THE VIEW FROM BELOW: THOUGHTS ON PHASE 3D GROUNDSTATION REQUIREMENTS Ed Krome KA9LNV Columbus, Indiana USA ka9lnv@amsat.org To date, much has been written about the features and technical aspects of the incredible OSCAR Phase 3D satellite. This paper, the "View from Below", looks at some of these almost magical things and what they mean to us on the ground. It also includes thoughts on both how to approach the development of an individual groundstation and describes some ideas on possible equipment and antenna configurations. It is not intended to be the last word on any of this subject, but merely the observations of an experienced OSCAR user and microwave experimenter. Features P3D has a myriad of features and characteristics, both of the intended orbit and of the satellite itself. Understanding these features and what they mean to us on the ground will help each individual make informed decisions on how to get the most out of P3D. This paper will also compare P3D characteristics to those of the existing Phase 3 satellites, AO-10 and AO-13, with which most of us are familiar. Orbit Phase 3D will be in a high inclination elliptical orbit similar to that of the other Phase 3 satellites, but higher and with a different period. The Ariane launch vehicle will leave P3D in a 10 degree inclination GTO, or Geostationary Transfer Orbit. From there, P3D's 400 Newton (or 90 pound force, for those of us who haven't quite gotten with the SI system yet) bi-fuel main motor will move the satellite to the desired orbit in several burns. The final orbit will be at 60 degrees inclination, with a perigee of 4000 km and an apogee of 47000 km. From there, the orbit will drift slightly until it attains about 63 degrees inclination. Compare that with AO-13's present orbit of 325 km perigee and 38500 km apogee, at 57 degrees inclination. The orbital position of P3D was designed to give a period of 16 hours, as compared to AO-13's period of 11.44 hours. This 16 hour period provides a unique characteristic of P3D's orbit; the fact that the satellite will reappear at the same position in the sky every 48 hours. While this characteristic will definitely not remove the requirement for computer tracking software, it will be very useful in setting up regular schedules and nets. When I first heard this "same place every 48 hours" phenomena, I wondered how it could be. It works as follows. A satellite's orbit is fixed around the Earth relative to the fixed stars. The Earth rotates as the satellite orbits. P3D attains apogee every 16 hours. Sixteen hours after the initial apogee (when the satellite attains apogee again), the Earth has rotated 240 degrees, or 2/3 of the way around. If the initial apogee occurred over (for example) Kansas City, Missouri, at 100 degree West Longitude, the apogee 16 hours later would occur at 340 degrees West, approximately above Tokyo. Sixteen hours later, at the next apogee, the Earth would have rotated an additional 240 degrees, or 480 degrees from the initial apogee over Kansas City, near Vienna, Austria. Sixteen hours later, at the third apogee, the Earth has rotated an additional 240 degrees, or 720 degrees total, and the satellite winds up back where it started, over Kansas City. Stabilization and Antenna Orientation The most important part of commanding a satellite is to keep the batteries charged. Without them, nothing else will work at all! Solar cells are the most practical method of charging batteries. AO-10 and AO-13 have 3 "arms" covered with solar cells. The position of the satellite in orbit must be consistent so as to insure efficient battery charging. Therefore, AO-10 and AO-13 are spin stabilized, where the satellite rotates about it's Z axis. The satellite is oriented so that the Z-axis is approximately perpendicular to a line to the Sun. As the satellite spins, each cell is evenly exposed to Sun and darkness. This rotating motion also insures that the satellite's internal temperature remains even. The spacecraft antennas are on the end of the bird, perpendicular to the Z-axis. Since the orientation of the spacecraft must be such that the solar cells receive adequate illumination, the antennas may or may not be actually pointing anywhere useful, like at the Earth. The ideal situation is with the antennas pointing at the center of the earth at apogee (nadir pointing), but this can only occur if the Sun is in the right position to adequately illuminate the solar cells. Part of the year it is, part it is not. Our tracking programs have "squint angle" calculations that tell us how far off the spacecraft antennas are pointing. This gives us a good idea of what to expect in the quality of the signals from the satellite; the larger the squint angle, the further off the main antenna pattern lobe we are and the weaker the signals become. P3D takes a completely different approach to stabilization and orientation, and consequently, to antenna pointing. Instead of the entire spacecraft spinning for stabilization, P3D has three internal "reaction wheels" mounted at right