The attached plots show the sampling jitter and drift in a
1 hour 15 minute series from the u.l1.4.75m sonic.
One can see the 9600 baud inter-character spacing in the vertical
spacing between the groups of points:
10bits/byte / 9600 bits/sec = .001 secs/byte.
The jitter is about +-0.005 sec. This is about 1/3rd the sampling
period, which isn't great.
The first plot is from u.l1.4.75m (CSAT SN 855), and the
second is u.u1.5.27m (CSAT SN 856).
The system clock was kept with NTP. Here are the ntp log messages
around that time.
May 3 12:43:29 localhost ntpd: offset -0.001755 sec freq 7.762 ppm error 0.001982 poll 10
May 3 13:43:29 localhost ntpd: offset -0.001508 sec freq 7.755 ppm error 0.002179 poll 10
May 3 14:43:29 localhost ntpd: offset -0.007275 sec freq 7.754 ppm error 0.025945 poll 10
11: data_system, Site marshall, Fri 07-May-2004 15:14:23 MDT, Raised data-system box
Kurt had noticed water in the bottom of the data-system
weather box after the rains at the end of April.
So today at around 3:00pm Tom raised the data-system box
onto the tower. This required disconnecting all the
data cables. We've checked with rserial to channel 200,204
that the cables are connected as before.
He also add some extra cable between the charge monitor and
the data-system - previously it was stretched tight.