[gpsd-dev] PPS glitch - Venus chip

[Hal Murray]

I have a SparkFun SkyTraq Venus GPS breakout board connected via USB.
  https://www.sparkfun.com/products/11058

The PPS is sometimes off by a second, but only sometimes.  Sometimes it is 
right.

>From gpspipe, grep for PPS.

2016-01-30 02:08:34: {"class":"PPS","device":"/dev/gps0","real_sec":1454148513
, "real_nsec":0,"clock_sec":1454148514,"clock_nsec":784014,"precision":-10}
2016-01-30 02:08:34: {"class":"PPS","device":"/dev/gps0","real_sec":1454148514
, "real_nsec":0,"clock_sec":1454148514,"clock_nsec":4781714,"precision":-10}
2016-01-30 02:08:36: {"class":"PPS","device":"/dev/gps0","real_sec":1454148515
, "real_nsec":0,"clock_sec":1454148516,"clock_nsec":884658,"precision":-10}
2016-01-30 02:08:36: {"class":"PPS","device":"/dev/gps0","real_sec":1454148516
, "real_nsec":0,"clock_sec":1454148516,"clock_nsec":4886409,"precision":-10}
2016-01-30 02:08:38: {"class":"PPS","device":"/dev/gps0","real_sec":1454148517
, "real_nsec":0,"clock_sec":1454148518,"clock_nsec":961487,"precision":-10}
2016-01-30 02:08:39: {"class":"PPS","device":"/dev/gps0","real_sec":1454148518
, "real_nsec":0,"clock_sec":1454148519,"clock_nsec":1010133,"precision":-10}

Note that some of the real_sec and clock_sec match and some are off by one 
and that the first pair and next pair arrived at the same time.

I was lucky to catch the above.  Normally it sticks at one or the other mode.

Has anybody encountered this before?

What's the right way to debug this?

Here is one second from:
  gpspipe -u -r -P

2016-01-30 03:57:58.002735: 
{"class":"PPS","device":"/dev/gps0","real_sec":1454155077, 
"real_nsec":0,"clock_sec":1454155078,"clock_nsec":258942,"precision":-10}
2016-01-30 03:57:58.064841: 
{"class":"TOFF","device":"/dev/gps0","real_sec":1454155077, 
"real_nsec":816999912,"clock_sec":1454155078,"clock_nsec":64363922}
2016-01-30 03:57:58.065373: 
$GPGGA,115757.817,3726.0916,N,12212.2611,W,1,07,1.1,37.3,M,-32.2,M,,0000*56
2016-01-30 03:57:58.124456: $GPGSA,A,3,12,29,05,25,21,26,20,,,,,,2.0,1.1,1.6*38
2016-01-30 03:57:58.205447: 
$GPGSV,3,1,12,29,72,358,40,20,68,158,23,25,46,195,18,05,43,056,33*7D
2016-01-30 03:57:58.282438: 
$GPGSV,3,2,12,21,37,266,17,26,19,317,18,18,16,205,,12,15,166,17*71
2016-01-30 03:57:58.360888: 
$GPGSV,3,3,12,02,14,064,07,13,09,110,16,15,09,147,,31,08,271,11*75
2016-01-30 03:57:58.443758: 
{"class":"TOFF","device":"/dev/gps0","real_sec":1454155077, 
"real_nsec":816999912,"clock_sec":1454155078,"clock_nsec":443360536}
2016-01-30 03:57:58.444315: 
$GPRMC,115757.817,A,3726.0916,N,12212.2611,W,000.9,180.5,300116,,,A*7A
2016-01-30 03:57:58.489447: $GPVTG,180.5,T,,M,000.9,N,001.6,K,A*0F

Note that there are two TOFF slots with quite different offsets.


-- 
These are my opinions.  I hate spam.