National Astronomy and Ionosphere Center Research and Development Laboratory 124 Maple Ave Ithaca, NY TECHNICAL REPORT

Transcription

1 National Astronomy and Ionosphere Center Research and Development Laboratory 124 Maple Ave Ithaca, NY TECHNICAL REPORT April 21, 1999 To: NAIC Staff From: Eugene Lauria Subj: Trap issue in reference to the L-band wideband receiver. This report is an update in reference to the trap issue brought to our attention from a memo presented by Riccardo Giovanelli, in January of this year. As pointed out in the memo, several tapped mode resonances ( traps ) have been observed with the L-band receiver. It is understood that this condition is unacceptable, and a solution has to be obtained. However, before a solution can be determined, some study and work had to go into understanding the mechanisms that cause the traps and what would be the easiest and most sensible way to remove them. There has been some progress in the understanding of the cause of these trapped modes, and what can be done to eliminate them. The main component that is thought to cause the trapped modes is the taper transition between the orthomode transducer (OMT) and the feedhorn. A major disadvantage in the design of our wideband L & S-band receiver systems is the very large diameter of the circular waveguide input to the OMT. Typically, the feedhorn diameter is at least or larger than the diameter of the input to the OMT of the receiver system. However, in our wideband receiver systems, the feedhorn is smaller in diameter than that of the mouth of the OMT. This requires a taper section that reduces in diameter between the OMT and the feedhorn. It is believed that due to the discontinuity from the probes in the back of the OMT (which convert the waveguide mode into a coaxial mode) and the fin structure, unwanted modes are produced and either get trapped in the fin structure or in the taper section. If the modes do get trapped in the fin structure, there is little if anything that can be done to eliminate them. This is an inherent problem with quad-ridge OMT s of which type this OMT is. Typically, these devices have three to four of these trapped modes; their number depending on the bandwidth of the device. The broader the bandwidth, the higher the likelihood that more traps will be present. Once they are present, they are

2 difficult if not impossible to remove. Typically, the way to try to remove them is to adjust the probe structure since they are responsible for launching the modes. However, this typically does not alleviate the problem. If a trap is located at a particular frequency of interest, the fin structure can be modified so that the trap is tuned to another frequency. The most troublesome modes that are believed to be causing so much trouble are the TE 21L [1] & the TM 11 modes. In the quad-ridge section of the OMT, the TE 21 mode degenerates to an upper TE 21U mode and a lower TE 21L mode [1]. The TE 21L mode is particularly troublesome. For fin heights that approach the radius of the circular waveguide that they are in, the cutoff frequency of the TE 21L mode is almost the same as the fundamental TE 11 mode. The TM 11 is known to be a very troublesome mode as well. In corresponding with Mal Sinclair and Bruce Thomas of the CSIRO (who manufactured the OMT) the TM 11 mode is known to cause a terrible cross polarization problem which causes a severe degradation in the beam of the antenna with their feeds. Luckily for us, we are using a different type of feedhorn and this problem has not been observed here in Arecibo. However, the TM 11 mode is also highly likely to cause trapping phenomena within the OMT and taper. In our particular case, we need to determine whether or not the numerous traps are generated by the taper transition. Several measurements were taken using a network analyzer with the OMT by itself and with the different taper sections attached to it. In all of the measurements, the OMT (or OMT/taper combination) was radiating into freespace with a piece of eccosorb placed in front of it. Reflection measurements were taken at the coaxial ports of the OMT as well as transmission measurements between the two ports to measure the isolation between the polarizations. Both types of measurements are useful in detecting the trapped resonances because they will show up as dips in the return loss and spikes in the isolation between the two polarizations. That is, at the frequency where the trap occurs, energy gets coupled to the orthogonal polarization and shows up as a resonant spike. In calibrating the network analyzer, a full 2-port calibration from 1 to 1.75 GHz taking 1601 data points was done. This number of data points gives a resolution of khz which ought to be adequate to detect the resonances. The OMT was placed on a table and its mouth was terminated into freespace with a piece of eccosorb in front of it. The horizontal coaxial port was connected to port 2 of the network analyzer, and the vertical polarization was connected to port 1. Four measurements were taken with the OMT. The first measurement was taken with the OMT alone. The second measurement was with a taper section that went from 9.21 inches (the diameter of the mouth of the OMT), to 8.33 inches. The third measurement taken was with the 9.21 to inch taper section. (This taper section and accompanying feed is currently being used with the wideband L-band receiver on the telescope.) In each of the measurements, a piece of eccosorb was placed in front of but not onto the mouth of the OMT or the OMT taper/feedhorn combination. The following table was produced cataloging all of the traps found with the different OMT/taper combinations.

