GPS glitches when using two GPS modules

[hsteinhaus]

A flight log with two active GPS modules shows 23 GPS glitches in 5 minutes of flight, while GPS conditions were OK (HDOP<2.0).

Flight log is available here: https://www.dropbox.com/s/9jp5dg32wzvu9bu/iris-dual-gps-326.BIN?dl=0

More information: https://groups.google.com/forum/?nomobile=true#!topic/drones-discuss/P5MqTgdF8N4

[nhadrian]

Hi,

the same here. I have a Ublox Neo-7N as main GPS and Neo-6N as secondary GPS. HDOP and NSats were OK but I had two Glitches during my test flight. The interresting is that the two modules had glitches exactly at the same time (GPS0 was always a slight earlier).

LOG and Parameters file can be found here: https://drive.google.com/folderview?id=0B7kxYjpc_dXQZHJVSm14bXI0alE&usp=sharing

Another log from yesterday, I was hovering some over my garden, had Glitches but HDOP and NSats were OK. Log here (parameters were the same): https://drive.google.com/folderview?id=0B7kxYjpc_dXQV3ZKczNkdXVSblk&usp=sharing

Checking my previous crash log from my TRI (#1560 closed issue) I noticed that I had similar error - the strange is that I didn't have GPS2 connected - how could I have GPS glitch on GPS2???

Regards Adrian

[nhadrian]

I analized my logs with Kevin's great analizer (#1560) and figured out that in all my analized logs, 5-7 s before GPS_glitch I had APM_Speed_errors.

I made screenshots about these here: https://drive.google.com/folderview?id=0B7kxYjpc_dXQTm5XRjdCU25NZ2c&usp=sharing

Maybe these related to each other?

[nhadrian]

Additionally, I couldn't find any "glitch" in GPS Lat, Lon datas. I don't know if it is important or not.

Here is a sample of the two GPS Lats.

[nhadrian]

Hi,

Any updates on this issue?

[jschall]

When it says "err gps-2" it means the error code is 2, which stands for glitch.

You won't get GPS errors relating to the second GPS (yet).

We don't use the second GPS yet. On Nov 20, 2014 7:54 AM, "nhadrian" notifications@github.com wrote:

Hi,

Any updates on this issue?

— Reply to this email directly or view it on GitHub https://github.com/diydrones/ardupilot/issues/1597#issuecomment-63829434 .

[nhadrian]

Ok, that makes sense, but I thought GPS2 is already used if sat number is more than on primary (GPS_AUTO_SWITCH parameter?) And what type of error does "err gps-0" means? What could cause the glitch protection to be activated? Why I can't see any sign of glitch in coordinates, HDOP, sat number? Also, is it normal that I got so many APM_Speed_errors with Kevin's analyser?

Many thanks for the replies, regards Adrian

[jschall]

0 means resolved. These are defined in defines.h in the ArduCopter directory.

On Thu, Nov 20, 2014 at 12:06 PM, nhadrian notifications@github.com wrote:

Ok, that makes sense, but I thought GPS2 is already used if sat number is more than on primary (GPS_AUTO_SWITCH parameter?) And what type of error does "err gps-0" means? What could cause the glitch protection to be activated? Why I can't see any sign of glitch in coordinates, HDOP, sat number? Also, is it normal that I got so many APM_Speed_errors with Kevin's analyser?

Many thanks for the replies, regards Adrian

— Reply to this email directly or view it on GitHub https://github.com/diydrones/ardupilot/issues/1597#issuecomment-63870493 .

[jschall]

Glitch protection is activated when the GPS suddenly jumps unexpectedly.

On Thu, Nov 20, 2014 at 1:51 PM, Jonathan Challinger < mr.challinger@gmail.com> wrote:

0 means resolved. These are defined in defines.h in the ArduCopter directory.

On Thu, Nov 20, 2014 at 12:06 PM, nhadrian notifications@github.com wrote:

Ok, that makes sense, but I thought GPS2 is already used if sat number is more than on primary (GPS_AUTO_SWITCH parameter?) And what type of error does "err gps-0" means? What could cause the glitch protection to be activated? Why I can't see any sign of glitch in coordinates, HDOP, sat number? Also, is it normal that I got so many APM_Speed_errors with Kevin's analyser?

Many thanks for the replies, regards Adrian

— Reply to this email directly or view it on GitHub https://github.com/diydrones/ardupilot/issues/1597#issuecomment-63870493 .

[nhadrian]

Today I had many test-flights, I had GPS glitches in every takeoff, even between takeoff when copter was on ground. I replaced the main GPS to a Neo-6, the same issue (I thought maybe I have a faulty GPS receiver), so I'm quite sure receivers are OK. (If you would like, I can share some logs as a sample, basically the same as before…)

All flights was made far from buildings, on a big field.

Does anybody else have such issues?

[jschall]

Do you have a dataflash log?

On Sun, Nov 23, 2014 at 5:52 AM, nhadrian notifications@github.com wrote:

Today I had many test-flights, I had GPS glitches in every takeoff, even between takeoff when copter was on ground. I replaced the main GPS to a Neo-6, the same issue (I thought maybe I have a faulty GPS receiver), so I'm quite sure receivers are OK. (If you would like, I can share some logs as a sample, basically the same as before…)

All flights was made far from buildings, on a big field.

Does anybody else have such issues?

— Reply to this email directly or view it on GitHub https://github.com/diydrones/ardupilot/issues/1597#issuecomment-64118521 .

[nhadrian]

Of course, here are three logs from today with GPS glitches: https://drive.google.com/folderview?id=0B7kxYjpc_dXQQnNUUTNRUFBTNEU&usp=sharing (Please note that I disabled GPS in arming to let copter be armed in stabilize without GPS lock.)

Now I'm going to reduce logged items according to Kevin's analyzer suggestion, to reduce "Apm Speed Errors" as called in the analyzer:

[nhadrian]

Today I had a flight of 7 minutes with only one GPS receiver (Neo-7) attached and log bitmask reduced to default. There was not a single Glitch. I'll try to make more takeoffs during this week to check this.

But, all those APM Speed errors still exists. So not related to log bitmask.

[silverburn]

Me too. 2x M8N's - plenty sats, no visible impact on flight. But tons of errors. Did another flight immediately after, as a loiter test...35 minute loiter, and not a single glitch. It only seems to error when moving...?

[rmackay9]

These are support issues and really shouldn't be logged in the github issues list. Instead they should go on the APM Forum. Sorry to be strict but we can't let the issues list become a support center.

[silverburn]

Happy to do that Randy, but I've tried posting on DIYdrones with no response, and in my entire use of ardupilot support, I've never had a response to any query; to the point I don't really post there anymore. (It's actually down at the moment as well). This appears to be the only place where an issue with dual GPS units is known.

[rmackay9]

This is strongly related to this other issue: https://github.com/diydrones/ardupilot/issues/1875

[magicrub]

will be fixed by https://github.com/ArduPilot/ardupilot/pull/5642

[rmackay9]

fixed in master so closing.