Msg # 2089 Type:B Stat:N To: ALL From: ON7KL Date: 29-Mar/1642 Subject: RUDAK EXP INFORMATION PART 2 Path: PI8HWB!ON7RC The Digital Transponder of the RUDAK Experiment by Hanspeter Kuhlen, DK1YQ, AMSAT/DL 8011 Aschheim by Munich AMSAT-DL Journal, Nr.2/15 - March/April 1988 (Translated by Don Moe, DJ0HC/KE6MN) Overview The new amateur radio satellite Oscar Phase 3C, now scheduled for launch into an elliptical orbit by the first Ariane 4 rocket in early May 1988, will be carrying the largest number of dif- ferent transponders ever flown on an amateur radio satellite. In addition to the two linear transponders for Mode B (70cm=>2m) and Mode L (24cm=>70cm), meanwhile widely known for their func- tion aboard Oscar 10, there are additional transponders avail- able for Mode J (2m=>70cm), Mode S (20cm=>13cm), and last but not least a regenerating transponder for digital amateur radio communications (RUDAK). The digital transponder "RUDAK" is basically an integral compo- nent of the Mode L transponder. In order to present the func- tional interrelations of the various transponders in a clearer fashion, the entire L transponder will be presented first. In the second part, I will go into the special RUDAK modules in greater depth. The Mode L Transponder The Mode L transponder linearly translates a 250 kHz wide fre- quency segment of the 24cm band into an equally wide segment of the 70cm band. All signals in the uplink passband are merely in- verted in frequency and retransmitted as received. "Frequency inverted" means that a signal sent in the lower end of the up- link band reappears at the upper end of the downlink band. At the same time, LSB signals from the ground stations become USB signals on the downlink. In contrast to Oscar 10, the transponder bandwidth is limited to 250 kHz, compared to 800 kHz. Even though technical problems on Oscar 10 prevented Mode L operation in the manner expected and desired, the chosen bandwidth appeared to be too large by a fac- tor of 2 or 3. Therefore the decision was made to opt for the narrower 250 kHz on P3C. The block diagram in figure 1 shows the structure of the entire transponder as installed in Oscar P3C. March 18, 1988 RUDAK Transponder Page 1 The receiver operates according to the dual conversion princi- ple, resulting in an intermediate frequency of 10.7 MHz after amplification and filtering at the input. The incoming signals are processed at this frequency by a limiting amplifier and sub- sequent radar pulse noise blanker before being made available to several different receivers. The limiting range of the IF ampli- fier and thus the entire receiver is approximately 30 dB. The experience gained with Mode B on Oscar 10 demonstrates the need for this high value, since on good days gain reduction of virtu- ally 20 dB was frequently necessary. This implies nothing more than that several users transmitted on these occasions at more than 100 times the normally sufficient effective radiated power level. Due to the automatic gain con- trol in the satellite's receiver, they thereby made the satel- lite simply too insensitive for QRP stations. Under such condi- tions, every receiver reacts with emergency measures in order to prevent distortion and intermodulation by-products, which are unavoidable at such excessive signal levels. Following the radar pulse blanker, the receiver branches into three completely independent receive legs. One leads to the com- mand receiver, which provides the data link to the onboard com- puter. A second carries the passband to the transmitter, which should effectively reradiate the entire 250 kHz HF bandwidth. The third leg leads to the RUDAK receiver. Initially let's pursue the second branch a bit further. The passband is converted up to an intermediate frequency of around 53 MHz. A mixture incorporating all signals destined for the downlink is prepared at this IF level. Among them are the so- called general beacon, which transmits the telemetry information from the onboard computer (IHU - integrated housekeeping unit) to the command stations, and the RUDAK beacon, which transmits the packet information from the RUDAK processor. Prior to injec- tion, both beacons are modulated at a bit rate of 400bps in BPSK using biphase-S encoding. The bandwidth of the beacon signal is limited by filtering to around 600 Hz. In the case of the RUDAK beacon, an NRZ-I encoded bit rate of 1200 bps can be transmitted alternatively. This signal is then compatible with the downlink signal from the Japanese packet ra- dio satellite, JAS-1/Fuji Oscar 12. Currently available 1200 bps BPSK demodulators can thus be used for both satellites. The frequencies of the beacons are situated just above (RUDAK) or below (GB) the passband and thus don't occupy any of the ana- log passband and can remain in operation, even when passband op- eration is disabled. The output power is 1 watt for the general beacon and 2 watts for the RUDAK beacon. Based on requests from a large number of users, Mode J was reintroduced, since in many regions on earth, such as Japan, the 2m band has become virtually unusable for downlink signals be- March 18, 1988 RUDAK Transponder Page 2 cause of strong interference from FM stations. For this option, an approximately 50 kHz wide segment of the 2m band is trans- lated to the 70cm passband. Prior to being radiated by either the 9dBi high-gain or the omnidirectional antennas, this signal mixture is amplified at high efficiency to a power level of approximately 50 watts PEP using the HELAPS technique especially developed by Dr. Karl Meinzer, DJ4ZC. As can be seen in the block diagram in figure 1, the signal path is separated into phase and amplitude portions. The phase por- tion is efficiently amplified in saturated amplifier stages, whereas the demodulated amplitude modulation drives the final stage. Depending on transponder usage, the overall efficiency of the linear amplifier can be improved by 5 to 10 percent compared to the normal "straight-ahead" method. The Digital RUDAK Transponder Now let's follow the IF signal path to the digital transponder section. An unusual feature is the controlled mixer oscillator at 19.7 MHz. All signals within approximately +/-7.5 kHz of the nominal center frequency are converted to 9Mhz before the crys- tal filter. A saw-tooth generator produces a search voltage which practically steers the receiver back and forth across this range. The receiver scans this range for uplink signals. Following the filter, another oscillator translates everything down to a frequency of 24 kHz for the coherent demodulator. Coherent means essentially "phase-exact". For coherent demodu- lation, the non-existent carrier has to be regenerated exclu- sively from the received signal. The fact is used that squaring a PSK modulated signal retains the carrier and loses the modu- lation. A portion of the signal is therefore extracted, squared, and fed to a phase comparator, which gets its reference from a voltage controlled oscillator running at the expected frequency. The incoming signal and the local VCO can then be synchronized through a filter. The local VCO remains phase-locked with the input signal. Mixing this VCO signal with the input signal re- sults in only the modulation, in our case the 2400 bps signal. Everything is known to be relative. The process just described takes place in this demodulator, slightly modified. We already saw that the receiver is set up as a scanner. Why can't this circuit also be used to regenerate the carrier? As shown in fig- ure 2, the oscillator in the first control loop is fixed and the receiver tuned. In other words, the local oscillator doesn't follow the receiver, rather the receiver follows the L.O. At a scan time of approximately 120msec, the receiver scans across the 15 kHz wide capture range. (zie vervolg deel 2 - pa0hwb) PI8DZI BBS (B,D,H,I,J,K,L,N,R,S,T,U,V,W,X,Y,?) >