Fig. 3.2. A Dicke microwave radiometer. From the left E. Beringer, R. Kyhl, A. Vance, and R. Dicke. Dicke is holding in front of the horn antenna a "shaggy dog," a good approximation to a source of blackbody radiation at room temperature. (From Five Years at the Radiation Laboratory, MIT, 1946.)

directions in the sky from which the radiation is received. The first two27 direct detections, in experiments at the Bell Telephone Laboratories and then at Princeton University, had a feature in common: both used antennas shaped like a horn or funnel. A waveguide at the small end of the horn leads to a detector. The early example in Figure 3.2 shows the horn in the Dicke radiometer described by Dicke et al. (1946). A horn antenna can suppress radiation coming from directions well away from the main beam. That is important because the ground is a strong source of microwave radiation, and the system must be well shielded from this unwanted ground noise.28 Hogg describes two horns used in the Bell communications experiments that detected the CMBR; these horns appear in Figures 4.1 and 4.2. Figures 4.22 and 4.23 shows the Princeton horn. The sizes are very different, but that is not important because the received energy flux from an isotropic sea of radiation does not depend on the horn size.29

27 Other histories of this subject mention earlier experiments that may have detected the CMBR. We note on page 63 that we have not been able to substantiate any.

28 Since the properties of horn antennas are important for the study of the CMBR we add here a few details. An example of the suppression of unwanted radiation incident from directions well away from the main beam of the horn is shown in Figure 4.13. The horns in Figure 3.2 and in the subsequent detections of the CMBR by the Bell Laboratories and Princeton groups have rectangular collecting areas. A reflector at the end of the Bell Labs horn brings radiation to a horizontal waveguide from a beam that may be swung across the sky. In communications applications using this Bel Labs design the waveguide is vertical and the beam horizontal. Burke (p. 179) compares this design to a scoop. Modern horns for measurements of the CMBR tend to have circular cross sections, as in a funnel.

29 The solid angle of the antenna beam — the angular area in the sky from which radiation is directed to the detector — is inversely proportional to the collecting area of the antenna. The product of the solid angle and the collecting area determines the rate of collection of radiation energy from a source that is broader than the solid angle. This product is independent of

Fig. 3.2. A Dicke microwave radiometer. From the left E. Beringer, R. Kyhl, A. Vance, and R. Dicke. Dicke is holding in front of the horn antenna a "shaggy dog," a good approximation to a source of blackbody radiation at room temperature. (From Five Years at the Radiation Laboratory, MIT, 1946.)

Unwanted radiation - noise - also originates in the receiver. The Bell Laboratories experiments used low-noise solid-state maser amplifiers developed there for the purpose of communication. The Princeton experiment used a detector with much larger noise, and they used a technique Robert Dicke pioneered to deal with it: rapidly switch the receiver between the antenna and a reference "load" that produces thermal radiation at a known temperature. The difference of the detector response subtracts the radiation originating in the detector and amplifier. The time average beats down the fluctuations in the difference. The Dicke radiometer thus yields a measurement of the difference Ts — Tl between the wanted sky temperature Ts and the known load temperature Tl.

The low noise of the Bell Laboratories' receivers allowed the engineers to make useful estimates of the noise originating in the maser amplifier and add it to estimates of the noise from all other known sources of radiation: the atmosphere, the ground, the antenna, and the waveguide leading from the antenna to the amplifier. That sum could be compared to what was detected. A persistent discrepancy between what was detected and expected led Penzias and Wilson to take the final step: use a cold reference load to check the amount of radiation originating in the system (that adds to what is incident on the antenna). The results forced them to conclude that there is a source of radiation outside the system that they could not identify. Hogg, Penzias, and Wilson recall these events in the next chapter.

As this was happening Roll and Wilkinson, just 30 miles away in Princeton, were building an instrument to search for a possible sea of microwave radiation. They recall in the next chapter that Dicke had proposed this project to test the idea of a hot big bang; we reviewed his thoughts in Section 3.3. Roll and Wilkinson also used a cold load. The difference was that their instrument rapidly switched between sky and cold load, so as to compensate for the drifting level of the relatively large noise originating in the detector. Penzias and Wilson needed only to switch occasionally because their system noise was much lower.

Dicke had invented the instrument Roll and Wilkinson were building, a Dicke radiometer, as part of war research at the Radiation Laboratory at the the horn size. The small horn in Figure 3.2 has poor angular resolution: Figure 3.3 shows the broad response of this radiometer to the relatively narrow warm objects. A larger collecting area gives better angular resolution, that is, sensitivity to compact sources. Figure 4.9 shows a dish reflector antenna, where incident radio or microwave radiation is directed by one or more reflecting surfaces — as in a dish — to a much smaller feed horn antenna leading to the detector. The size of a primary reflector can be made larger than the size of a horn, improving angular resolution. But that sometimes comes at the expense of poorer rejection of radiation incident from directions well away from the source, including ground noise, which can be a serious problem for observations of the CMBR.

Massachusetts Institute of Technology. An early application was the measurement of the emission - and hence absorption - of microwave radiation by the atmosphere, which at the time limited the push to develop radar at shorter wavelengths for better resolution.30

Figure 3.3 is Dicke's illustration31 of the sensitivity of his radiometer. In this example, the reference load was at room temperature, and the switching was done by a wheel that swung the load into and out of the waveguide connecting the detector and the antenna. Later measurements used electronic switching and loads with temperatures that are colder and more closely matched to the CMBR temperature. A Dicke radiometer "sees" thermal microwave radiation wherever the horn is pointed, whether at the ground, or people, or the atmosphere. The strip-chart recording in Figure 3.3 shows the variation in the response when the antenna was pointed at the sky and at chimneys that were in use and so slightly warmer than their surroundings.

The top line in Figure 3.3, measured with the antenna scanning at an angle of 75° from the zenith, indicates a more or less uniform temperature of about 125 K. Variations in the temperature from one part of the sky to the other are small, less than about 10 K. This means the instrument used at MIT was capable of detecting temperatures as small as 10 K. Note also that the temperature measured well away from hot chimneys increases as the angle from the zenith is increased from 75° to 90°. Some of the increase in detected temperature at the larger zenith angle is the result of larger microwave emission by the Earth's atmosphere. This effect is the basis for a measurement of the radiation emitted by the atmosphere, as follows.

When the instrument is aimed closer to the horizon, it looks through a longer path through the atmosphere. The longer path length means the atmosphere produces more radiation along the line of sight.32 By measuring how the temperature of the received radiation varies with the angular distance from the zenith, or with distance through the atmosphere, one

30 Since microwave emission from the atmosphere is another important part of the story it is worth remarking here that if material at a nonzero temperature absorbs radiation then it also emits radiation. It is the balance of absorption and emission that produces blackbody radiation. Atmospheric absorption of radiation means that ground-based measurements of the microwave background have to deal with radiation produced by the atmosphere. Wilson, on page 163, and Wilkinson, on page 203, emphasize another point to bear in mind. The standard technique used by radio astronomers to measure the radiation received from an object outside the atmosphere is to compare the energy flux received when the detector beam is on the source to the flux received when the beam is directed to a point in the sky slightly off the source. The subtraction eliminates a good deal of the noise from the atmosphere as well as from the ground and detector. But this technique does not work for observations of the temperature of CMBR because it is uniformly distributed across the sky.

31 The data were taken by Dicke in the summer of 1945. A redrawn version of this figure is in Lawson and Uhlenbeck (1950).

32 In the approximation of the atmosphere as a plane-parallel slab of emitting material the detected atmospheric emission varies with the secant of the zenith angle.

0 0

Post a comment