angles to each other. These are like large, variable speed gyroscopes, which, at steady state, impart a resistance to change. If, however, the speed of the rotating mass wheels is varied relative to each other, controlled angular motion is imparted to the spacecraft. This 3-dimensional control allows the orientation of the spacecraft itself to be precisely controlled. Additional control is available through magnetorquers, similar to those used on '10 and '13. P3D will also have ATOS, a mono-fuel arcjet motor. These stabilization and control systems allow P3D to always be oriented so that its antennas (on the end of the spacecraft) are pointed toward the Earth. Normally, the squint angle will be zero, optimum for communications. This freedom from spin allows the satellite to have a set of "wings" covered with solar cells, which are folded up around the spacecraft at launch and deployed when in orbit. These wings are oriented for most advantageous sun angle by rotating the entire spacecraft about its Z-axis. Internal temperature stabilization, which was automatic with spin stabilization, is handled by a heat pipe system. This whole package is controlled by the IHU (Internal Housekeeping Unit), the onboard computer. Communications on the Satellite P3D is a real flying antenna farm. It has "24 things sticking out of it". Directional antennas for all bands sprout from the +Z (main motor) end and side. Omni-directional antennas for 3 bands (146, 436 and 1269 MHz) are on the -Z (ATOS motor) end. Antenna design involved figuring out how to cram that many antennas on an 8.5 foot wide piece of real estate and getting them all to actually work! Each antenna has to have gain (which is beamwidth) consistent with the requirements of the orbit and path loss. When P3D is at apogee (47000 km), the Earth appears to be 13 degrees wide. The antennas were designed to have a beamwidth that provides signal strength consistent within 2 dB between the midpoint of the beam and the edges of the Earth. On the other hand, at perigee, the Earth appears to be about 68 degrees wide. The antenna pattern should cover this also. Path loss varies with distance. For example, on 1269 MHz, the path loss at apogee is about -185 dB. It perigee, it is - 165 dB. So there is a 20 dB improvement in path loss while the required beamwidth of the antennas changes from 13 degrees to 68 degrees. The pattern can therefore be 20 dB weaker (at the edge of the Earth) at perigee than at apogee to provide similar signal strength on the ground. Directional antennas consist of 3 folded dipoles on 146 MHz, 6 patch antennas on 436, a short backfire (a cavity antenna resembling a "dish", but not concave) on 1269 MHz and small dish antennas for 2400 and 5670 MHz. Horn antennas are used for the 10 GHz and 24 GHz downlinks. 29 MHz uses a 2-element beam (a driven element and a director) on the side of the craft. All (except HF) are right hand circular polarized. Internally, all the antennas are connected to individual transmit or receive converters, which all share a common Intermediate Frequency of 10.7 MHz. Each of the IF inputs and outputs are through a switch matrix that can connect any separate transmitter and receiver together as a crossband transponder. Theoretically, it would be possible to have a 24 GHz downlink with a 29 MHz uplink, but this is not feasible for other reasons. It is not possible to connect a transmitter and receiver in the same band. Each receiver (uplink) is connected to LEILA, the alligator killer. LEILA senses stations that are creating too strong an uplink and warns them of their transgression. If they do not reduce power, LEILA will put a deep notch right on top of the offender's signal, removing him from communications through the satellite. Hopefully, LEILA will not get much use. So much versatility in selection of up and downlinks has created a requirement for a new mode naming convention. On previous satellites, "Mode B" meant a 2 meter downlink and a 70 cm Uplink. That was easy enough, but when you have 36 different combinations, you run out of letters! So a 2 letter convention has been chosen, "Uplink/Downlink". Bands are still identified by their old conventional or radar designators. 146 MHz is called "V" for VHF. 436 MHz is "U" for UHF, 10 GHz is "X" and so on. Therefore, what we used to call "Mode B" will now be called "Mode U/V". Why Microwaves? All this brings up the old grumbling question of "why is everything going to microwaves"? There are several answers; all important. First, goodness of a communications channel is largely determined by the signal to noise ratio. To get a better comm link, you must either improve the signal strength or reduce the noise; preferably do both. Use of microwave frequencies allows narrow, concentrated beamwidths with physically small antennas. These narrow beamwidths allow efficient use of available transmitter energy; a clean narrow pattern concentrates energy that would be lost in broad, ragged pattern. And real estate is expensive and hard to come by, especially