This is an odd digital signal I first stumbled across on 9 August 2009 at about 1745 UTC (1045 local). I first saw it on a frequency of 13881.3 kHz.

Talking with others in the #wunclub IRC channel I found that several, including Jon-FL and TheWeb, had been looking at this off and on for the last month or so. No one appeared to have any real idea what the signal is for or where it originates.

The signal has been seen on other frequencies, and sometimes on all known frequencies simultaneously. Currently known frequencies include 6081, 7431, 7481, 9531, 9581, 9631, 9831, 9881, 10031, 11631, 11681, 11831, 11881, 11981, 12031, 13881, and 15381 kHz. All of those frequencies have been RXed at my station. If you note the trend of those freqs you will see they all end in either 31 or 81, and several show a paired freq 50 kHz away. If this trend holds there are at least 4 frequencies still to be detected. The signal has added and subtracted frequencies (form the known set) in mid transmission, without taking the transmitter off the air.

The signal consist of pulses of about 8 msec duration, so far I have seen no major deviation from that length. Every tranmission I have measured comes out to 8 msec PW within the error of the software I am using.

Pulse repetition interval varies, so far I have seen 11, 22, 33, 43 (possibly 43.666), 55, 76, and 88 msec. I have seen the signal change PRI in mid transmission, it occurs as a step, not a drift.

Each pulse consist of multiple tones. So far I have seen tone spacings of 6.6, 11.8, 13.1, 18.5, 19.5, 23.2, 31, and 46 Hz, however I have not been able measure every transmission I have heard. As the RF drifts up and down the tone spacing stays constant, even during the step in frequency. I have seen the transmission change tone spacing in mid signal without going off the air.

Either the transmitter is not particularly stable, and drifts up and down in frequency, or the intelligence on the modulation drifts in frequency. It is unknown if this drift is intentional or not. The transmitter or the intelligence also periodically jumps in freq +/- about 230 Hz. Both of these factors lead to different reported center freqs, for example on 13881 the center freq has been seen to be anything from 13881.05 to 13881.9 kHz. Jumps as short as one pulse have been noticed, or it may set on one freq (with the mentioned drift) for many 10's of minutes. It has been reported that the signal jumps both up and down in its steps, meaning at least three "channels" or center freqs. So far I have only seen it jump one step, normally down from where it starts, leading to two center freqs about 230 Hz apart. Of course, as they both show the same drift it is hard to define the two channels until you see the step itself.

Around mid September the signal shifted up a bit in frequency, most commonly about XXXX1.7, ie. 13881.7, 11981.7, etc. Sometime later, possibly in December of 2009, it dropped a couple hundred Hz, now most commonly seen at about XXXX1.5 kHz center freq (13881.5 kHz, 11981.5 kHz, 6081.5 kHz, etc). It still exhibits the same drift, so these are approximates. After this shift up in frequency the occasional 230 Hz "step" seen before has been much less common.

The signal is not always present, but does generally follow a given schedule.

Jon-FL has seen the signal drop pulses on one freq and include them on the other in a 50 kHz pair. Specifically he has seen pulses dropped on 11681 appear on 11631. I have not yet seen this but then my reception of the 11681/11631 pair is a little marginal most of the time.