High precision 1Hz to 100kHz frequency source, derived "off-air"

During the design stage of the Synchronous Clock Driver, featured in EPE Sept '01, doubts arose as to the accuracy of the frequency meter being used to check and adjust the output frequency of the project. Since this instrument was the author's primary means of measuring frequency, the difficulty arose as to how it could itself be checked and, if necessary, adjusted.

One of the broadcast radio carrier signals appeared to be the best way of obtaining a suitable reference. Most British readers will know of the time signal transmitted at 60kHz from Rugby, but this isn't really suitable for frequency testing since it is pulsed on and off by the data signals it carries. Another source that seemed better suited for the purpose was the 198kHz "longwave" carrier for Radio 4. Originally this was intended for use as a national frequency standard and its accuracy is still maintained to an incredible level, having a Rubidium frequency source as its reference with constant monitoring by the National Physical Laboratory.

In fact, the accuracy is claimed to be one part in 1011, which translates to about a third of a millisecond per year of error. This should be more than adequate as a standard for most home workshops!

Various circuits are available for receiving and using this signal "off air", so one was soon obtained through the good offices of one of our better radio magazines and hastily constructed. In fact, the "error" of the author's meter turned out to be of insignificant proportions, but by then the "design bug" had bitten. Could a better version be built offering more useful output frequencies, such as 100kHz and the decades beneath it, 10kHz, 1kHz and so forth? Such a source would be very useful for calibrating all sorts of equipment, including oscilloscopes and frequency meters, and perhaps also in the testing and adjustment of clocks.