[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gpsd-users] Communication loss with GR-701W

From: Gary E. Miller
Subject: Re: [gpsd-users] Communication loss with GR-701W
Date: Mon, 20 May 2019 11:39:32 -0700

Yo Benoît!

On Mon, 20 May 2019 18:35:35 +0200
Benoît Thébaudeau <address@hidden> wrote:

> > 3.18 had a bug, which is why 3.18.1 cam out shortly therafter.  
> I've upgraded to 3.18.1, but it does not help.

No, you just have not realized what it fixed yet.  :-)

> > When you hotplug the GPS gets reset to stored configuration.  You
> > should check how you have your GPS configured.  It is easy to 
> > program the u-blox to ouput no data on startup, that can cause
> > issues.  
> Actually, there is data at the very beginning. See the log below.

I guess I'll get there in a bit.

> > The exact error would be nice.  Better yet the full log.  
> Here is the full log:
> call gpsd:PROG: switching to match packet type 1:
> $GPGSV,2,2,08,20,40,119,37,21,52,059,37,26,58,164,39,27,53,290,20*7C\x0d\x0a

So you have NMEA.

> gpsd:IO: <= GPS:
> $GPGSV,2,2,08,20,40,119,37,21,52,059,37,26,58,164,39,27,53,290,20*7C
> gpsd:WARN: GPGSV field 3 value of 8 != actual count 4

Started mid-cycle.  Things will sync up.

> gpsd:INFO: Sats used (4): gpsd:IO: <= GPS:

And they did.

> gpsd:PROG: selecting u-blox driver...

Changed to -u-blox.

> gpsd:IO: UBX checksum 0x1b85 over length 108, expecting 0x3030

Bad checksums are not good.  Is this serial, RS-232 or USB?

> gpsd:IO: <= GPS: $GPTXT,01,01,00,trk tick 2 96000 dt 23881*3A
> gpsd:WARN: TXT: Error: trk tick 2 96000 dt 23881
> gpsd:IO: <= GPS: $GPTXT,01,01,00,trk tick 3 144000 dt 8694*36
> gpsd:WARN: TXT: Error: trk tick 3 144000 dt 8694
> gpsd:IO: <= GPS: $GPTXT,01,01,00,trk tick 5 240000 dt 5595*38
> gpsd:WARN: TXT: Error: trk tick 5 240000 dt 5595
> gpsd:IO: UBX: len 10

More error messages from the GPS.  Also not good.  The GPS is telling
us it is sick.

> As you can see, after the trk tick error, the only thing still
> working is the PPS.

Yeah, not good.  Of course that error is undocumented by u-blox.

> > Have you tried ubxtool to reset the configuration?  
> Is it safe?

Yes.  The default config is in ROM, by design it is the safest thing
to do.

> I mean, I do not know the configuration the GR-701W ships
> with,

Yup, but likely it is the factory defaults.

> nor if gpsd may change this configuration on its own,

It certainly did, as shown by your logs.

> so I do
> not want to risk getting a malfunction because of this.

Too late, you already have a malfunction.

> > Sending different types of reset is easy with ubxtool.  
> Sending a MON-HW2 poll (echo -ne '\xb5\x62\x0a\x0b\x00\x00\x15\x49'
> > /dev/ttyUSB0) results in: 00000000  b5 62 0a 0b 1c 00 fa 61  0f 5f
> > 70 00 00 00 ff ff  |.b.....a._p.....|
> 00000010  ff ff ff ff ff ff ff ff  ff ff 00 00 00 00 00 00
> |................| 00000020  00 00 5e
> 0f                                       |..^.|

That way is to hard, I'm not even going to try to hand decode that.

And no point, UBX-MON-HW2 has nothing to do with the current issue.

> So the module replies properly, which means that the baudrate is
> correct on both sides.

We already knew that.

> But it does not talk more after having sent
> this command.

Which one again sounds like a serial port issue.

> I still have to dump the whole internal state of the module to see if
> it reports anything abnormal.

Don't bother.  Just try to configure it properly, without saving or

    # ubxtool -d NMEA
    # ubxtool -e BINARY

> Sending a CFG-RST/hotstart/HWreset (echo -ne
> '\xb5\x62\x06\x04\x04\x00\x00\x00\x00\x00\x0e\x64' > /dev/ttyUSB0)
> makes the module recover,

Wow, you like doing things the hard way.  How about:

    # ubxtool -p COLD

> but the issue can then still happen after
> replugging the module a few times.

Yes, because that reverts to the saved config.

> I cannot succeed in duplicating the issue when stracing gpsd, which
> may indicate that it is timing-related.

Which points again to potential serial problems.  Also potential
buffer overruns sending to the GPS at low speeds.

> > Also, since this happens on hotplug, how is your udevd setup?  
> I'm just using the SysV gpsd.hotplug and 25-gpsd.rules installed by
> gpsd 3.18.1.

Good.  And since it does seem to detect the GPS, b4 failing, that is
not an issue.

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can’t measure it, you can’t improve it." - Lord Kelvin

Attachment: pgpjMdasYApdd.pgp
Description: OpenPGP digital signature

reply via email to

[Prev in Thread] Current Thread [Next in Thread]