they obviously trust you. Reading thru that dumpster fire it seems this may (and I stress 'may') be a hardware issue . . . and that is discouraging knowing the history of DJI and owning a significant hardware issue.

So . . . now what? Another waiting game?

Looks to be another waiting game for now until they can get to the bottom of the issue or just leave it as that and keep pushing out small updates in the hope that we either forget about it or it gets fixed. I have posted before that for me I have no gps issues now and I was so certain I had a faulty drone but the FW I am on proves that theory wrong- but then why does this update work for some and not everyone is a real mystery. I honestly would recommend that if people can't live with it then start a return process for a drone swap out, this will put Dji under pressure and will be forced to get to the bottom of this issue, their so called Mavic 3S plan, if true will have to wait until this is resolved if there starts to be many returns.

 

[ZoneAerial]

Suren, I applaud you sir and thank the aerial GODS for our moderators. I just spent an hour in that forum and ZoneAerial is right. It is a zoo, inside of a circus next to a train wreck (and that's being polite so our mods don't thrash me)

You are somewhat of a DJI 'point man' in both forums and they obviously trust you. Reading thru that dumpster fire it seems this may (and I stress 'may') be a hardware issue . . . and that is discouraging knowing the history of DJI and owning a significant hardware issue.

So . . . now what? Another waiting game?

The forum makes for great laughs. It’s like a turducken. Except stuffed with (Mod Removed Language)Kind of like the Mavic 3 is at this moment, now that I think of it.

 

[mcde]

The thing is, GNSS is standard fare, it was established years ago, it's not exactly experimental stuff and DJI after all these years have tons of experience on this. It's not an issue anybody expects to face when buying a new device.

It's also known that downloading the ephemeris directly from the GNSS satellites is excruciatingly slow, so TTFF in the order of seconds requires some kind of assistance, usually a network connection and an A-GPS server in the user's network. Without A-GPS, TTFF may be very long. I remember back in the day, when the first gps-enabled smartphones surfaced (that's about twenty years ago, these devices did not have an always-on connection to the internet, and A-GPS was not implemented) that it took ages to get a GPS lock and with only a few satellites, maybe 4-6.

It's also very difficult to believe that GNSS chips responsible for receiving and handling the data are problematic as they, too, are widely used parts.

So, all things considered, it's very disturbing that this issue has persisted for so long and it's also really disappointing. For example, I sold my Mavic 2 pro back in December because I planned to buy a Mavic 3. I'm still waiting though. The Mavic 2 pro was an excellent, very reliable drone and I do not want to buy anything less than that.

 

[MrEddieNunez]

I did the .600 firmware update on my Mavic 3, 8 minutes for a GPS lock in NYC. The issue is not resolved.

 

[ZoneAerial]

I did the .600 firmware update on my Mavic 3, 8 minutes for a GPS lock in NYC. The issue is not resolved.

Eddie: Many of us are in the same boat. I saw a huge thread on the forums about this, and DJI deleted it - probably because it exploded and people went nuts on each other. Anyhow, I think the only way DJI will deal with this (if they can) is if every person who has a problem reports it to DJI and if they ask, give them logs. Again. That's what I did. I hope you do the same. That's the only thing we can do. Pretty much no difference for me either - in open areas in central Virginia.

 

[beermat]

EDIT: No doubt, @beermat made some very interesting evaluations on the GNSS board. I really appreciate him doing that. But I think he would also agree, we have not discovered a smoking gun with respect to the slow boot times on the Mavic 3 and why they're so different and got worse in Dec.

Correct. I resisted the urge to speculate, even then, tempting as it was In my testing (via u-center and my own code) of the M3 GPS module I received, even with the unpopulated chip, it functioned exactly the same as the Mavic Air chip did, and functioned as expected and documented, through power cycles, warm starts, cold starts, even with A-GNSS via uBlox AssistNow.

There is one simple thing I always throw into the mix as I agonize on theories on what causes this, and it's important in the context of the latest "theory" passed off as fact, and that is - hot swapping the battery after the home point has been established. The M3 can only be using persistent storage (flash, battery-backed RAM, etc) for ephemeris data at this point, and it works fine in this case in "instantly" establishing a home point, even for those experiencing long initial acquisition times. Removing the main battery power will clear all data stored in volatile RAM and leave only persistent storage available. Whether power is removed for 30 seconds or 12 hours, that is still true so how can it have space to retain that data when powered up 30 seconds later but not 12 hours later? That doesn't make sense.

