The "GPS glitch"

Joined
May 4, 2015
Messages
2,134
Reaction score
820
Age
62
As everyone knows 3DR is battling a GPS issue that they attempted to band-aid with the last firmware.

In my case it just slowed down the time it takes to get to Fly GPS mode.

This week I had a tiny issue which I would not have ruled out as pilot error. Basically, upon landing it was dancing around more than normal and when I tried to set it down it caught a leg and flipped over. It basically had landed but when it flipped over it spun the props on concrete which destroyed them before I could AB - Pause.

I looked at log file and everything looked fine except the GPS location on the final second or two showed Solo about 50' from actual.

So just for the heck of it and to test the tech support submit log function I submitted the log.

I got an initiation repression asking for complete details, proof of purchase, was anything damaged, etc.

The second response was this.

Hi Rich,


I went through your logs.
Apparently, you experienced a GPS glitch.
The GPS horizontal position didn't match the real position of the drone for an instant.
This made the Solo move to that incorrect position.
The new update release, that you already have (1.0.5), handles the GPS glitch better by switching to Fly: manual mode when there's inconsistent GPS data.
That mode is not GPS-dependent, so it there's a GPS error, the Solo's position will not be affected.
You probably will not experience anything like that now that you've updated your system.



Let me know if you have any questions, Rich.


Best regards,

Diego
[Technical support engineer]


I was a little confused by his response, so I replied that this was my replacement (#2) Solo and that it actually was shipped back to me with the new firmware. It was running the new firmware when this happened.

His response back was.


HI Rich,


Thanks for the info.
I'll share your log with my colleagues and have a second look at it.
Were the propellers the only damage to the Solo?



Best regards,

Diego
[Technical support engineer]


So either the glitch problem continues with the new firmware or tech support is not paying attention and really evaluating logs.

I did find it odd the logs showed the aircraft moved 50' in the final 2 seconds when it really did not.
 
This has happened to me a few times when it dances around trying to land. Once, i was actually able to AB PAUSE luckily.

For the times it did this, it actually knocked the entire gopro frame piece from its clipped in position and the gopro was just hanging by the cable. Again luckily it was near the ground.

Fortunately, no tip overs. I know i just said that and probably jinxed myself.

Maybe i'll just manually land from now on to avoid the dancing problem and the gps glitch? I like the auto land feature but I'm thinking to avoid any problems until they fix it, correct by manually landing just prior to touch down during auto land.
 
I got the impression my response was either by somebody with very limited knowledge reading off a script or simply was just skimmed over and they pasted a general response. My actual issues really weren't addressed just said fly in a better spot and land on flat ground, duh.
 
This has happened to me a few times when it dances around trying to land. Once, i was actually able to AB PAUSE luckily.

For the times it did this, it actually knocked the entire gopro frame piece from its clipped in position and the gopro was just hanging by the cable. Again luckily it was near the ground.

Fortunately, no tip overs. I know i just said that and probably jinxed myself.

Maybe i'll just manually land from now on to avoid the dancing problem and the gps glitch? I like the auto land feature but I'm thinking to avoid any problems until they fix it, correct by manually landing just prior to touch down during auto land.
Yep that's what I'm doing as well it's not really convenient when just shooting vids but it's my safest option, all my issues have been about 10 feet or less off the ground. I've had flawless flights since I gave up on using gps to take off and land. Till a for sure fix I'll just keep this up, it's not a big deal but would be nice if I had the option to auto land and takeoff in fly mode since it's what's recommended.
 
Yep, lol, guess we gotta get back to our "roots" of taking off and landing manually. Haha. I think sometimes, these companies [not specifically quadcopters, tech in general, like those cars that drive themselves, altho it wud be nice to sleep in traffic] make it too easy for people and add auto this and auto that.

And then i dont mind manual takeoffs and landings cuz im having fun piloting this thing from start to finish...
 
I got the impression my response was either by somebody with very limited knowledge reading off a script or simply was just skimmed over and they pasted a general response. My actual issues really weren't addressed just said fly in a better spot and land on flat ground, duh.

I am not ruling that other :)

And I normally manual land all the time. Pretty sure if I had been auto-landing in this situation it would have took off and been much worse.
 
