10. Recommendations
On the basis of the investigations carried out, there are three areas which should be pursued further.
Firstly, it was evident that the quality of the results obtained for a metal post in a rectangular waveguide depended heavily on the quality of the waveguide test jig. In the measurement, the function 'port extension' was used extensively to shift the measurement plane. As it had been explained, port extension only changed the phase of the scattering parameters. The assumption taken was that over the distance 'extended', that section of the waveguide (test jig) was assumed to be perfect. There was no imperfection in the way that could change the phase locally. In practice, this assumption was only true to a certain extent. One reasonable way to improve the results can be achieved by reducing the length of the waveguide test jig. The shorter the length, the less significant the assumption is. Better still, should the test jig be redesigned, it would be excellent if the test jig could have provisions so that calibration to the plane of the post was possible.
On the other hand, the existing test jigs were having too many 'gaps' - they were basically held in place by screwing many flaps together. The electrical contacts were poor. Although not being able to be seen by eyes, at microwave frequencies the gaps could be significantly large. The new design can consider to have more parts that are cast out.
Second, concerning the program NWAMODEL, experience had shown that there were occasions when the program 'went out of sequence'. Essentially, what happened was that if the user did not use the program in the way that he/she was instructed as in the message box, chances were that the program would go out of sequence and that the NWA would interpret those incorrect steps as error. The user would then have to terminate the program and start it again. This matter should be pursued further. Measures should be taken to guard against this from happening.
Thirdly, it was found that the capacity of the PC used to monitor the NWA was poor in that it always stalled. If the LabVIEW program, say NWAV-1b, was left idle for too long - such as 10 minutes, it was very likely that it would hang. On the other hand, it is advisable to increase the RAM of that PC. The current memory of 8 Mbytes was too little to meet the requirement for its applications.
Return | Table of Contents | Proceed
|
|