in space! P3D is quite large by amateur satellite standards, but it is still only 8.5 feet wide and must provide a platform for antennas for many different bands. Microwave links allow small antennas with efficient, predictable performance. Noise, the bottom half of the S/N equation, is lower on microwave bands. Noise comes in two forms, man-made and natural. Man-made noise from machinery and electrical equipment drops off greatly by 30 MHz. RF interference decreases as the frequencies increase both because there are fewer RF generating devices (not many 2.4 GHz and up handheld radios around) and those that exist also use gain antennas to concentrate their radiation where the owner wants it, rather than randomly as on HF. Natural noise comes in atmospheric and stellar varieties. We are all familiar with atmospheric noise. Lightning noise is crushing on 160 meters, not much on 10 meters and non-existent on 2 meters or above. Stellar noise is largely from electron motion and is measured in terms of temperature. The Earth is 300 degrees Kelvin; deep space on 2400 MHz is only a few degrees Kelvin. Cold and quiet. Microwave bands offer wide frequency allocations without a lot of competition. Users don't have to crawl over each other. Those familiar with digital communications know that the bandwidth required to send data is directly related to the data rate. If you combine the excellent signal to noise ratios available on microwave links with the broad bandwidths available, it is obvious that you have the makings for high speed, robust digital communications links. An additional advantage of having a ham radio presence on the microwave bands is described as "use it or loose it"! As technology progresses, commercial communications users are becoming more and more interested in the higher frequency bands and are fighting hard for all the band allocations they can get. And they have money. In the USA, we lost part of the 220 MHz band and share several bands with commercial and military interests, including radar on 400 MHz and "wireless" LAN's on 2.4 GHz. There is at present an assault on even 2 meters and 70 cm, the most popular VHF bands, for use with commercial "little LEO" satellites. The commercial people really like the results from the Microsats. An amateur presence on the microwave bands will help keep them for our use. Also, much of the microwave activity built into P3D involves downlinks on the highest frequencies. It is much more difficult and expensive to generate transmit power than it is to add a converter and preamp in front of a receiver. The P3D microwave downlinks will therefore be comparatively inexpensive for implementation in a groundstation.. Finally, this is cutting edge technology, and great fun! Up and Downlink Requirements AMSAT-DL has published several descriptions of P3D systems and statistics (in German; I have recently seen English translations by John Bubbers). Frank Sperber DL6DBN has published the calculations of up and downlink power and gain requirements for the various frequencies. These have been related to as close to common antenna configurations as possible, including short yagis for the bands up through 1269 MHz and the use of a 60cm (about 2 feet in diameter) parabolic dish antenna for the higher bands. The 60cm dish provides reasonable received signal strength on all microwave bands. Transmit power on all bands is no more than 10 watts with reasonable sized antennas. Of course, larger antennas may be used with lower power transmitters or less sensitive downlink setups. Omni-directional antennas may be useable up through 70cm. Problems, problems. With so many up and downlink combinations, how does one figure out what bands to operate and how to make everything work together? There are no firm plans on what will be the most used uplink/downlink combinations. A committee will be formed from the major contributors who will dictate schedules, consistent with technical requirements. While mode B is the current favorite mode (the only mode on AO-10 and the one with 92% of the time on AO-13), the 2 meter downlink is becoming almost unusable in much of the world due to RF congestion. Try rag chewing with a JA in in Tokyo on mode B. They are almost deaf from local interference. Big cities in Europe and the USA are almost as bad. Therefore, we will probably be seeing increased usage of the higher bands on P3D. Mode V/U (144 MHz up and 436 MHz down; the present mode J) will probably see a lot of use, although many operators have self-interference difficulties with a downlink harmonically related to the uplink. Mode L/S (1269 MHz up and 2401 MHz down) has been touted as the best of all worlds and should offer outstanding performance with inexpensive and easily attainable equipment and really small antennas. We can probably expect to see high percentage of time devoted to these combinations. The higher frequencies are definitely for the experimenters. Combinations using the X band downlink should be popular since there is a surprisingly large