[time-nuts] LTE Lite time error
michael.cook at sfr.fr
Wed Jan 28 05:42:35 EST 2015
I have verified the Skytraq claim on my stand alone NS-T.
Tue Jan 27 06:52:06 UTC 2015 $GPGGA,065206.000,4847.3506,N,00216.3005,E,1,06,1.3,191.0,M,47.0,M,,0000*59
Tue Jan 27 06:52:06 UTC 2015 $GPGLL,4847.3506,N,00216.3005,E,065206.000,A,A*52
Tue Jan 27 06:52:06 UTC 2015 $GPGSA,A,3,04,01,22,28,27,11,,,,,,,2.1,1.3,1.7*3F
Tue Jan 27 06:52:06 UTC 2015 $GPGSV,2,1,07,04,83,016,40,11,77,287,36,01,57,288,47,28,29,299,42*7E
Tue Jan 27 06:52:06 UTC 2015 $GPGSV,2,2,07,03,28,223,09,22,25,056,41,27,21,149,26*42
Tue Jan 27 06:52:06 UTC 2015 $GPRMC,065206.000,A,4847.3506,N,00216.3005,E,000.0,000.0,270115,,,A*65
Tue Jan 27 06:52:06 UTC 2015 $GPVTG,000.0,T,,M,000.0,N,000.0,K,A*0D
Tue Jan 27 06:52:06 UTC 2015 $GPZDA,065206.000,27,01,2015,00,00*53
Using the GNSS gui, Query Timing, get a NAK. Monitoring 1PPS gets blank.
When booting from Flash, the quantization error of +/-6ns was verifiable both with Monitor Timing on the GUI and on my scope triggered from my T-Bolt..
When booting from ROM it doesn’t seem to be any worse than +/-10ns .
I would think that the lower quantization error would enable better phase locking in a GPSDO, especially with a long time constant, which is why I bought the device in the first place. So anyone with an LTE-Lite may just prefer to not use the NMEA stream until a new firmware becomes available, rather than use the WA. It would be of interest to others to see what booting from ROM makes in practice with the LTE-Lite.
One observation that I will make is that TTFF is MUCH faster when booting from ROM, and the device is MUCH more sensitive. I can pull in twice the number of SATS when booting from ROM. I am getting HDOP figure <1…wow! The reason why my receiver is not glued to a PRS10 at the moment is that I kept losing 3D Position fix according to the GUI when booted from Flash, even though when in Static mode the 1PPS looked good.
"Ceux qui sont prêts à abandonner une liberté essentielle pour obtenir une petite et provisoire sécurité, ne méritent ni liberté ni sécurité."
> Le 27 janv. 2015 à 02:29, Bill Beam <wbeam at gci.net> a écrit :
> Be careful with 'eyeball data'. GPS receiver does not generate NMEA time data and the leading edge of PPS at the same time.
> Programs like Tac32 (totally accurate clock) and Lady Heather increment the time display at the leading edge of PPS with a
> value 1 second greater than the previous NMEA data time.
> I am able to run multiple GPS receivers into multiple computers running Tac32. The LTE-lite displays one second earlier than
> all the others.
> Prior to the announcement of "Leap Second Pending" in the GPS data stream the LTE-lite agreed with all other units. Now it does not.
> On Mon, 26 Jan 2015 22:59:36 +0100, Mike Cook wrote:
>> Yes there is certainly an error here:
>> With my timing module I was just eyeballing the output on a windows platform , comparing GUI data.
>> I have just linked the module up to a BeagleBone Black syncG�d with NTP and this is the NMEA msg log:
>> root at bb3:/home/mike/serial-ports# while read GGA; do echo `date` $GGA; done < /dev/ttyO4
>> Mon Jan 26 21:51:07 UTC 2015 $GPGGA,215106.000,4847.3526,N,00216.3005,E,1,04,2.8,192.8,M,47.0,M,,0000*5C
>> Mon Jan 26 21:51:07 UTC 2015 $GPGLL,4847.3526,N,00216.3005,E,215106.000,A,A*56
>> Mon Jan 26 21:51:07 UTC 2015 $GPGSA,A,3,25,12,06,31,,,,,,,,,3.0,2.8,1.0*3A
>> Mon Jan 26 21:51:07 UTC 2015 $GPRMC,215106.000,A,4847.3526,N,00216.3005,E,000.0,173.5,260115,,,A*60
>> Mon Jan 26 21:51:07 UTC 2015 $GPVTG,173.5,T,,M,000.0,N,000.0,K,A*0D
>> Mon Jan 26 21:51:07 UTC 2015 $GPZDA,215106.000,26,01,2015,00,00*54
>> Mon Jan 26 21:51:07 UTC 2015 $PSTI,00,2,0,4.6,,*30
>> The time data is all a second late so they appear to have a serious issue.
>>> Le 26 janv. 2015 +� 02:47, Bill Beam <wbeam at gci.net> a +�crit :
>>> Odd indeed. My LTE-lite is one second late, appears to have already added the pending leap second.
>>> I can compare with four other GPS timeing receivers using time pulse on DCD line. The NMEA data
>>> reports in error.
> Bill Beam
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts