[LEAPSECS] BBC article

Peter Vince pvince at theiet.org
Sun Nov 6 06:33:49 EST 2011

On 6 November 2011 08:51, Paul J. Ste. Marie <paul_j at ste-marie.org> wrote:


> WWVB includes a UTC-UT1 offset that will overflow if too many leap seconds

> are skipped.  I forget exactly how many bits it occupies, but the format has

> a hard limit of 60 bits.

NIST have a nice diagram at <http://tf.nist.gov/stations/wwvbtimecode.htm>

WWVB normally uses 200ms second markers to indicate a "1" data bit,
but for the DUT1 data they are 500ms. In Europe our MSF (Britain) and
DCF-77 (Germany) just use 200ms markers, with the period between 200
and 300ms after the second edge being a second level data digit.

MSF: <http://www.npl.co.uk/science-+-technology/time-and-frequency/time/msf-time-date-code>
DCF: <http://www.ptb.de/cms/index.php?id=1791&L=1>

MSF uses these second-level bits for the DUT-1 data at the start of
the minute, and towards the end of the minute for parity and "Daylight
Saving Time" bits. Perhaps these bits throughout the rest of the
minute could be employed to transmit a larger range DUT1 code? Five
BCD digits (20 bits) would allow coding up to an hour in tenths of a
second steps. Going for seven digits (28 bits) would allow finer
resolution down to millisecond level.


More information about the LEAPSECS mailing list