This signal is one I have been watching since late 2011, in my log it is called “Big Tones”. Originally it started with 1 kHz tones (pulse repetition rate), and has also used that pulse spacing recently, but most often it uses the 300 Hz as you have shown here.
I cannot tell you what it is (I simply do not know, although I have a few rough guesses) but I can tell you what it is not. It is almost certainly not a radar, and it is almost as certainly not a sounder.
So, why not a radar?
With any radar you can tell certain potential performance parameters based on the transmitted signal. This does not give you the complete picture of its capabilities, but it does define a few boundaries of performance. Depending on the radar type (pulse, FMCW, pulse Doppler, etc) you can tell different things from its transmissions.
With a simple pulse transmission like this one (there does not appear to be any modulation on the pulses that are significant for radar applications) three parameters are very easy to determine. Maximum possible unambiguous range and minimum range resolution. Maximum unambiguous range is simply the maximum range at which the radar can detect a target and be sure of the range. Minimum range resolution is the minimum distance between two targets that is possible and the radar still be able determine that there are two targets there, not just see the two (or more) as a single target. The same factors that determine minimum range resolution also determine minimum possible detection range, the closest target that can be seen.
Maximum unambiguous range is determined with the formula R = (c x t) / 2. R is the max range, c is the speed of light, and t is the radar PRI (pulse repetition interval, time between pulses, leading edge to leading edge). The speed of light and the time need to be in the same time unit, normally seconds. Whatever distance unit is used for c will determine the distance unit for R.
We will call c = 300000 km/sec. The PRI for this signal is 3.333 msec, or 0.003333 seconds (this is 300 Hz, as in your example, but I have seen it as high as 303 Hz).
(300000 x 0.003333) / 2 = 499.95 km
So the maximum unambiguous range is 500 km. OK, this seems a reasonable max usable range for a radar, even if a bit short by some OTHR standards.
Minimum range resolution is determined using the formula Rr = (c x PW) / 2. Looks familiar, ah? Radar is very time dependant, and radar wave travel at the speed of light after all. Rr is the minimum distance apart two targets can be and be detectable as individual targets. c is as defined above in the maximum range formula. PW is the pulse width.
The pulse for this signal is 900 microseconds, or 0.0009 seconds (I have seen shorter pulses also, down around 835 microseconds).
(300000 x 0.0009) / 2 = 135 km.
(edit) I originally failed to mention that Pulse Width also determines the minimum range at which a simple pulsed monostatic radar can be used. It cannot listen while the pulse is being transmitted, and so during this pulse transmission time, and for some recovery time after the pulse ends, the radar is “blind”. Since time is distance then the minimum range a radar using the above pulse width can see would be a bit over 270 km.
So, if this signal were a radar it would have a maximum usable range of 500 km, a minimum usable range of 270+ km, but it would not be able to tell two targets apart unless they were over 135 km apart. Basically it would end up with less than 2 usable range cells in its full range. This means that if there were 2 targets in its illumination cell that were each 130 km from the other then it would see one large target from min range to max range.
Not a very affective design
![Wink ;)](https://www.hfunderground.com/board/Smileys/default/wink.gif)
And the 1 kHz version of the signal is even more limited, since it uses a 50% duty cycle (square wave). So, in my opinion and based on a few years in the field of radar, almost certainly not a radar. There are also a couple of other modulations seen on the signal, such as a 4.5 kHz sine wave of low modulation depth.
I have been seeing this signal since late 2011. It uses several frequencies, and seems to hit them pretty dependably / repeatedly. It generally starts at the lowest freq and then moves up as time goes on, but it frequently goes back and revisits
Some of the frequencies seen (not a complete list, but the most common ones I have seen, each used at least 3 times on separate dates in my logs):
4040
4810
5385
6400
6970
7595
7990
9050
9803
11065
12045
13530
16060
17048
18036
19270
20510
21460
23180
24450
26875
Video of the signal on 6970 in 2012 (PRF and PW slightly different from your example, but the same signal):
http://www.youtube.com/watch?v=3AIZLc-ZqW0Picture of the spectrum (if you have no other way to detect the pulse width the reciprocal of the time from peak to first null will yield the PW, or half the frequency from first null on low freq side to first null on high freq side):
![](http://www.pbase.com/token/image/145847556/large.jpg)
Hope some of this helps.
T!