Hi all.
Just looking for some input on GPS glitching. I had a number of them in todays flight.
In general it didn’t prove to be a real problem other then once tossing an EKF error that signaled a land event. But the rest where cleared in short order. But I am curious as to what are the possible causes of GPS glitches.
Sorry I can’t attach the log as I still get the file size warning.
Oh and it was very windy and cloudy…other then that it was a wide open field.
(Get ready for an over simplification) GPS Glitches occur when the GPS reports a position that is far from the copters current position. The flight controller knows it is wrong, decides not to listen to the GPS until it reports a sane position, and pops up the message. If it is in a flight mode that requires GPS and the condition doesn’t improve, it will land.
Causes… Really just bad GPS reception. Interference on your vehicle. Trees, mountains, buildings, etc, blocking the signal. Maybe clouds in your case.
1 Like
Hey @wicked1
Ok, that makes sense and just dumbed down enough for me lol.
So I see two possibilities then…Clouds, it was cloudy and the clouds where water laden so first possible choice.
Second I had replaced the RC receiver with a larger and more complex receiver. I suppose it could emit some 433 MHz interference. So today I added some shielding to the top and sides of the receiver and will add a ground to the shielding first chance I get.
Going back to the clouds. I just realized my first 2 flights where uneventful but the cloud had not rolled in. So perhaps it was just the cloud.
Two years ago, I was suffering glitches on a daily basis. LiPo was on top of the frame, behind the GNSS antenna, and whenever the quad started a run heading North - front with the antenna lower and rear with the LiPo higher - GPS was prone to glitch. Heading North sideways or backwards was fine.
A longer mast didn’t help 100%, as neither did a heavily shielded USB3 piece of cable between GPS and FC. The solution was a better GNSS antenna.
Your run-of-the-mill 25x25 mm ceramic antenna was perfect in size and weight back in the days of single-constellation GPS days. But it runs excessively out of bandwidth today with GPS+Glonass, for our NEO-M8Ns. And is a joke for the newer GPS+Glonass+Beidou+Galileo M9N.
Better antenna is a dual-feed dual-frequency ceramic patch, with the filtering, amplifying and mixing stage. You’ll find them in higher-end DJI units. A2 sports an older type, but still above and beyond anything offered in our side of the drone market, and they have original HMC5883 mags onboard. A3 sports the latest model antenna, but have clone QMC883 magnetometers that are sometimes a nightmare. They also involve some soldering, to get rid of DJIs proprietary canbus and connect to GPS serial and mag I2C. The gem is the Inspire2 replacement GPS. It doesn’t have a mag, so you’ll have to add one, and the RM3100 is a prime target, but it requires no soldering, as it’s a standard serial NEO-M8N GPS. And it’s just $60. That’s what just the ceramic patch retails to in electronic component stores 
Thank you for that.
I don’t think its the battery in this case as its below the GPS. I have had little issue with this unit other then this day so I am wondering if it was just the weather.
That said I am not silly enough to not consider other possibilities.
I think your referring to this gps
https://www.amazon.ca/iMusk-Original-Inspire-Module-Repair/dp/B07L5JPRS4
IF so I see that it has more then 4 wires going to it so I am curious what the pinout is.
Also I have never seen an M9n but will have a look for that one.
That’s it. First 4 are the usual +5V TX RX Gnd, the next two are for the onboard security chip, and are read by the Inspire to check the GPS is genuine DJI unit, and the last two go to free pads that don’t have anything soldered on them.
Awesome thanks man
I will keep it in mind if the glitch remains a problem. Batteries are charging now
I have also been struggling with gps glitches for a while. I suspect that the 3D speed reported by the gps is the problem. Looking at my logs the gps position looks more realistic than the estimation from the fc.
Running 4.1dev and bn-880 gps.
Tonight I found this m9n alternative, https://store-drotek.com/917-DP0801.html
Have ordered one to see if there will be any difference.
Anyone having experience with it?
Hi!
Started with arducopter some months ago, but keep having problems with GPS glitches. Now I rebuilt the frame to separate GPS for all electronic, but no help.
From previous evaluation I think it’s not really caused by GPS, but more with the compass variance. Is it common?
My quad is powered with LiIon battery which is magnetic and even I do my best to separate it from compass, it’s impossible to avoid compas influence. Ardupilot some way can automaticaly compensate compass inacuracy so after few initial “toilet bowl” circles in LOITER copter settles down and holds position OK, but I have problems in RTL at bigger distances and altitudes, where copter keeps LOITERing instead of RTL while slowly gaining altitude (see in attached log between times 12:44:29 and 12:45:07)
Current GPS and compass is BN-880, before I tested completely different one and problem was the same.
My log is here: https://drive.google.com/file/d/128ycNJe3MfOS6EOLZJayCY-nmL_JlrUP/view?usp=sharing
Any ideas, advices or even further questions wellcome!
Thanks!
Hi Roman
I am no expert in GPS…there are a number of guys on this forum that are far greater experts then I. But I have one observation I can share… That is the BN-880 GPS…I had one and had to toss it…GPS glitches.
Also @Patrik_Emilsson is also having the same issues on the BN-880 GPS…so it’s something to consider.