3 F (GHz) OMT Only 8.33 dia taper dia taper x x x x x x x x x x x x Total: Another (the fourth) measurement was performed by attaching the feedhorn onto the diameter transition. However, attaching the feedhorn had no effect in adding, subtracting, or in the tuning of the traps. Several observations can be derived from the preceding table. First of all, it is important to note at what frequencies the TE 21 and TM 11 modes are cutoff at the mouth of the 9.21 diameter OMT, and what frequencies can propagate through the smaller opening of the taper section for these modes. The following table lists the cutoff frequencies for these modes for the diameter of the OMT and the two taper sections: Diameter (inches) fc TE 21 (GHz) fc TM 11 (GHz) 9.21 (OMT) The first observation to be made is that the 4 traps marked by an x were recorded in all of the measurements. These traps are restricted to the fin structure, and cannot propagate through the mouth of the OMT. As a result, these traps are unaffected by the taper section or the feedhorn. Note that the cutoff frequency for the TE 21 mode is at GHz for the mouth of the OMT. This implies that if the TE 21 mode will not propagate past the fin structure or down the taper section below this frequency. However, there are several traps that do

4 occur above this frequency, and more importantly, shift to a higher frequency when the inch diameter taper is used. This is an important observation because it does suggest that the taper section does create these traps. This is because as the diameter reduces in size the diameter of the taper reaches the cutoff frequency and the mode becomes trapped in the OMT/taper section. The upward shift in frequency of these traps by the inch diameter taper occurs because the cavity produced by the OMT and taper is shorter in length than that of the cavity produced by the 8.33 inch diameter taper. Since both taper sections are the same in length, the angle of the inch diameter taper is steeper than the 8.33 inch diameter taper thus, shortening the length of the cavity producing the trap for the smaller diameter taper section. Another important observation to make is that the cutoff frequency for the TE 21 mode is at GHz for the 8.33 inch diameter taper, and at 1.44 GHz for the inch diameter taper. Notice that there is an additional trap at GHz for the inch diameter taper but not in the 8.33 inch diameter taper. This suggests that above these frequencies, the TE 21 can propagate past the opening of the taper section and no traps should occur. Therefore, the traps above these frequencies are most likely not from the TE 21 mode. A closer look at the data suggests that these higher frequency traps (except the one at GHz) may be from the TM 11 mode. The cutoff frequency for the TM 11 mode for the mouth of the OMT is at 1.56 GHz. Also, if these traps were generated by the TM 11 mode, they would all be caught within the taper section because the cutoff frequencies for the 8.33 inch taper and the inch taper for the TM 11 mode are at GHz and GHz, respectively. Since all of the traps measured are below these frequencies, the are indeed trapped inside the taper for the TM 11 mode. The only unfortunate result from performing the measurements with the OMT in the lab is that none of the traps that were observed in the lab corresponded with the traps that have been observed with the OMT being used with the receiver system on the telescope. The reason may be because the two OMT s are not mechanically identical. It is very difficult to make two of these devices exactly alike (especially for the probe section). For example, we have two S-band OMT s with the same model number. However, there are enough mechanical differences between them so that they cannot be interchanged with the existing dewar package. These mechanical differences can cause differences in the electrical behavior of the device. This doesn t mean that one OMT may produce fewer or more traps, (because the electrical design is the same), but the traps can occur at different frequencies with respect to one another. In conclusion, it looks quite hopeful that removing the downward taper and putting a straight section of circular waveguide between the OMT and feedhorn should remove many of the traps that we are observing with the current configuration. Currently, a new section of circular waveguide is being fabricated in Ithaca, and it will be tested with the OMT in the lab. If this works out, we will attempt to redesign the feedhorn so that it matches the diameter of the OMT, and also has higher gain to reduce the spillover on the tertiary. Results from the straight section will be presented in a later report once it has been fabricated.

5 Reference: [1] M.H. Chen, G.N. Tsandoulas, and F.G. Willwerth, Modal Characteristics of Quadruple-Ridged Circular and Square Waveguides, IEEE Transactions on Microwave Theory and Techniques, August, 1974, pp