RE: Accurate Timings

From: Ted Molczan (
Date: Fri Jun 02 2000 - 16:08:58 PDT

  • Next message: Matson, Robert: "RE: Accurate Timings"

    Patrick wrote:
    > I've been trying to make accurate timings of satellites for a
    > while, and I
    > can't figure out how you guys do it.  The closest I can get
    > is about two or
    > three seconds, because I have to look down at my stopwatch
    > and then back up
    > at the satellite, which takes some time.  So my question is,
    > what's your
    > strategy?
    The strategy is to use a stopwatch referenced to an accurate time source, such
    as WWV. WWV broadcasts on several shortwave frequencies, including: 5 MHz, 10
    MHz, 15 MHz, 20 MHz; or you can hear it by phoning 303-499-7111.
    My $30 Casio wristwatch has a stopwatch function with a memory that can record
    30 split times.
    Shortly before I go out to observe, I start the stopwatch. Then I phone the WWV
    number, and record several splits at one minute intervals.
    I write down the actual UTC time of each split, so that I have a way to relate
    the elapsed time of the stopwatch to UTC.
    If I am reasonably alert, I should be able to consistently hit the stopwatch to
    within a range of a few hundredths of one second.
    I do not take the splits at the start of a minute, because WWV has a several
    second pause with no clock ticks prior to the long tone at the start of the
    minute. Instead, I count off 5 seconds after the minute, and then take a split.
    This should reduce the reaction-time error, but I can anticipate each tick of
    the clock.
    By this process, I have in effect calibrated the start time of the stopwatch.
    When I return from observing, I phone WWV again, and record a few more splits,
    to enable me to check how much the stopwatch has drifted. A cheap watch like
    mine will drift quite a bit over several hours, requiring that I correct my
    observed times.
    For a short observing session, during which negligible drift is expected, I can
    skip the pre-observing calibration against WWV. However, even if I don't plan a
    long session, I may do this calibration as a safety measure, in case I should
    accidentally stop the watch after I have recorded a few positions (it is all
    too easy to do that in the dark). Having done the pre-observation calibration
    means that even if I stop the watch, I can still compute the UTC time of each
    split that I have recorded. Without the calibration, the data would be lost.
    In any event, it is necessary to perform at least one calibration, either
    before or after observing.
    Most positional observers time the passage of the satellite between pairs of
    closely spaced stars. The closer, the better the positional accuracy, and the
    time accuracy. It is easier to accurately judge the moment that a satellite
    crosses the imaginary line between a pair of closely spaced stars, than more
    widely spaced ones.
    I try not to use stars separated by more than 1 deg, and preferably less than
    0.5 deg. As a rough rule of thumb, an experienced and careful observer should
    be able to achieve an accuracy of about 5 percent of the separation between the
    stars, and 0.1 s in time.
    Ted Molczan
    Unsubscribe from SeeSat-L by sending a message with 'unsubscribe'
    in the SUBJECT to

    This archive was generated by hypermail 2b29 : Fri Jun 02 2000 - 16:09:14 PDT