As a preliminary to describing radio telescopes, it is useful to have a look at the transparency of the atmosphere to electro-magnetic waves of different frequencies. Figure 3.1 is a plot of the height in the atmosphere at which the radiation is attenuated by a factor of 2 as a function of frequency. There are only two bands at which radiation from outerspace can reach the surface of the Earth, one from - the optical/near-infrared window, and one from a few mm to tens of meters - the radio window. Radio waves longer than a few tens of meters get absorbed in the ionosphere, and those shorter than a few mm are strongly absorbed by water vapor. Since mm wave absorption is such a strong function of the amount of water vapour in the air, mm wave telescopes are usually located on high mountains in desert regions.
The optical window extends about a factor of in wavelength, whereas the radio window extends almost a factor of in wavelength. Hence while all optical telescopes `look similar', radio telescopes at long wavelengths have little resemblance to radio telescopes at short wavelengths. At long wavelengths, radio telescopes usually consist of arrays of resonant structures, like dipole or helical antennas (Figure 3.2). At short wavelengths reflecting telescopes (usually parabolic antennas, which focus incoming energy on to the focus, where it is absorbed by a small feed antenna) are used (Figure 3.3).
Apart from this difference in morphology of antennas, the principal difference between radio and optical telescopes is the use of coherent (i.e. with the preservation of phase information) amplifiers in radio astronomy. The block diagram for a typical single dish radio astronomy telescope is shown in Figure 3.4. Radio waves from the distant cosmic source impinge on the antenna and create a fluctuating voltage at the antenna terminals. This voltage varies at the same frequency as the cosmic electro-magnetic wave, referred to as the Radio Frequency (RF). The voltage is first amplified by the front-end (or Radio Frequency) amplifier. The signal is weakest here, and hence it is very important that the amplifier introduce as little noise as possible. Front end amplifiers hence usually use low noise solid state devices, High Electron Mobility Transistors (HEMTs), often cooled to liquid helium temperatures.
After amplification, the signal is passed into a mixer. A mixer is a device that changes the frequency of the input signal. Mixers have two inputs, one for the signal whose frequency is to be changed (the RF signal in this case), the other input is usually a pure sine wave generated by a tunable signal generator, the Local Oscillator (LO). The output of the mixer is at the beat frequency of the radio frequency and the local oscillator frequency. So after mixing, the signal is now at a different (and usually lower) frequency than the RF, this frequency is called the Intermediate Frequency (IF). The main reason for mixing (also called heterodyning) is that though most radio telescopes operate at a wide range of radio frequencies, the processing required at each of these frequencies is identical. The economical solution is to convert each of these incoming radio frequencies to a standard IF and then to use the exact same back-end equipment for all possible RF frequencies that the telescope accepts. In telescopes that use co-axial cables to transport the signal across long distances, the IF frequency is also usually chosen so as to minimize transmission loss in the cable. Sometimes there is more than one mixer along the signal path, creating a series of IF frequencies, one of which is optimum for signal transport, another which is optimum for amplification etc. This is called a `super-heterodyne' system. For example, the GMRT (see Chapter 21) accepts radio waves in six bands from 50 MHz to 1.4 GHz and has IFs at 130 MHz, 175 MHz and 70 MHz3.1.
After conversion to IF, the signal is once again amplified (by the IF amplifier), and then mixed to a frequency range near 0 Hz (the Base Band (BB) and then fed into the backend for further specialized processing. What backend is used depends on the nature of the observations. If what you want to measure is simply the total power that the telescope receives then the backend could be a simple square law detector followed by an integrator. (Remember the signal is a voltage that is proportional to amplitude of the electric field of the incoming wave, and since the power in the wave goes like the square of its amplitude, the square of the voltage is a measure of the strength of the cosmic source). The integrator merely averages the signal to improve the signal to noise ratio. For spectral line observations the signal is passed into a spectrometer instead of a broad band detector. For pulsar observations the signal is passed into special purpose `pulsar machines'. Spectrometers (usually implemented as ``correlators'') and pulsar machines are fairly complex and will not be discussed further here (see instead Chapters 8 and 17 more more details). The rest of this chapter discusses only the first part of this block diagram, viz. the antenna itself.