Ku-Band Satellite Payload

Academic project, coursework inside the SYSCom track at ENSEEIHT. No simulated curve, computed budget or measurement is published yet.

A transparent payload: the satellite receives an uplink band, converts it, amplifies it and retransmits it on the downlink, without demodulating what it carries. That makes the payload an RF chain whose imperfections land directly on the user's signal.

• Input sectionThe 17.3 to 18.1 GHz uplink band, selected by filtering and amplified where the noise figure of the whole payload is decided.
• Frequency conversionMixing down towards around 12.1 GHz, with the local oscillator and the image and spurious products that come with it.
• FilteringBand selection before conversion, image rejection around it, and output filtering before the high-power stage.
• Automatic gain controlHolding the drive into the output amplifier where it is wanted, as the input level varies.
• Output amplifierA travelling-wave tube amplifier, which is efficient near saturation and nonlinear exactly there.

The engineering question of the project. A TWTA is most power-efficient close to saturation, and close to saturation it distorts. Backing the drive off restores linearity and costs output power, which costs link margin. The design point is a trade, not an optimum, and the point of the study is to see the trade end to end: from a compression curve to a constellation.

• Saturation and compressionThe AM/AM and AM/PM behaviour of the tube: amplitude compression, and the phase shift that comes with drive level.
• Input and output back-offIBO and OBO as the control knob, and what each dB of back-off buys in linearity and costs in radiated power.
• QPSK degradationHow the operating point shows up in the received constellation: warping, phase rotation, spectral regrowth into the adjacent channel, and the resulting error rate.
• End to endThe relationship the project is built to expose: an RF-level quantity, the amplifier operating point, expressed as a communication-quality quantity.

Back-off measured from the saturation point at each side of the amplifier. The two are not equal, because compression means the output falls more slowly than the input: that difference is the whole reason the trade is interesting.