[time-nuts] 60 Hz data
hmurray at megapathdsl.net
Thu Jun 30 09:36:03 UTC 2011
Tom's suggestion of using a modem control signal is a winner.
Here is a graph:
Hour 0 is the start of Jun 30 UTC, 5PM PDT.
It's gained almost 200 cycles over 9 hours. There are enough wobbles in
there that I think the data is good. It will be interesting to see what the
rest of a 24 hour slot looks like.
I'm in California. Does anybody in the same power grid have data that looks
similar? (or different?)
Here is the same data plotted as frequency measured over 10 seconds.
There is a lot of noise. I don't know how much is on the line and how much
is jitter in the capture system. (I'll have to think about that tomorrow.)
I'm sampling every 10 seconds. If it picked up an extra cycle, the frequency
would be 60.1 Hz. There is one data point (about 4 hours) that's in that
I'm using a wall wart transformer. It's nominally 12V RMS but I measured 16V
no load. I used a couple of 1K resistors as a divider.
I'm using the Linux PPS stuff.
[murray at jim 60Hz]$ cat /sys/class/pps/pps0/assert
[murray at jim 60Hz]$
and a python hack to write stuff to a log file.
Here is a log file if anybody wants to play:
First two columns are MJD and seconds this day copied from NTP's log file
3rd column is unix time stamp of most recent event. 4th column is number of
events since ???
5th and 6th columns are delta time/count since the previous line (10 seconds
55742 4.259 1309392004.255156 232038 10.014635 601
55742 14.270 1309392014.267448 232639 10.012292 601
55742 24.281 1309392024.279595 233240 10.012146 601
55742 34.291 1309392034.275325 233840 9.995730 600
55742 44.298 1309392044.287679 234441 10.012354 601
55742 54.309 1309392054.299189 235042 10.011509 601
The first graph uses columns 2 and 3.
The frequency graph uses columns 5 and 6.
These are my opinions, not necessarily my employer's. I hate spam.
More information about the time-nuts