Having read many, if not all, of the gps 'glitch' posts here, I get the feeling it would most likely take an actual hardware modification to sort out our gps issues? I'm no expert when it comes to the dev side of firmware code changes and implementations, but at the end of some of these conversations conclusions are drawn in regard to improper equipment layout etc. as I quoted from another thread.
"I never stated the GPS hardware was defective. In fact I stated the opposite. It is 3DR's implemention of the GPS components in solo that is defective."
from here:
http://www.3drpilots.com/threads/so...-3mm-active-patch-antenna-test.544/#post-8244

Are we currently reaching the limit of improvement a firmware update can provide, and perhaps we might have to do a bit of hardware design change? (i.e. relocating the gps to mount externally etc.) I'm confused.
 
Yep, lol, guess we gotta get back to our "roots" of taking off and landing manually. Haha. I think sometimes, these companies [not specifically quadcopters, tech in general, like those cars that drive themselves, altho it wud be nice to sleep in traffic] make it too easy for people and add auto this and auto that.

And then i dont mind manual takeoffs and landings cuz im having fun piloting this thing from start to finish...
Have been flying for awhile (RC and Full Size) I prefer to do the flying, with the exception of Smart Shots. I just find it more enjoyable to be the pilot, but that's a personal feeling and I know everyone doesn't agree. To go out and push a few buttons to watch it fly just isn't as much fun. Plus the less electronics you rely on the greater the odds of successful flights.
 
so should we keep flying the solo with all theese reported gps failures or ground them until 3dr comes up with a fix for this?
 
so should we keep flying the solo with all theese reported gps failures or ground them until 3dr comes up with a fix for this?
Dirby, only you can answer that. Have you been flying and having a good time with it? Do you rely solely on GPS or do you fly other modes? If you want to take off and land manually and use GPS as needed, you should be fine. But that is just my opinion and I haven't seen a glitch yet. (yes, I did say yet..)
 
all my flights have been glitchless but reading all of the horor stories kinda scares me.
 
all my flights have been glitchless but reading all of the horor stories kinda scares me.
Understood. But try to keep it in perspective.. How many users you have seen with a problem and 1000s of Solos in the air. Most like you and I that have not had a glitch. It's kind of like deciding not to drive any more because you saw where someone had a wreck today. Is it possible you will have a wreck next time you are in a car? Absolutely! So do you stay home or enjoy life?
But hey, that's just me.
 
  • Like
Reactions: GoodnNuff
Understood. But try to keep it in perspective.. How many users you have seen with a problem and 1000s of Solos in the air. Most like you and I that have not had a glitch. It's kind of like deciding not to drive any more because you saw where someone had a wreck today. Is it possible you will have a wreck next time you are in a car? Absolutely! So do you stay home or enjoy life?
But hey, that's just me.
agree.
no truer words were ever spoken.
and there is a solo at my local bb if i should happen to bend this one i can go get that one.
 
  • Like
Reactions: Raybro
As everyone knows 3DR is battling a GPS issue that they attempted to band-aid with the last firmware.

In my case it just slowed down the time it takes to get to Fly GPS mode.

This week I had a tiny issue which I would not have ruled out as pilot error. Basically, upon landing it was dancing around more than normal and when I tried to set it down it caught a leg and flipped over. It basically had landed but when it flipped over it spun the props on concrete which destroyed them before I could AB - Pause.

I looked at log file and everything looked fine except the GPS location on the final second or two showed Solo about 50' from actual.

So just for the heck of it and to test the tech support submit log function I submitted the log.

I got an initiation repression asking for complete details, proof of purchase, was anything damaged, etc.

The second response was this.

Hi Rich,


I went through your logs.
Apparently, you experienced a GPS glitch.
The GPS horizontal position didn't match the real position of the drone for an instant.
This made the Solo move to that incorrect position.
The new update release, that you already have (1.0.5), handles the GPS glitch better by switching to Fly: manual mode when there's inconsistent GPS data.
That mode is not GPS-dependent, so it there's a GPS error, the Solo's position will not be affected.
You probably will not experience anything like that now that you've updated your system.


Let me know if you have any questions, Rich.


Best regards,

Diego
[Technical support engineer]


I was a little confused by his response, so I replied that this was my replacement (#2) Solo and that it actually was shipped back to me with the new firmware. It was running the new firmware when this happened.

His response back was.


HI Rich,


Thanks for the info.
I'll share your log with my colleagues and have a second look at it.
Were the propellers the only damage to the Solo?


Best regards,

Diego
[Technical support engineer]


So either the glitch problem continues with the new firmware or tech support is not paying attention and really evaluating logs.

