OBSERVATIONS BMEWS'ed Al GM1SXX & John LA2QAA
Since winter 2003 when John and Al were collecting A-O-7 telemetry frames and analysing them, we noticed something strange. As the bird approached on its ascending node, frequently over Russia before taking its trip South again toward the UK and Norway, the transponder would go from working fairly well while the telemetry system was not quite so good, to a state of weakened signals and severe FM'ing of the CW beacon.
We spent quite a bit of time wondering what caused this signal degradation since the transponder normally had no 'big signals' from radio amateurs overloading it when we were monitoring. In fact during the climb Northwards, signals could be quite good and consistent on both our modest antenna systems .. 8 element yagis for both 2m & 70Cms (linearly polarised) at both stations.
And yes, we *did* consider that not using CP polarisation could have been a factor BUT twisting the antennas through 90 degrees made no difference to the large signal fades we encountered.
With the bird in good illumination at the time the solar panels should have held the transponder output power fairly constant, but it seemed that 'something'' was pulling down the transponder and that 'something' was not radio amateur activity on the bird. When the transponder performance deteriorated, it was often accompanied by a crackling noise on SSB on the 2M down-link, at least in Scotland, while the beacon would warble and sometimes quit altogether.
At the QTH of Al gm1sxx, a background crackling noise was frequently present in the downlink passband during passes heading toward the UK. To test that this was not a local effect, the 2M rig and antenna were taken out into the countryside and another pass monitored. The crackling noise was still audible in the transponder passband.
We started looking causes for the desense and crackling and had strong suspicions that the BMEWS RADARS in the Northern hemisphere were the most likely culprits with the finger of blame pointed at RAF Fylingdales (which is actually a mainstay of the US BMEWS system).
This suspicion was reinforced after monitoring many orbits over the Christmas holidays in December 2003. Al wrote some simple software to decode the telemetry frames while John collected CW telemetry data and forwarded it for analysis. It was *very* obvious that something was severely 'clobbering' the transponder as the bird flew over Europe.
One thing to remember is that when AO-7 was launched (November 1974) the RADAR systems in use then were not the same types as today's modern electronically steered computer driven pulsed repetition types.
A web search showed that BMEWS has been implicated in other de-sense situations in the past.
File photo of Fylingdales BMEWS SSPAR.
About BMEWS
The 40-meter high truncated pyramid at Fylingdales that forms part of the BMEWS SSPAR RADAR system has 3 faces each 84 feet across containing an array of 2,560 transmit/receive modules each with a circularly-polarised `Pawsey stub' antenna. Each module produces a transmitter power of 340 Watts and this gives an overall mean power output from the three faces of 2.5 Megawatts Since the RADAR uses gain antennas, the ERP will be far higher than this figure. The Fylingdales SSPAR RADAR was built by Raytheon.
The RADAR operates in the 420-450 MHz band, has a range of over 3000 miles and is able to operate over a full 360degrees in azimuth . The main radar beam is directed under computer control and limits are in place so that the main antenna lobe(s) are at least 3degrees above the horizontal. The residual side lobes however can reach ground level.
The array has a practical deflection angle 60 degrees each side of the main lobe with a vertical radiation angle that covers from 3 to 80 degrees of elevation ... that's 2 x 60 degrees = 120 degrees in azimuth and 73 degrees in elevation ... from 3 sides ... = 360 degrees.
The Pulse Repetition frequency is adjusted by the system's computers to give the optimum results and the beams are swept in azimuth and elevation under computer control.
With regard to the Oscar-7 transponder de-sense problem it's true that Karl Meinzer added a form of "radar filter" to the spacecraft electronics but that was designed for the radars anno 1974. We seriously doubt that it has much effect on today's modern pulsed radar systems. The RADAR filter was nothing more than a careful choice of transponder AGC characteristics to attempt to cope with the Pulse Repetition Rate PRF, of the RADARs in use at that time.
This RADAR ‘filter’ is unlikely to keep out the signals of a modern BMEWS installation which sports variable PRF’s to suit the objects being scanned for.
Al gm1sxx contacted an amateur radio club member in the Scarborough area close to Fylingdales for feedback on the RADAR’s effects on the 70Cms band. He was told that the interference sounds like the noise made by a zipper fastener being ‘unzipped’. That of course is what the interference sounds like when close to the RADAR installation.
Whatever is interfering with Oscar7 generally is sometimes inaudible while other times having the characteristic 'crackling' noise made by a 'dry joint' being touched.
There are three other similar BMEWS arrays in the Northern hemisphere. not counting the ones in the former USSR and in the USA..