amount of terrestrial activity on 10 GHz, and equipment is available in both kit and assembled form. Amplifiers for uplinks on 2.4 GHz will be available, but will probably be in the 2 watt range, which would dictate use of larger antennas. There are no amplifiers commercially available to amateurs for 5.7 GHz, though there is reputed to be sporadic availability of surplus amplifiers. Down East Microwave is in process of having their 2.4, 5.6 and 10 GHz equipment (both kits and built-up) redesigned to make them more "user friendly". SSB Electronic (Germany) will have built-up equipment available for all bands through 10GHz. The 24 GHz downlink will probably see limited use since I am only aware of two European sources of mixers and preamps for that band, and they are definitely not plug and play. Groundstation Matrix Each individual will want to put together a groundstation that will best suit his operating habits. Just figuring out what the options are is confusing. One of the best ways to determine what makes sense for your own operating is to put together a matrix comparing modes to available bands. It is not necessary to put together every possible combination; some will never be implemented (such as a mode H/X or any up/down in the same band) and few of us will want to work them all. There are a few ground rules that make things fit together. This is the "what do you have and where do you want to go" part. First, figure out what equipment that you have or intend to obtain. The most common arrangements seem to be either a multi-band satellite rig (such as the Yeasu FT-736R or ICOM 970) with 2 meter and 70 cm modules or separate 2 meter and 70 cm transceivers, such as the ICOM 275/475. Either of these arrangements are useful for modes V/U or U/V.. The addition of a 23cm module to the multi-band rigs or the purchase of a separate 23cm transceiver (such as the ICOM 1271) will add L band. These arrangements can all operate full duplex, satellite crossband style, with the uplink in one band and the downlink in the other, but cannot go up and down in the same band. Either arrangement will allow operation on 6 separate modes (not counting HF), which should keep an operator busy. Many other arrangements are possible. For example, I have never owned a commercial radio that would tune above 30 MHz, but have relied on converters up through 10 GHz. Adding additional bands can be done most economically by sticking with available uplink bands. Additional downlink bands can be implemented by adding receive converters in front of your existing transceivers. Converters and preamps are cheaper than RF power amps. This is where a written matrix comes in handy. Converters can be obtained with various IF outputs. The most common IF's are 144 MHz and 432 MHz. In order to operate crossband full duplex, you must use one of your existing radios for uplink, and the other for downlink. Therefore, you want to chose converters that have IF's on the opposite band as the required uplink. For example, 10 GHz converters are available using either 144 or 432 MHz IF's. If you intend to work mode V/X, you would not want a converter with a "V" (144 MHz) IF, since you could not transmit and receive on 144 MHz at the same time. So a 10 GHz converter with a 432 MHz IF is the best choice. On the other hand, the modes that use S downlinks (U/S and L/S) have either 436 MHz or 1269 MHz uplinks, so a 144 MHz IF would be best. Operating both of the S up and downlinks will be challenging, since it will require two separate converters with different IF's. Typical P3D Groundstation Equipment Matrix Assume: AO-13 station with FT-736R (2 meter and 70cm) add: 1269 MHz transmit module 2400 MHz Receive Converter (144 MHz IF) 10 GHz Receive Converter (432 MHz IF) Up/Down 146 ( 146 ( 436 ( 436 1269 ( V/U ( ( V/X ( X( U/V ( ( U/S S( ( L/V ( ( L/U ( ( L/S S( ( L/X X( ( Microwaves pose some interesting problems that influence equipment decisions. As we have mentioned, transmit power is much more expensive than receive gear. A real problem is that coaxial cable losses become so high that it is necessary to use as little cable between the antenna and the active stage (whether transmit or receive) as possible. Cable losses also add directly to receive system noise figure. On AO-13 Mode S (2400 MHz downlink), the standard practice is to mount a low noise preamplifier directly at the antenna feedpoint. Many operators even mount the receive converter itself "up on the pole" near the antenna to keep line losses on 2400 MHz as low as possible. The trip to the shack is done at 144 MHz, where even RG-58 is satisfactory. This technique will be mandatory at higher frequencies. On the transmit side, there is no point in loosing all your expensive RF watts to line losses, so final amplifiers on the highest bands should be tower mounted. On 1269, this is convenient due to the availability of small "brick" linear amplifier modules that produce 10 to 18 watts RF out for a few milliwatts of RF in. Since 1269 uplink requirements are predicted to