Campbell Scientific LoggerNet Datalogger Support Software - Manual (Page 112)

Section 4. Setting up Datalogger Networks
4.2.6 Setting the Clock
A datalogger’s Clock tab can be used to define a schedule at which an
automatic clock check will be performed. The datalogger’s clock will be set if
it varies from the LoggerNet server’s clock more than the amount of time
specified in the Allowed Clock Deviation field.
Because it is important to maintain accurate time stamping of your data, there
are a few things to take into consideration when setting up a clock check
schedule.
Your datalogger clock should deviate no more than ±1 minute per month.
Typically, this drift is less than what will be experienced with a personal
computer. Therefore, unless you have a scheduled task on your computer
which synchronizes your computer’s clock with an atomic clock or other
accurate time keeping device, the datalogger’s clock may be more accurate
than the PC’s clock.
Another point to consider is how the clock checks may affect the time stamp
for your data. Let’s say, for instance, that you have a data collection schedule
of one minute with a clock set if the two clocks deviate more than two minutes.
Over time, the clocks may drift sufficiently that the datalogger’s clock is set. If
the datalogger’s clock is 12:02:00, and the LoggerNet computer clock is
12:04:15 the datalogger’s clock will be set to 12:04:15. Therefore, there will
be no data for the time stamps 12:03 and 12:04. Conversely, if the datalogger’s
clock is a few minutes faster than the LoggerNet computer’s clock, the result
would be duplicate time stamps that contained different data.
For the CR10X-TD family of dataloggers, when an instruction
P84 Output Record is executed in the datalogger program to create
an interval based table and a record is stored, the time stamp is not
stored along with the record. Instead, when data is retrieved from
the datalogger, the datalogger uses the time stamp of the last
record stored and the table interval to calculate the time stamp for
any previous records. This calculated time stamp is then stored by
the server as part of the data record along with the other data
values when data is collected. Because of this time stamping
method, if the datalogger clock is changed such that it passes an
output interval a discontinuity could occur in the records that
could cause the time stamps to be incorrect. Event-based data
storage does not rely on a calculated time stamp. Data stored to a
table based on an event includes a time stamp in the table.
In table-based dataloggers the record number can be used along
with the time stamp to assure that records are in order, and no data
has been missed.
Changing the computer system clock while the display screens are
running will terminate the connection for most of the screens.
This can also affect LoggerNet operations or even crash the
program.
NOTES
4-56