Absorption by the contents of an equipment enclosure, particularly printed circuit boards (PCBs), affect the enclosure's shielding performance. At high frequencies, this absorption can be quantified using the angle of arrival and polarization averaged absorption cross-section (ACS). However, there is no available data on the high-frequency absorption characteristics of modern PCBs. In this study, we apply a reverberation chamber to the determination of the average ACS of a large number of PCBs taken from contemporary information and communication technology (ICT) equipment to provide a unique and comprehensive dataset. The ACS was found to range from 4 × 10 −4 -10 −2 m 2 from 2-20 GHz and different classes of PCB could be identified according to their surface characteristics. The "shadowing effect" of densely packed PCBs was also quantified for a subset of the PCBs. It was found that the ACS of a PCB in the stack was reduced by 20%-40% compared to its value when isolated. By way of a review of the general power balance analysis of an electrically large populated equipment enclosure in an external environment, we show how the acquired data will be useful for future qualification methodologies for ICT enclosures and PCBs.

No abstract

Abstract-The electromagnetic environment inside a shielding enclosure is affected by the absorption characteristics of the contents, which should therefore be represented in shielding measurements and simulations. At frequencies up to a few gigahertz, lossy dielectric materials have previously been used as surrogates for printed circuit boards in enclosure shielding assessment, both experimentally and in simulations. However, no systematic methodology for the design of these surrogates and their calibration against real hardware at high frequencies has been elucidated. In this paper we show how both lossy dielectric material and microstrip transmission line based "representative contents" can be designed and calibrated against real printed circuit boards over the frequency range 2-20 GHz using power balance concepts. The calibration is made by matching the average absorption cross-section of the surrogate to an average value for a class of real contents measured in a reverberation chamber. The surrogates are designed using efficient power balance models for layered media and field-excited microstrip lines and verified using full-wave simulation. The fabricated surrogates are validated by shielding measurements. The methodology presented could form an important part of future standards for enclosure qualification measurements that more accurately represent the internal environment of real equipment.

No abstract

This paper describes a greatly simplified model for the prediction of phase noise in oscillators which use a negative resistance as the active element. It is based on a simple circuit consisting of the parallel addition of a noise current, a negative admittance/resistance, and a parallel (Qlimited) resonant circuit. The transfer function is calculated as a forward trans-resistance (VOUT/IIN) and then converted to power. The effect of limiting is incorporated by assuming that the phase noise element of the noise floor is kT/2, i.e., -177 dBm/Hz at room temperature. The result is the same as more complex analyses, but enables a simple, clear insight into the operation of oscillators. The phase noise for a given power in the resonator appears to be lower than in feedback oscillators. The reasons for this are explained. Simulation and experimental results are included.