Acquire a home point, fly a Hyperlapse mission and use any of the other new features introduced in the Dec firmware, and then remove the battery for 45 seconds, re-insert, and it will lock the home point in seconds. How does that fit with the newly proposed theory?

Also, with the M3 uBlox GPS module (which I identified in that thread a while ago as being a different variant of the typical chip used by DJI), I downloaded almanacs and ephemeris data for *every* GNSS system and *every* satellite, regardless of current visibility, via the AssistNow service, stored it in the battery backed RAM, and power-cycled it, and it retained it. It has enough memory on-chip. I don't recall the uBlox M8 specifications giving any way for an end-user to use the BBR storage independently as a generic data store, either.

The other thing is, in this age of reliable high-speed internet, people tend to have the viewpoint that a long download time is indicative of size of the download, and thus the ephemeris and almanac data must require a lot of storage, and that's why this latest "theory" makes sense, squeezing it into memory is an issue. In fact, it's just that satellites have incredibly low bandwidth that makes it slow - the full set of almanac and ephemeris data for all GNSS systems and all satellites is around 10Kb. That's less than the "Google" image that downloads when you go to the deliberately-minimal ww.google.com page. It's a tiny amount of data.

So, as you ultimately said - who knows? The latest theory is hard to take as it doesn't make technical sense, knowing how the uBlox chip functions and can be used, and also the behavior of the OP in subsequent follow-ups makes it hard to believe. He claimed to only be passing on information, but he wasn't - he was presenting a theory backed with information that was secretly only available to him (red flag as that is unverifiable, classic conspiracy tactic) and information that he claimed was on other forum posts, which turned out to be false or non-existent/misunderstood.

 

[DelBeachFlyer]

@beermat so what would you do if you had an M3 still within the return window? While I love to reverse engineer with the best of them, the net result is the same right now. I think it comes down to "do we have confidence that DJI will be able and will resolve the issue?" Their track history isn't exactly stellar.

 

[beermat]

@beermat so what would you do if you had an M3 still within the return window? While I love to reverse engineer with the best of them, the net result is the same right now. I think it comes down to "do we have confidence that DJI will be able and will resolve the issue?" Their track history isn't exactly stellar.

Honestly, it's tough to decide what to do The hard part of this is that it is impossible to get a good feel of the scope, scale and type of the issue. I know some people have this problem; I know some people don't. What I don't know is, out of all the people who own a Mavic 3, what percentage actually have the problem? All I see are forum posts, which represent a minority of overall users, and skewed to a certain type of posting, and many of the users who post about the issue are artists, photographers, videographers, etc., not data analysts, scientists, GNSS experts, not people who know how to conduct a repeatable test that isolates the variables, etc. (nor should they be, modern expensive tools should just "work") so a lot of the information about the issue is noise, incomplete testing or just plain wrong!

The question is - can you live with 4-8 mins startup time in every circumstance? For some users, with planned flight times or lazy Saturday afternoon flying days, they can plan around with it and live with it. For pros on the go, for busy people chasing twilight shots or other shots which require spontaneity, it's a deal breaker. I mostly shoot twilight now, so this issue ruins most of my use cases. If I was someone inspecting infrastructure with the drone, it's livable. 6 minutes isn't very long when arriving at a job site and setting up; 6 mins waiting while sunrise is happening around you is eternity

 

[sar104]

...
There is one simple thing I always throw into the mix as I agonize on theories on what causes this, and it's important in the context of the latest "theory" passed off as fact, and that is - hot swapping the battery after the home point has been established. The M3 can only be using persistent storage (flash, battery-backed RAM, etc) for ephemeris data at this point, and it works fine in this case in "instantly" establishing a home point, even for those experiencing long initial acquisition times. Removing the main battery power will clear all data stored in volatile RAM and leave only persistent storage available. Whether power is removed for 30 seconds or 12 hours, that is still true so how can it have space to retain that data when powered up 30 seconds later but not 12 hours later? That doesn't make sense.

Ephemeris data are refreshed every 4 hours.

 

[Raffaello_IZ0QWM]

Hello folks,
sorry, I only arrived now and I haven't read all the threads that talk about the problem.
In Italy I noticed that despite updating the new firmware, someone solved it and someone didn't.
So I wonder and I don't know if anyone has asked, isn't there EMI problems inside the Mavic 3 that for some reason drown out the GPS RF signal?

As someone who builds drones knows, imperfect EMI protection can cause problems with GPS (as well as compass) reception.