require 10 watts into a 12 turn helix, a very nice package could be made of a "brick" amp mounted on the back of a short helix and serving as its counterweight on the cross boom. A few milliwatts of RF at 1269 could be brought up from the shack in standard RG-213 cable. No hardline needed here. Mode S (2400 MHz) is unique in that it has both up and down links. One possibility is to mount a complete S band transverter with separate T/R inputs and outputs up on the tower in a weatherproof box, then mount a few watt amplifier behind a transmit antenna and a preamp and feed at a separate receive antenna. A pair of RG-58 cables could connect 144 MHz to and from the shack. Once again, cable expenses and losses are minimized. The Antenna Farm The attached figure of a proposed P3D groundstation antenna arrangement incorporates these ideas and more and crams complete 7 band operation into an antenna farm only 5 feet long, 7 feet wide (fits on a standard 6 foot fiberglass cross boom) and about 2 feet thick. Probably the most prominent characteristic of the system is the lack of a 2 meter antenna. The 2 meter antenna was left out for two reasons, the first being that they are physically large. Second, with the previously mentioned problems being experienced with 2 meter downlinks in Japan and Europe, 2 meters will not be a popular downlink. If, however, a 2 meter uplink is desired, the numbers indicate that a separate omni-directional antenna and 50 or so watts of transmit power will provide a satisfactory communications link. This is convenient, since RF power is common and inexpensive on 2 meters. The second characteristic is the presence of two separate parabolic dish antennas. One is the uplink and one is the downlink. Both dishes use clustered feeds (feeds for several different bands located next to each other). Clustering feeds works satisfactorily with the following cautions. First, put the highest frequency feed in the center, since it is the most critical to pointing and feed positioning (the "phase center" of the feed must be at the focal point of the dish). Second, several feeds will block and reduce the effective capture area of the dish. A larger dish may be required to compensate. Third, stick with "shallow" dishes (say, .5 to .6 f/d) so the curvature of the dish does not vary radically; it still must concentrate energy on the feedpoint and with multiple feeds, this is easier to do with the less precise focus offered by a shallow dish. Fourth, expect to off-point the dish slightly to get maximum signal to those feeds not directly in the center of the dish. Since all transmitting is done with one dish and all receiving with the other, it is not necessary to protect delicate receive preamps from the RF energy of transmitters. When using a single dish for both transmit and receive, it is necessary to use some sort of coaxial relay arrangement to disconnect the inputs of any receive systems from their dish feeds to keep from frying the preamp front ends. This need to protect receive front ends means that full duplex operation will not be possible with a single dish doing double duty. Coaxial relays are expensive and heavy. And any mechanical device on the antenna is just something else to fail at the least opportune times. Also, spacing the transmit and receive dishes physically apart reduces receiver desense experienced when a transmitter is operating in close proximity to a sensitive receiver. The two dish arrangement offers several other advantages. Switching is simplified, especially on S band, where separate dishes and separate feeds fed from separate ports on a transverter completely eliminate relays and switching. I have seen some proposed dish feed arrangements which mount 1269 MHz (for uplink) and 2400 MHz (for downlink) helix feeds concentrically. This looks good at first, but will prove impractical, since it would require expensive relays for preamp protection and still not allow full duplex crossband operation on mode L/S. A much better arrangement is to use a separate L band uplink helix antenna (as previously described) and put the S band downlink in a small dish. Parabolic Dish Antennas Parabolic dish antennas come in many forms. The size-independent parameter that characterizes any dish is the f/d ratio. This ratio gives the relationship between the focal length, f (where the "phase center" of the dish feed is positioned) and the dish diameter, d. Most dishes have an f/d between .3 (called a "deep" dish) and .8 ("shallow"). The characteristics of the feed must be matched to the f/d ratio, in that the feed must have a beamwidth that illuminates the dish from edge to edge, typically at the -10 dB points. If the beamwidth is too narrow, not all the area of the dish is utilized. If the beamwidth is too wide, on receive the feed will "see" the warm Earth behind the dish, receiving excessive noise along with the signal. Many references exist on the subject of matching dish feed to dish geometry. Note that f/d ratio is just that, a ratio, and is not related to the actual