One *theoretical* way to find out whether the BMEWS RADAR affects AO-7 is to listen to the beacon and check the voltage (via the telemetry) when the satellite passes through the main lobe of the radar. Given a good tracking program and an "educated guess" as to the timing of the satellite's transit through the RADAR beam ... this ought to feasible but not easy.
AO7 telemetry is slow… its sent as CW and a complete single frame takes quite a while to copy. In reality, it would be a difficult task.
A linear transponder satellite with an AGC telemetry channel and a 70Cms uplink would be the ideal test-bed for checking the effect of BMEWS RADARS on satellites in this band.
The article reproduced below shows that RADAR has been a considerable nuisance in the past to amateur radio birds with 70Cms uplinks.
It is noteworthy that when the Fylingdales RADAR was switched off for maintenance, the 70Cms command uplink on Oscar13 went from being heavily interference plagued to almost perfect. Note the following paragraph, culled from a document translated from the AMSAT-DL Journal by Don Moe and reproduced in full later)
---------------------------------------------------------------------------------------------------------------------------------------------
The article said….
OSCAR-21, like AO-13, has a command receiver and uplink frequencies in the 70cm band also. The situation here (Germany… gm1sxx) is likewise quite difficult. Only low passes in the East can be used at all when the English radar is below the horizon. The success ratio for uploads of long data files is unfortunately mostly below 30% so that much time is lost to numerous repetitions. On March 25 and 26 the radar was coincidentally out of service for maintenance and the success ratio rose to over98%! Unfortunately this is only seldom the case and for obvious reasons at irregular intervals and without prior notice.
---------------------------------------------------------------------------------------------------------------------------------------------
This is a very strong case for blaming the Fylingdales SSPAR for desensing Oscar7’s uplink. If you can copy it, the interference takes the form of a ‘crackling’ noise. It sounds a bit like a loose connection or dry joint in a radio.
It is unlikely that we will ever be able to positively prove the case against BMEWS. Oscar-7 telemetry is intermittent and even when working properly is a very slow mode of data transfer for a 'real-time shot' of the situation.
LA2QAA & GM1SXX Jan 2006
===============================================================================================
Circular polarisation is ONLY circular down the boresight of an antenna. Put simply, if either the satellite or the ground station (or as is more often the case both) are not pointed DIRECTLY at each other, you DO NOT have CP. Instead you may have elliptical or slant polarisation but it simply cannot be circular.
BMEWS... Ballistic Missile Early Warning System
SSPAR..Solid-State Phased-Array RADAR
PAVE PAWS is another name for the SSPAR type RADAR systems. PAVE is an Air Force program name, while PAWS stands for Phased Array Warning System.
PAVE PAWS is the US component of the BMEWS system. The RADAR at Fylingdales while located at an 'RAF Station' is a major source of orbital data to the USA and is essentially the same Raytheon RADAR system used by PAVE Paws although unlike others, the Fylingdales RADAR covers 360 degrees in azimuth.
===============================================================================================
The following is from an article written in 1993 for the AMSAT-DL Journal and translated by Don Moe, DJ0HC/KE6MN).
It followed comments on the spontaneous activation of the LEILA 'anti-alligator' system on the 70Cms uplink while the transponder was free of any amateur signals.
What we see on LEILA is probably related to this?
PAVE PAWS
Since 1987 the US Air Force has been operating a total of four so-called "PAVE PAWS" radar systems located in the states of Texas, Georgia, California and Massachusetts. The first such installation went into operation in 1979.The radar system serves primarily to recognize launches of ship-based and intercontinental ballistic missiles (ICBMs) and to track their trajectories. Additionally it also simultaneously observes and tracks more than 6000 satellites and other objects in Earth orbit. The data are sent to NORAD in Colorado for analysis. Orbital parameters and Kepler elements for nearly all satellites are administered and published there, including among others weather and naturally amateur radio satellites. This data is available at no charge over a variety of computer networks including Celestrak and Space-Track.
Each of these gigantic "PAVE PAWS" radar antennas consists of so-called phased arrays [1]. These phased arrays are built around thousands of separate Pawsey-Stub antennas in order to generate a directional radar beam to scan the skies.
In contrast to customary radar installations in which a parabolic dish is mechanically turned and aimed, the directional control is determined by changing the phase angle between the individual dipole elements. Rather than taking minutes, a computer can electronically aim at any desired point in the sky within a few milliseconds!
A "PAVE PAWS" radar is built around a three-sided, 32 meter high building and can cover an arc of 240 degrees. Phased array antennas with a diameter of 31 meters are mounted on two sides of the building. Each side can hence scan 120 degrees of azimuth and 80 degrees in elevation. The transmitter power is provided by transistorised amplifier stages consisting of four 100Watt transistors combined in each transmitter module to drive a single radiating element. This is repeated for a total of 1792 elements.