1 Like
Hi Patrik, I am testing the m9n:
[Edit]
I raised an issue:
Ricky, thanks for the idea, but previously I had the same experience with different GR-BD GPS ( https://images.app.goo.gl/kZMoDQiM5QAFjQZv8 ) and in the flight log are very strange messages like “EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE” so it seems to be something deeper. Great would be if there is someone with deep understanding of internal software process and looked into my logfile - maybe the answer can be found there…
Hey Roman
I get the same messages when my machine first loads up. It appears that it needs a minute or two to settle before I arm it. If I arm it right away when there are enough sats. I get the error.
IF I wait a minute and then try I don’t get the error and I am free to fly.
Here is a log snippet, basically on this flight I bunged the battery in and immediately tried to arm it. I could see in QGC that there where 16 sats so i figured I was good to go…ah nope.
However all the other flights where I waited for a minute had no such errors.
- ERRORS * TIME Subsys ECode Description
- 19:00:26 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
- 19:00:26 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
- 19:00:29 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
- 19:00:37 24 1 EKF PRIMARY: OCCURED OR FAILED TO INITILIASE
- 19:00:37 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
- 19:00:37 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
- 19:00:42 24 0 EKF PRIMARY: ERROR RESOLVED/CLEARED
- 19:00:42 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
My vehicle booted up at 12:35:14 and I armed and airborne at 12:36:49 with 14 locked satellites, but as you can see below, errors were coming repeatedly much later.
12:38:27 24 1 EKF PRIMARY: OCCURED OR FAILED TO INITILIASE
12:38:31 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
12:38:31 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
12:38:32 11 2 GPS: GPS GLITCH
12:38:41 16 0 EKF CHECK: EKF CHECK VARIANCE CLEARED
12:38:41 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
12:38:49 11 0 GPS: ERROR RESOLVED/CLEARED
12:40:47 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
12:40:47 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
12:40:47 11 2 GPS: GPS GLITCH
12:40:54 16 0 EKF CHECK: EKF CHECK VARIANCE CLEARED
12:40:54 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
12:41:03 11 0 GPS: ERROR RESOLVED/CLEARED
12:42:06 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
12:42:06 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
12:42:06 11 2 GPS: GPS GLITCH
12:42:15 16 0 EKF CHECK: EKF CHECK VARIANCE CLEARED
12:42:15 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
12:42:23 11 0 GPS: ERROR RESOLVED/CLEARED
12:42:50 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
12:42:50 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
12:42:52 16 0 EKF CHECK: EKF CHECK VARIANCE CLEARED
12:42:52 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
12:43:23 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
12:43:23 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
12:43:24 11 2 GPS: GPS GLITCH
12:44:08 16 0 EKF CHECK: EKF CHECK VARIANCE CLEARED
12:44:08 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
12:44:13 16 2 EKF CHECK: EKF CHECK BAD VARIANCE
12:44:13 17 1 EKFINAV FAILSAFE: OCCURED OR FAILED TO INITILIASE
12:44:17 29 1 : OCCURED OR FAILED TO INITILIASE
12:44:22 10 5 FLIGHT MODE: VEHICLE WAS UNABLE TO ENTER FLIGHT MODE LOITER
12:44:22 24 0 EKF PRIMARY: ERROR RESOLVED/CLEARED
12:44:23 16 0 EKF CHECK: EKF CHECK VARIANCE CLEARED
12:44:23 17 0 EKFINAV FAILSAFE: ERROR RESOLVED/CLEARED
12:44:26 11 0 GPS: ERROR RESOLVED/CLEARED
12:44:38 29 0 : ERROR RESOLVED/CLEARED
I think it’s something else than GPS but don’t know what to test next.
Hi, have now got a chance to test this GPS on my quad and I also had issues with the auto configuration. Did it manually in Ublox config tool and turned off auto config in ArduCopter. The result is fantastic I got 30satelites today and hdop of 0.47!
Have not seen any glitches in my log anymore so I’m very happy. I guess one reason for the better reception is the much bigger antenna, approx 4times the size compared to 880GPS
1 Like
Hey Patrick which GPS are you referring to.
Might be time to go shopping.
Hi,
This one from Drotek with Ublox M9N
So far very happy with the performance, though a little troublesome setting it up.
I can help if you go for it and get into trouble.
1 Like
Thanks Patrick. Do yo have teh dimensions of the GPS unit as a whole. She looks large. Space is a bit of a concern on my two drones…that said…GPS glitches are ticking me off.
Hi guys,
just wanted to add, that I discovered my GPS glitches are caused by vibrations and have nothing to do with GPS.
When I fly slowly, vibrations are weak and IMU position corresponds to GPS position. Once I start fly fast (for my 9’’ quad it’s about over 45km/h) vibrations increases and disturbs IMU readings the way that position became inacurate. Then there is inconsistency betwen IMU and GPS position (even it’s correct) and GPS glitches are reported. May be usefull to know for you…
Roman
Good to know Roman.
Unfortunately in my case I get the glitch the minute I arm the bird. I disarm it and it clears, I arm it it glitches. I wait a minute it clears and I can arm. I then may get a glitch while flying…not as often…but from time to time. However I do have issues with Vibrations…so I will look into it. Anything to improve on it’s performance.