Leap second - MSF time signal
https://www.youtube.com/watch?v=FpBxB2Yqh-U
Time signal -
0000000000111111111122222222223333333333444444444455555555556
0123456789012345678901234567890123456789012345678901234567890
Starting at 23:58:00 -
A 100000000000000000001011010010110001110100011101100101111110
B 100000000111100000000000000000000000000000000000000000001000
Starting at 23:59:00 -
A 1000000000000000000001011100001000001000000000000000001111110
B 1111111000000000000000000000000000000000000000000000000111100
So the new DUT1 was transmitted starting at 23:59:01, but the year 2017,
which is normally transmitted starting at 17s, was transmitted starting at
18s.
So it appears that the leap second was actually inserted at 23:59:17?
On Sun, Jan 01, 2017 at 02:28:23AM +0000, Deirdre O'Byrne wrote:
Nice.
Starting at 23:58:00 -
A 100000000000000000001011010010110001110100011101100101111110
B 100000000111100000000000000000000000000000000000000000001000
Starting at 23:59:00 -
A 1000000000000000000001011100001000001000000000000000001111110
B 1111111000000000000000000000000000000000000000000000000111100
My decoder seemed to give something like this:
A: 00000000000000000001011010010110001110100011101100101111110
B: 00000000111100000000000000000000000000000000000000000001000
Raw: 31/12/2016 23:59 Sat -400 No-DST
Ctime: Sat Dec 31 23:59:01 2016
A: 00000000000000000001011100001000001000000000000000101111110
B: 11111100000000000000000000000000000000000000000000000111000
Raw: 1/1/2017 0:01 Sun 600 No-DST
Ctime: Sun Jan 1 00:01:01 2017
Which shows it switching at the right time. There's a long wave spectrum
at:
http://www.maths.tcd.ie/~dwmalone/time/leap2016/leap2017010100.png
or an animated verion at:
https://www.youtube.com/watch?v=9Y_DQZN88HQ
I think the time codes relate transmitted relates to the minute following
the one in which they are transmitted according to:
http://www.npl.co.uk/upload/pdf/MSF_Time_Date_Code.pdf
Could that explain what you saw?
David.
Your decoded timecode only has 60 data points for the 61-seconds of the
last minute of 2016 UTC, so it's impossible to say what your decoder did
with the leap second. (Also I think you have an error in your parity bit
for the time for midnight (bit 57B)).
The long wave spectrum seems to show the same fourteen 0.1+0.9-second
pulses between 23:59:07 and 23:59:20 (inclusive), which tends to suggest
that the leap second was indeed inserted between the DUT1 code and the Year
code.
Interestingly the long wave spectrum also shows DCF77 - a project for when
I get bored! :)
On 1 January 2017 at 12:02, David Malone dwmalone@maths.tcd.ie wrote:
On Sun, Jan 01, 2017 at 02:28:23AM +0000, Deirdre O'Byrne wrote:
Nice.
Starting at 23:58:00 -
A 100000000000000000001011010010110001110100011101100101111110
B 100000000111100000000000000000000000000000000000000000001000
Starting at 23:59:00 -
A 1000000000000000000001011100001000001000000000000000001111110
B 1111111000000000000000000000000000000000000000000000000111100
My decoder seemed to give something like this:
A: 00000000000000000001011010010110001110100011101100101111110
B: 00000000111100000000000000000000000000000000000000000001000
Raw: 31/12/2016 23:59 Sat -400 No-DST
Ctime: Sat Dec 31 23:59:01 2016
A: 00000000000000000001011100001000001000000000000000101111110
B: 11111100000000000000000000000000000000000000000000000111000
Raw: 1/1/2017 0:01 Sun 600 No-DST
Ctime: Sun Jan 1 00:01:01 2017
Which shows it switching at the right time. There's a long wave spectrum
at:
http://www.maths.tcd.ie/~dwmalone/time/leap2016/leap2017010100.png
or an animated verion at:
https://www.youtube.com/watch?v=9Y_DQZN88HQ
I think the time codes relate transmitted relates to the minute following
the one in which they are transmitted according to:
http://www.npl.co.uk/upload/pdf/MSF_Time_Date_Code.pdf
Could that explain what you saw?
David.
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/
mailman/listinfo/time-nuts
and follow the instructions there.
Hi Deirdre,
Well done getting such a clear recording! Yes, the leap-second is
effectively inserted after the DUT1 code on MSF as you say, even though
that isn't crystal clear from the MSF spec document that David linked to.
The DUT1 codes refer to the "current" minute, whilst all the rest of the
time and date information refers to the next minute epoch!
Peter
On 1 January 2017 at 15:01, Deirdre O'Byrne deirdre.dub@gmail.com wrote:
Your decoded timecode only has 60 data points for the 61-seconds of the
last minute of 2016 UTC, so it's impossible to say what your decoder did
with the leap second. (Also I think you have an error in your parity bit
for the time for midnight (bit 57B)).
The long wave spectrum seems to show the same fourteen 0.1+0.9-second
pulses between 23:59:07 and 23:59:20 (inclusive), which tends to suggest
that the leap second was indeed inserted between the DUT1 code and the Year
code.
Interestingly the long wave spectrum also shows DCF77 - a project for when
I get bored! :)
Didn't the DUT1 code change from -0.4s during the minute starting 23:58:00
to +0.6s during the minute starting 23:59:00?
On 1 January 2017 at 16:06, Peter Vince petervince1952@gmail.com wrote:
Hi Deirdre,
Well done getting such a clear recording! Yes, the leap-second is
effectively inserted after the DUT1 code on MSF as you say, even though
that isn't crystal clear from the MSF spec document that David linked to.
The DUT1 codes refer to the "current" minute, whilst all the rest of the
time and date information refers to the next minute epoch!
Peter
On 1 January 2017 at 15:01, Deirdre O'Byrne deirdre.dub@gmail.com wrote:
Your decoded timecode only has 60 data points for the 61-seconds of the
last minute of 2016 UTC, so it's impossible to say what your decoder did
with the leap second. (Also I think you have an error in your parity bit
for the time for midnight (bit 57B)).
The long wave spectrum seems to show the same fourteen 0.1+0.9-second
pulses between 23:59:07 and 23:59:20 (inclusive), which tends to suggest
that the leap second was indeed inserted between the DUT1 code and the
Year
code.
Interestingly the long wave spectrum also shows DCF77 - a project for
when
I get bored! :)
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/
mailman/listinfo/time-nuts
and follow the instructions there.
Hmmm, yes, you're absolutely right. Now I'm confused!
Peter
On 1 January 2017 at 16:21, Deirdre O'Byrne deirdre.dub@gmail.com wrote:
Didn't the DUT1 code change from -0.4s during the minute starting 23:58:00
to +0.6s during the minute starting 23:59:00?
On 1 January 2017 at 16:06, Peter Vince petervince1952@gmail.com wrote:
Hi Deirdre,
Well done getting such a clear recording! Yes, the leap-second is
effectively inserted after the DUT1 code on MSF as you say, even though
that isn't crystal clear from the MSF spec document that David linked to.
The DUT1 codes refer to the "current" minute, whilst all the rest of the
time and date information refers to the next minute epoch!
Peter