1792 times 400 Watt requires over 700 Kilowatt of pure transmitter power, not considering the enormous antenna gain. Objects with a surface area of 10m2 can thus be observed at a distance of nearly7000 km.
Two older early warning systems for ballistic missiles (BMEWS) also existed in Thule, Green-land and Fylingdale, Yorkshire, England. In the meantime both BMEWS systems have been converted and modernized to PAVE PAWS radar installations. The PAVE PAWS radar in Thule, Greenland went into operation in June, 1987. The English (actually British...GM1XX) system followed a few years later. In contrast to the radar in Greenland, the English PAVE PAWS radar has even a third antenna array and can thus cover an entire 360 degree circle.
Each of the three antenna surfaces consists of more than 2500 individual antennas. Unfortunately the operating frequencies of PAVE PAWS radars lie within the 70cm amateur radio band, and in particular between 432 MHz and445 MHz. One of the three broad-band main carriers is almost exactly on 435 MHz and a second one is close to 439 MHz. The frequency spectra also overlap to some extent. The 435 MHz satellite segment is thus particularly affected. Ingeniously the range below 432 MHz is not occupied at all.
Only around 422 MHz do we find another carrier. At the enormously high radiated power levels of these radar systems, consequences to satellite operation are unavoidable. After the new PAVEPAWS radar installations went into service in Greenland and England, our situation became correspondingly earnest, particularly over the past few years since the launch of AMSAT OSCAR-13, but OSCAR-21 with RUDAK-II is also seriously affected. To support command station operations,OSCAR-13 relies upon separate frequencies to allow the command stations direct access to the onboard computer. Depending on which transponder is active, the command station must call either the Mode-B command receiver on70cm or the Mode-L command receiver on 23cm.Since Mode-B operation predominates, particularly during emergencies or during the launch phase, the 70cm command receiver plays a very important role.
Shortly after the launch of AMSAT OSCAR-13in June, 1988, I discovered, as the primary command station in Europe, that during orbits whenOSCAR-13 covered large portions of the USA, only very poor command control was possible.
In contrast, command control on easterly orbits was quite reliable even at very low power levels. Therefore we were able to accomplish our tasks. Initially we did not even consider radar interference, but suspected faults in our own stations. Over the course of time, the apogee of OSCAR-13 has continued to drift further north and the satellite correspondingly covers more of the northern hemisphere for longer periods of time. Unfortunately this also includes the radar installations in the northern parts of the USA and of course the modernized radar sites in Greenland and England.
There is practically not a single orbit where one of these PAVE PAWS systems is not in view of OSCAR-13. In the early stages, this was not too serious for the reasons just mentioned. Incidentally, in February, 1993 the apogee of OSCAR-13 has reached its northern-most excursion at 57.6 and is now gradually moving to-wards the equator. On the analogue transponder of Mode-B, the radar interference can scarcely be noticed and nobody seems bothered by it. However the interference means the end of digital communications via satellite.
During the first two years following the launch ofAO-13, a reload of the software was indeed required on two occasions and the last time it took nearly 6 hours to reprogram the onboard computer because of the radar interference. A portion of the upload had to be performed by Graham Ratcliff, VK5AGR, in Australia. Because Graham lives in the southern hemisphere, he does not suffer from radar interference, but his access periods are much more limited than ours are in the northern hemisphere. The problems that we here in Europe have with the 70cm command uplink affect not only OSCAR-13, but also AO-21 with its RUDAK-II experiment, for example. In this case, the digital uplink frequencies are in the satellite band just above 435 MHz. The downlink is at the upper end of the 2m satellite segment. Although AO-21 is in a significantly lower orbit compared to AO-13, it is not spared by the radar in Fylingdales.
Again we have noticed that command access is extremely poor when the PAVE PAWS radar is within view. Inspection of the disrupted uplink blocks have shown an agreement with the pattern of interference to AO-13. In FM mode, the pulse interference can be heard quite well and sounds like crackling noises. On AO-13 or AO-10 these same irregular pulses can be readily heard on a weak CW carrier as interruptions or crackling noises.> [.....]>
OSCAR-21, like AO-13, has a command receiver and uplink frequencies in the 70cm band also. The situation here is likewise quite difficult. Only low passes in the East can be used at all when the English radar is below the horizon. The success ratio for uploads of long data files is unfortunately mostly below 30% so that much time is lost to numerous repetitions. On March 25 and 26 the radar was coincidentally out of service for maintenance and the success ratio rose to over98%! Unfortunately this is only seldom the case and for obvious reasons at irregular intervals and without prior notice.
(Due acknowledgement is made to both AMSAT and
Don Moe in reproducing the above item)