Guest Editorial -- How Accurate Must We Be?

Guest Editorial

How Accurate Must We Be?
by Richard Factor

The ARRL, to which many SETI League members belong, is and has been since the beginning of ham (amateur) radio the umbrella organization for its practitioners in the US. ARRL initially stood for American Radio Relay League; through some marketing sorcery, it now apparently stands for National Association for Amateur Radio. A very large percentage of us join, although we are under no obligation to do so, and its magazine, the engagingly titled "QST" is an important historical and technical journal in our radio lives. The ARRL sponsors various activities and promulgates the rules and dates in QST. For example, "Field Day" is an annual event at which time hams will leave their comfortable homes to erect antennas in public parks, farm fields, and mountain tops and attempt to contact others so-situated over the radio. One major contributor to the non-profit ARRL is the mosquito repellant manufacturers association.

The majority of ARRL-sponsored activities could be considered "social" in the sense that, as does Field Day, they involve a bunch of hams getting together to do something. Others are social in the sense that the goal is for hams to contact each other on the radio and exchange information. But one activity is both solitary and nerdish: The "Frequency Measuring Test." In the FMT, the ARRL uses its headquarters radio station W1AW to transmit signals at several different frequencies in the amateur radio bands, and it is up to the entrants to measure the frequencies as precisely as they are able to. For example, the contest announcement may say that W1AW will be transmitting on (approximately) 1853, 3586, and 7039 kHz. The mission would be to refine that to 1853,059.2 Hz, etc.

In order to do this, you have to have a number of capabilities:

You have to find the signal with your receiver. Although this is usually easy, you must contend with "band conditions" and weak signals, not to mention possible interference from other users who share the frequency.
You have to identify the signal. They identify in Morse Code, so if you don't speak Morse you might pick the wrong signal.
You have to measure the frequency of the signal as precisely as possible. Depending on your equipment, which you have hopefully recently calibrated, you might be limited to measuring to the nearest 100 or 10 Hz. It actually takes a bit of precision work to get to the nearest Hz and fraction thereof. (A "Hz" is just a "cycle per second" in new clothes. An AM broadcast station, e.g., 1010 WINS, transmits on 1,010,000 Hz.)

Needless to say, the FMT is one of my favorite ham (and nerd) activities. I can sit, (alone, of course,) amongst the glowing lights and screens of my test equipment and reduce hours of effort and kilobucks of hardware to three or four precise numbers, each of which agrees closely with its counterpart number measured at ARRL Headquarters over a hundred miles from here. What's not to enjoy? But the best part of the FMT for me was that I was able to prove the ARRL wrong in one of their assumptions about the contest. As every SETI League member knows, there is no greater nerd satisfaction than doing something that you are told can't be done. My joy was ameliorated to some extent in that I think the "can't be done" assertion wasn't well thought out. In early announcements of the FMT in QST, ARRL asserted that accuracies greater than 1 PPM weren't possible. Hah! (1 PPM is 1 Hertz per megahertz. For example, it's "impossible" to measure the 7039 frequency to better than 7Hz. Did I mention "Hah!"?) This surprised me when I saw it, and in later years they dropped the claim. I'd like to feel that I contributed to their change of theory.

So, how well do I do in the FMT? Between 2003 and 2006, my very worst result was .26 PPM, and most were .1PPM or (much) better. Is this close enough for SETI work? Probably. Our Project Argus stations typically monitor the spectrum surrounding 1420 MHz, the precession frequency of interstellar neutral hydrogen atoms. An error of .1 ppm at the hydrogen line translates to 142 Hz of frequency uncertainty. If our digital signal processing bins are (typically) 10 Hz wide, this means we can be off by fourteen bins in assessing any received signals, and never know it. Of course, since our software monitors many thousands of bins at a time, this might not particularly matter, at least in the signal acquisition phase of our search.

This editorial is an adaptation of a Richard Factor blog item, used by permission.

Disclaimer: The opinions expressed in editorials are those of the individual authors, and do not necessarily reflect the position of The SETI League, Inc., its Trustees, officers, Advisory Board, members, donors, or commercial sponsors.

email the
Webmaster

Top of Page