 

[beermat]

Ephemeris data are refreshed every 4 hours.

Yes. I know that. That wasn't the point. I'm talking in the context of the recent theory of "not having enough space" to store all the ephemeris; how does that theory hold water in the context of a hot restart, such as swapping a battery? If lack of storage is the claimed issue causing the long startup time when you turn on the drone 12 hours later, it should also apply when you remove power for 30 seconds, but it doesn't.

 

[sar104]

Yes. I know that. That wasn't the point. I'm talking in the context of the recent theory of "not having enough space" to store all the ephemeris; how does that theory hold water in the context of a hot restart, such as swapping a battery? If lack of storage is the claimed issue causing the long startup time when you turn on the drone 12 hours later, it should also apply when you remove power for 30 seconds, but it doesn't.

With respect to that then yes - agreed.

 

[Mad_Angler1]

Who cares? if they can’t fix it then I’d just as well believe him as I would Mad Angler. And the only people who know the real issue (DJI) obviously aren’t telling what’s up. That forum is an unmitigated disaster.

It’s actually really important that hearsay and incorrect information is not allowed to become “fact”

The reason is in future when other DJI drones or other manufactures use this chipset people will look at it and say ohh it’s the bad one beciase that guy said it is or it’s badly designed.

Hearsay and rumours quickly become fact online. The OP post is very cleverly written in such a way that it’s hard to go back with “facts” as such. It’s a classic example of “prove me wrong, ohh you can’t.” Well we can actually.

I have spent half the day talking with one of the main hardware designer of one of the flight controller manufactures. While I can’t share docs as it’s all under NDA with Ublox they are more than happy with what DJI appear to have done with the module. Antenna is good, ground plane is good, battery in place and no issues with any of the chipsets chosen. Nothing alarming initially. I have ordered a Mavic 3 GPS module my self tonight for £100 and that will be here this week and I will also do some digging my self and sharing findings.

My hardware guy did comment on the use of BeiDou though and it was not exactly good.

To the the fantastic user who hooked his upto Ucenter can you share any info you will save me an hour on connection pin out ect.

Edit:

Dang just found you thread, ok your image confirms no external flash. For peoples info that QFN unpopulated in not flash. It’s not the right pinout. Too many infact, also the M8030 has 2Mb internal storage that’s more than enough for most GPS usage as as been stated.

I personally am not expecting any issues with the module it’s self and tbh I’m leaning on interfearance. We are seeing similar issues with HDZero wiping out the GPS too. While everything is speculation the module overall looks fine. What is really possible it’s it’s a packaging issue in the drone. A wire in the wrong place or clock signal from one of the CPUs. Silly mistakes do happen.

I hope everyone is well.

 

[ZoneAerial]

It’s actually really important that hearsay and incorrect information is not allowed to become “fact”

The reason is in future when other DJI drones or other manufactures use this chipset people will look at it and say ohh it’s the bad one beciase that guy said it is or it’s badly designed.

Hearsay and rumours quickly become fact online. The OP post is very cleverly written in such a way that it’s hard to go back with “facts” as such. It’s a classic example of “prove me wrong, ohh you can’t.” Well we can actually.

I have spent half the day talking with one of the main hardware designer of one of the flight controller manufactures. While I can’t share docs as it’s all under NDA with Ublox they are more than happy with what DJI appear to have done with the module. Antenna is good, ground plane is good, battery in place and no issues with any of the chipsets chosen. Nothing alarming initially. I have ordered a Mavic 3 GPS module my self tonight for £100 and that will be here this week and I will also do some digging my self and sharing findings.

My hardware guy did comment on the use of BeiDou though and it was not exactly good.

To the the fantastic user who hooked his upto Ucenter can you share any info you will save me an hour on connection pin out ect.

Edit:

Dang just found you thread, ok your image confirms no external flash. For peoples info that QFN unpopulated in not flash. It’s not the right pinout. Too many infact, also the M8030 has 2Mb internal storage that’s more than enough for most GPS usage as as been stated.

I personally am not expecting any issues with the module it’s self and tbh I’m leaning on interfearance. We are seeing similar issues with HDZero wiping out the GPS too. While everything is speculation the module overall looks fine. What is really possible it’s it’s a packaging issue in the drone. A wire in the wrong place or clock signal from one of the CPUs. Silly mistakes do happen.

I hope everyone is well.

Let DJI correct it if it’s wrong. You say it’s so important to “set the record straight”. But DJI doesn’t appear in a hurry to do that. And you don’t even work for DJI.

