Tracking Briz-M on 18 November 2007

Sven Grahn

I watched the launch of the Proton carrying the Swedish Sirius-4 satellite (built by Lockheed-Martin) from Baikonur at ILS' website at 2239.47 UT. Then went to bed and set the alarm clock for 0225 UT. I had set up the radio to use my 900-1000 MHz helix that points at low elevation low towards the southeast. As a preamp I had connected the SSB Electronic amplifier for 1000 MHz I had used in 2006 to track Molniya satellites.
 
My antenna set-up in Sollentuna
I was monitoring 1020.5 and 1018.5 MHz as per data in the Proton user's manual available online. he relevant page is reproduced below.
Proton-M radio systems

First reception, on 1020.5 MHz, 0233-0239.51 UT

I must say I expected maybe a weak diffuse wideband bump on the screen. Nothing had prepared me for what happened. The alrm clock was set at 0227 UT. I was flipping back and forth between frequencies on the AR 8600 and then, at 0233 UT, I saw the strong carrier on 1020.5 MHz, with a fast rising doppler! The C/N shot up to 16 dB. It faded in and out and then just vanished at 0239.51 UT.  The waterfall display shows the fading signal and rising Doppler during this first reception of the signal.
First reception on 1020.5 MHz
 

Second reception, on 1018.5 MHz, 0240.20-0246.54 UT

I retuned to 1018.5 MHz and first saw a single carrier at 0240.20 UT, soon replaced by two carriers +/- 6 kHz from the centre. I now turned on the demodulator and heard audible modulation using USB. Sort of a purring sound. Doppler was still rising. Then the signal vanished again at 0246.54 UT.
 Signals on 1018.5 MHz


The spectrum from my Adobe Audition software shows what the demodulated (Upper SideBand) looks like. By clicking here you can hear a short piece of the signal.
Audio spectrum of 1018.5 MHz signal

Third reception, on 1020.5 MHz, 0248.50-0258.20 UT

I found the signal again on 1020.5 MHz at 0248.50 UT. Again the signal was a strong, very strong carrier with audible modulation, but the modulation was different, a pulse train... C/N increased to 24 dB! The doppler rate was now slowing down and the doppler maximum was reached at about 0252 UT. I then started taking down frequencies by hand:

1020.517667 MHz at 0253.30 UT
1020.517633 MHz at 0253.42 UT
1020.517545 MHz at 0254.56 UT
1020.517409 MHz at 0256.04 UT
1020.517274 MHz at 0257.13 UT

 The waterfall picture below shows that the signal was quite strong during this reception.
Third reception - on 1020.5 MHz


Here is what the signal on 1020.5 MHz looks like in overview. It seems that there is some kind of “frame” marker every 6.6 seconds.
Frames every 6.6 secs?


On the smallest scale the signal looks like below. By clicking here you can hear a short segment of this signal.
Pulse train on 1020.5 MHz

 

Single-carrier mode splits to dual-carrier mode at command-off at 0258.20 UT

The signal was commanded off at approximately 0258.20 UT. Just before the signal vanished the single carrier on 1020.5 MHz had split into two carriers similar to what I had seen on 1018.5 MHz. However, the carriers were now at +/- 16 kHz instead of +/- 6 kHz apart as on 1018.5 MHz. The waterfall picture below shows it clearly.
 Command-off

After this I set the waterfall display to one update every 60 seconds and the AR 8600 receiver to 1020.5 MHz. The Briz-M now climbed towards apogee at 35800 km where rocket engines would start again and then the Sirius-4 satellite would be separated.

It is a bit funny that I saw a doppler maximum. If one runs an orbit simulation a doppler minimum would be expected- The time of the minimum is 0249 UT, which agrees well with the maximum! Often one can see reverse doppler on the SDR-14 display. A strong signal just outside the range of the screen "folds back" into the display.

Fourth and fifth receptions, on 1020.5 MHz,  at 0400-0430 UT and 0700-0741.14 UT

When I woke up again just after 8.20 local time I rushed out of bed to look at the waterfall that has been running on slow scroll throughout the morning hours. It was immediately obvious that the signal on 1020.5 MHz had come up again at about 0700 UT (the time stamp on the SDR-14 screen is such that the top of the lettering corresponds to the actual time). Again two carriers (I have measured them to be at exactly +/- 16 kHz apart). I could see the doppler maximum a few minutes before 0725.  The signal level increased after the maximum had been reached. Just a few minutes later the frequency curve did not turn so sharply down, it sort of distinctly leveled out. Unfortunately I fumbled and lost the spectrogram where one can see this. However, the piece of the waterfall below shows the lower of the two carriers and a clear maximum. The orbit simulation shows no maximum. Why is there a maximum? Possibly because of transmitter drift. 

Doppler maximum

Indeed, there is actually another single-carrier trace with a maximum on the overnight waterfall at about 0400-0430 UT. If one compares the frequency at the 0415 UT maximum with that of the the frequency precisely between the dual-carrier maxima at 0712 UT the frequency has drifted down 7.6 kHz. Since the display is frequency-inverted there is actually an increase in frequency of 7.6 kHz. Whe  running simulations using the pre-launch orbit data from ILS and element sets from SpaceTrack the frequency increase is estimated at 7.6 kHz - the same number. I promise, I measured the frequency shift before computing. it. The figure below shows the entire waterfall that I saw when waking up.
Briz-M spectrogram 18 November 2007
It is instructive to compare the little maxima at 0415 and 0712 UT. It appears that there were "dropouts" in the signal with its maximum frequency at 0415 UT. However the shape of the "maxima" are the same. See picture below:
Briz frequency transient



The signal was commanded off at 0741.14 UT.  Just before that I tried to measure the doppler rate, but it was not easy to read two numbers at the same time. The last reading can be off in time, so the two first are the most reliable since they are far apart in time:

1020.515147 at 0739.06
1020.515113 at 0740.10, i.e. -34 Hz/64 sec = -0.53 Hz/s, or actually +0.53 Hz/s because of the spectrum inversion
1020.515106 at 0740.36, i.e  - 7 Hz/16 sec = -0.44 Hz/s, or actually +0.44 Hz/s because of the spectrum inversion

Orbit simulations show that the doppler rate was 0.52 Hz/s, a reasonably good agreement.


I thought that this was the end and that the spacecraft had separated from the Briz-M. But I had not checked flight plan. If I had done that I could have seen that Briz ignition was planned for 0730.03 UT. This is probably when I saw the doppler curve level off. Briz burnout was planned for 0737.15 UT. Both these events occurred within this reception period.
 

Sixth reception, on 1020.5 MHz, 0750.20-0754.41 UT

To my surprise the signal (+/- 16 kHz carriers) came back strongly on 1020.5 MHz at 0750.20 UT. There was almost no doppler shift. Just before LOS I measured the two carriers to be located at 1020.513 MHz and 1020.481 MHz (the three decimal beyond these numbers were identical!), so the center fx was 1020.497 MHz. Command-off occurred at 0754.41 UT.

Nominally, spacecraft separation should have occurred at 0752.47 UT. So, after this reception I did not hear from Briz-M again.

Post-Script on spectrum inversion

In December 2007, when tracking satellites on VHF, I discovered that my receiver, the AR8600, inverts the internediate frequency output at 10.7 MHz for input frequencies around 150 MHz. It probably does the same at 1 GHz, while at S-band it does not. The receiver changes operations mode for various frequency ranges, so this is a probable explanation to the spectrum inversion giving me so much headache above.

Back to Space Tracking Notes