Three little unsolved riddles from early radio tracking of Soviet satellites

Sven Grahn

Switching sequences

During the flight of Kosmos-115, 20-28 April 1966, I observed for the first time a pattern in the frequency-shift pulse-duration modulation transmissions on  19.995 MHz that became a regular pattern for a certain type of Soviet reconnaissance satellites. The transmitter on 19.995 MHz was heard to be switched off twice during a pass over Europe and each time it was switched off the silent interval was exactly the same - 25 seconds. The ground track of Kosmos-115 during that first observation is shown below.

This phenomenon was dubbed by the "Kettering Group" and happened only during the last "radio-active" pass over Soviet Union each day. It affected only Zenit-2 type of reconnaissance satellites. It was last observed by myself from Kosmos-696 in 1974 - eight years after the first observation..

The silent interval varied from satellite to satellite, but was always constant for a particular satellite. The silent interval fell in the range 25-55 seconds. The period between the silent intervals varied and was sometimes as large as 6 minutes. A table showing some the silent period duration for a group of Zenit-2 satellites.

I have never found an explanation of this phenomenon and it is difficult to imagine one. I have toyed with idea that onboard power was needed for some urgent purpose and all non-critical loads were switched off, but what was the urgent purpose? Anyway, I invite ideas to solve this little mystery.
 

Two tone signals

Soviet reconnaissance satellites of the Zenit-4MK "Germes" type are often called 3rd generation high resolution reconnaissance mission. These satellites were manoeuvrable and started flying regularly in 1970 after their probable introduction with Kosmos-317 in  December 1969. This type of satellite was observed to transmit on 19.989 MHz. The signal on this frequency cycled back and forth between two adjacent frequencies about 1100 Hz apart. The cycling rate, i.e. the time it took to return to the initial frequency was approximately 1.6 seconds. The cycling rate varied a little from satellite to satellite - something that probably was not intentional, but was used by amateur trackers to distinguish between satellites of the same time in orbit simultaneously - which happened from time to time. So, the signal was simple simple frequency-shift-keying that has absolutely no information content. I have always wondered about this fact. Why bother to transmit this signal if nothing was transmitted? (The spectrogram on the right shows a segment of signals from Kosmos-364).

Four year after the Kettering Group had discovered this signal we stumbled upon additional information about this strange signal, which we called "two-tone" in everyday exchanges. We discovered that the two-tone signals were present on two frequencies. A signal that sounded exactly like the signal on 19.989 MHz was detected on 39.978 MHz, i.e. on exactly double the frequency of the short-wave signals. What was this? An unattenuated unintended harmonic of the 19.989 MHz transmitter? Well, the signal on 39.978 MHz was very strong and not just a weak harmonic. Was it a separate transmission modulated by the same signal as that on 19.989 MHz. This controversy could have been cleared up if a recording of the 39.978 MHz signal had been preserved, but unfortunately no such recording exists. If the 39.978 MHz signal was a simple harmonic the frequency shift of the 39.978 MHz signal would have been double that of the 19.989 MHz. If not there could have been separate transmitters or two transmitters fed from the same basic oscillator but modulated after the carrier had been multiplied to the appropriate frequency. In either case  the signals would have been coherent.

What possible use could coherent transmitters have? Perhaps as an ionosphere beacon system to provide, at a very low cost, an almost continuous system for monitoring electron content and other parameters of the ionosphere. Is there any indication in Soviet scientific publications of such data? Two-tone signals continued to be used until the early 1990's. I tracked my last signal on this frequency on 10 March 1991 when I picked up Kosmos-2136.
 

The fourth frequency on the Soyuz control panel: where was it?


"Telemetry bursts" were observed on 20.008 MHz from Kosmos-140, -186/188 and 212/213 as described in another article on this web site. The signal was switched "on" for 30 seconds and the "off" for 90 seconds. This pattern matches nicely the sequencer on the Soyuz instrument panel that was capable of cycling between four frequencies (see figure on the right). The explanation reads:

"Each time the SHORT WAVE XMTR I FREQ button is depressed the frequency cycles through FREQ I, FREG II, FREQ III, and FREQ IV and the lights 25,26,27, and 28 so indicate".


The "30-90" pattern certainly indicates there were indeed four "slots". We know that the frequencies 15.008, 18.035 (later 18.060 MHz) and 20.008 MHz used by Soyuzes. One of these frequencies would would fit propagation conditions and "get through" to the receiver. But what was the fourth "slot"? Was it empty or was there another frequency used by Soyuz that we do not know of? Ideas are herewith invited!


Back to "Space Tracking Notes"