 

[Mad_Angler1]

Let DJI correct it if it’s wrong. You say it’s so important to “set the record straight”. But DJI doesn’t appear in a hurry to do that. And you don’t even work for DJI.

Bad information is bad information, I think most can see what's being attempted here. Make enough noise about something that forces them to react, trust me I have done that more than anyone however doing that with incorrect information wont get the issue fixed. No amount of shouting, screaming or bad press will make them do something that they don't know what the cause is or don't want to discuss yet, what will is Data.

Saying they have a bad GPS module with incorrect information will just be completely ignored.

I helped get the FPV Goggles issues resolve and the I2 issues sorted by backing them into a corner with the facts not just making things up. I have delt with DJI for a very long time, as has Ian. I even helped run their Facebook groups for over 2 years as non staff admin so I know my way around them a little. It also needs to be added that COVID will also be slowing things with lockdowns in china again being an issue. Does this help customers no but its still reality.

There also has to be some context here on how large the issue is, how many are affected and is a number of the reports not quite the same issue. I have worked in electromechanical engineering for a long time and reality is often far more complex than its just not working properly for everyone. The internet amplifies everything , it could be 1% of users, it could be 100% we don't know, however we cant assume its a massive wide spread issue and everyone is facing the same issue just because of forum posts. Everyone is first to shout when something is wrong, but often stay quiet when its not.


I know some people are not happy now but the simplest way to get a result asap on that in the short term is vote with you wallet. That is what will hurt them more when its RMA spike and then they have a large issue with used stock on hand. If 500 Mavic's are RMA due to GPS issues its going to get noticed faster than forum posts and articles online. the Press side of things just gets to PR and they will spin it what ever way they want.

DJI is a large siloed company now and getting to the right people can be a challenge. but not impossible.

 

[ZoneAerial]

Bad information is bad information, I think most can see what's being attempted here. Make enough noise about something that forces them to react, trust me I have done that more than anyone however doing that with incorrect information wont get the issue fixed. No amount of shouting, screaming or bad press will make them do something that they don't know what the cause is or don't want to discuss yet, what will is Data.

Saying they have a bad GPD module with incorrect information will just be completely based and ignored.

I helped get the FPV Goggles issues resolve and the I2 issues sorted by backing them into a corner with the facts not just making things up. I have delt with DJI for a very long time, as has Ian. I even helped run their Facebook groups for over 2 years as non staff admin so I know my way around them a little. It also needs to be added that COVID will also be slowing things with lockdowns in china again being an issue. Does this help customers no but its still reality.

There also has to be some context here on how large the issue is, how many are affected and is a number of the reports not quite the same issue. I have worked in electromechanical engineering for a long time and reality is often far more complex than its just not working properly for everyone. The internet amplifies everything , it could be 1% of users, it could be 100% we don't know, however we cant assume its a massive wide spread issue and everyone is facing the same issue just because of forum posts. Everyone is first to shout when something is wrong, but often stay quiet when its not.


I know some people are not happy now but the simplest way to get a result asap on that in the short term is vote with you wallet. That is what will hurt them more when its RMA spike and then they have a large issue with used stock on hand. If 500 Mavic's are RMA due to GPS issues its going to get noticed faster than forum posts and articles on line. the Press side of things just gets to PR and they will spin it what ever way they want.

DJI is a large siloed company now and getting to the right people can be a challenge. but not impossible.

“Bad information is bad information”

You should be spending 24/7 on Twitter and Facebook as a self-anointed arbiter of truth with a singular mission of filtering, and correcting information. I’m sure DJI is well equipped to make its own decisions on how to react or not react without the expert advice from a former volunteer Facebook group monitor for them which you presume as a badge of your expertise in their PR strategy (or lack thereof).

You have expressed your opinion/s about what may or may not happen (albeit in long-winded manners with self-bolstered interpretations.) Thank you.

 

[ianwood]

Mad is exactly right. Which isn't surprising as he is uniquely qualified to shed light on the issue.

If you're shocked and surprised about how opaque DJI is on the issue, maybe you've been lucky and haven't had to deal with something like this before. Welcome. Others have been dealing with DJI on issues like this for the better part of a decade. General rule is to not expect much.

Finally, I'd like to remind everyone that this thread is not for complaints. There are plenty of those threads already. This thread is for constructive analysis of the issue that may help us to actually identify a REAL cause and/or put pressure on DJI.