[time-nuts] no 23:59:60 in TAC32
John Ackermann N8UR
jra at febo.com
Sun Jan 1 08:55:19 EST 2006
Bill Beam said the following on 12/31/2005 11:45 PM:
> I have since compared UT+ raw data with TAC32 display
> as follows:
> UT+ raw data TAC32 display
> 12/31/05 23:59:59 same
> 12/31/05 23:59:60 1/1/06 00:00:00
> 12/31/05 23:59:60 1/1/06 00:00:01
> 1/1/2006 00:00:00 1/1/06 00:00:01
> 1/1/2006 00:00:01 same
Isn't there an extra extra second there? The UT+ is showing two :60s;
shouldn't there be only one?
Another thing I haven't dug into, but if the GPS is outputting the time
at the *next* PPS signal, the sentence will appear to be ahead compared
to a display that's attempting to show the *current* time. Not sure if
that's the case with TAC32.
[ snip ]
> I obtained my data using a Nikon D70 digital still camera
> operated in 'burst' mode taking about three frames per second
> of the computer screen. Both the TAC32 display and UT+
> raw data are shown together on the screen.
I had my D70 in burst mode pointed at the Spectracom 8170 and HP5065A
clocks; unfortunately, the 8170 took 4 minutes to update, so I didn't
capture the transition. I did get a couple of shots after the change
showing the 1 second disparity between the two, but it's just not the
same as capturing 23:59:60 (which, alas, the Spectracom never did
display -- it sent :59:59 twice).
More information about the time-nuts