I did find it odd the logs showed the aircraft moved 50' in the final 2 seconds when it really did not.
Rich,
I found a similar cursory overview in my situation, which I had to show was eronious. I also had similar GPS "innacuracies" in my log file, and had a "horiz variance" error. Mine however was in the air, and took a tumble.

I am guessing (though educated) at this point, but I will bet that the problem is with the actual implementation of the GPS unit and placement of the antenna, along WITH momentary simultaneous problems with another sensor (such as baraometer or accelerometer) in combination with the code that flips between them. The code has no way of accounting for the errors that are coming up (not sure it even could) and we are getting unexpected results when that happens in flight and quick decisions are necessary..

I don't know enough about the details yet, and I am just "thinking outloud"here but an example would be when the GPS gives wildly wrong information, that the Solo looks at position from the accelerometer and if that is wrong, it tumbles, or if the barometer thinks it is to high it drops, if the compass thinks it has the wrong orientation it turns, or combinations thereof.

You can try to fix that in code, but the reality is the code isn't the real problem.

As I said, I am no expert here, but I see this ALL the time in other control systems, where they try to outsmart the physical problem with controls. Sometimes they can, and sometimes they can't, but when they can't the controls aren't the real problem. I am hoping they find a way to isolate the GPS (ala the P2) and it fixes it...
 
  • Like
Reactions: rrmccabe
The GPS glitch is more than just the software. It's linked to a radio issue with the controller. On July 4th my drone flew two flights with no problem. The third flight the drone took off on autopilot then wouldn't hold altitude properly to hover. It began flying erratically from side to side and manual controls just made it go crazier. At some point I started worrying about this thing flying away and injuring someone so I attempted to have it return to home. The controller was completely locked out and although it audibly made the return to home sound I couldn't control it anymore. The drone was just a few feet off the ground and was unsafe to attempt to grab it. Finally it tried to land without ever raising up to 48' first and the landing gear touched down and then flipped itself over. While upside down the motors finally kicked in and hit full speed shredding the rotor blades up. It was trying to reach automatic landing altitude and when I grabbed it with my hand and turned it over to try and shut it down the drone was pulling my arm into the air it was pretty strong even with busted rotor blades.
The 3DR Supoort Team came back after 12 days of reporting issues and admitted the controller had bad RF levels and caused a GPS Glitch.
The ironic part is they kept receiving my logs and sending me calibration instructions to run with my computer through the Linux commands. They knew it was having issues and their oversight almost caused an accident. They need automatic scripts to read through logs and quickly alert them to issues that may cause bodily harm or property damage. We all can't wait for 12 days for someone to actually analyze the log files once they hear a bad crash has happened. They should be preventing these crashes in advance by having us ground our air craft or replacing it quickly if the scripts alert them to a serious issue.
Most Best Buy stores get rotor blades every week. I have 4 new sets and they discounted them due to the Solo issues causing damage.
 
Last edited:
Can you post 3DRs actual response? I am not understanding. Being controller related with surprise me.
 
The first response was the canned "GPS Glitch" the admission of the controller RF issue was on a telephone call to me. I don't think they wanted to put that in writing. I will check my other emails and see if they ever said it after mentioning the GPS Glitch. They immediately replaced the Drone and controller.
 
For what it's worth, GPS, I had my Son help with a little video.

As some of you already know I have never been an [out of the box] person.
I have and probably always will build my own stuff , I'm old school, , that way I can make changes as needed as most people do.
I'm sure someone will let me have it on this video telling me the do's and don't but please save your breath if it works the proof is in the video..

I have been testing a 8 over the stock 3DR 6 GPS receiver and I personally am having very good results and in this video you will see the test and this test is one that the Solo failed miserably.. So when the Solo can pass this test I will once again purchase but I think it will take a year or so. .

Link to video GPS test..
To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
I wish I could say I am incredibly impressed but I do that almost every day to some degree or another. I just got back from a Tower test flight and "walked my SOLO down the street toward me about 6 inches off the street. I fly out of my back yard (with WIFI all over the place), fly around and land between 2 trees on a 18"x18" square every time. I am not saying your demo isn't to be noted, but the implication the SOLO is way outclassed is just wrong. Don't want to get anything started just that I believe I could put my SOLO behind yours and do the same thing. Jus' sayin'
 

Members online

Forum statistics

Threads
13,096
Messages
147,752
Members
16,067
Latest member
Minh44