physical size of the dish. The distance of the phase center of the feed from the dish is calculated from the f/d and the actual dish diameter. In practice, this is usually optimized by peaking on Sun noise. Dish construction varies from solid metal forms to almost invisible wide open mesh designs. The advantage of open mesh is that it has much less wind load than an equivalent solid dish. The "quality" of the dish, it's trueness to the actual parabolic form and the allowable spacing between wires in open mesh dishes, is determined by the frequencies over which it will be used. A rule of thumb, from optical practice, is that there is no measurable difference between a 100% accurate dish form and one where the maximum deviation from theoretical is less than 1/10 wavelength at the highest frequency of use. This also applies to allowable mesh spacing in open mesh dishes. Remember that this is not a hard and fast rule. Greater variations will only result in a progressive deterioration of performance (reduced efficiency) on form or increased noise level and reduced efficiency on mesh size. This rule indicates that a dish which will be used at or below 2400 MHz (13cm) need only be accurate within 1/2 inch of true form. I have used small, homemade, stressed construction dishes covered with 1/2 inch mesh "hardware cloth" for years on mode S and experienced quite satisfactory performance. For 5670 MHz, the form and mesh needs to be .21 inch or less. On 24 GHz, this must be .06 inch. This accuracy will probably dictate use of solid commercial dishes. Additional Thoughts One possible method of reducing cabling and line losses is to mount certain equipment up on the antenna tower. We have already mentioned the advantage to mounting an entire S band transceiver and transmit amplifier on the tower (with the receive preamp mounted on the dish feed). Two other good candidates for tower mounting would be a 5668 MHz transmit converter and amplifier and a 15-20 watt L band (1269 MHz) "brick" amplifier. If the 5668 converter uses a 1269 MHz IF, a single length of RG-213 coax can be used to supply low level RF from the shack-mounted 1269 MHz exciter to either the amplifier or converter, switched by means of a small RF relay. The milliwatt power levels involved may even allow switching with an inexpensive TO-5 can type RF relay. Of course, 10 GHz and 24 GHz receive converters must be tower mounted, most preferably directly at the dish feedpoints. To keep the weight at the feedpoints down, it may be desireable to mount only the RF stage and mixer at the dish feed, with the local oscillator mounted remotely. If the LO is mounted behind the dish antenna, it will help serve as a counterweight. A short run of UT-141 copper hardline could be used to connect the LO and the mixer. There is lots of room for creativity here. An extremely important consideration is mechanical stability and ruggedness of the antenna system. OSCAR antennas can become large and complex, which means they are heavy and present large wind loads to the rotors. It is poor practice to rely on the elevation rotor and the strength of boom clamps to support off-center loads applied to the cross boom. Parabolic dish antennas and their feed arrangements can be quite heavy and are extended away from the cross boom. Even mesh dishes can have large wind loads. Always counterbalance your antennas in elevation and equalize loads in azimuth directions. Small dish antennas frequently use a non-metallic center support to which the dish and feed are attached; this also serves as the attachment point to the cross boom. Extending this support behind the cross boom gives an ideal attachment point for microwave transmit amplifiers, local oscillators and other devices which are necessary to operation. At the same time, these make very practical counterweights. Also, attempt to equalize wind load on either side of the azimuth rotor. The double dish arrangement (with one on each side of the azimuth rotor) accomplishes this. That layout also has the balance of a short yagi on one end of the cross boom and a helix (with it's wind catching mesh reflector plate) on the other end. When not in use, the entire antenna should be rotated for minimum or equalized wind load, which (for dish antennas) may be at 90 degrees elevation, cross boom parallel to the prevailing winds. This orientation is sometimes referred to as the "birdbath" position. Under snow and ice conditions, dish antennas should be pointing to the horizon, as this will prevent them from gaining a lot of weight. And so... This paper was written for two reasons; first, to explain how P3D's physical and orbital characteristics relate to the groundstation user and, second, to suggest some possible configurations for a multi-purpose groundstation. The concentration is on antennas and intermediate frequencies, since those take the most "juggling" to coordinate. None of this is the last word on the subject, but maybe will give users a place to start in assembling versatile groundstations for Phase 3D.