Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.787974
Title: Developing a multi-tone load-pull system for the direct extraction of Cardiff behavioural model coefficients
Author: Al-Rawachy, Azam
ISNI:       0000 0004 7973 0806
Awarding Body: Cardiff University
Current Institution: Cardiff University
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
The main objective of this thesis is to develop and utilize a high-speed measurement system based on PXIe modules for multi-tone measurements. This thesis addresses challenges that have been tackled during developing an accurate LabVIEW software to measure the multi-tone signal. Having solved these problems, the system's functionality was demonstrated by using it to extract some useful data such as stability and gain information for RF designers in, 'real time' during active load-pull measurements. In the final part of the thesis, the system was used to aid the development of the Cardiff behavioural model. One of the main challenges in the development of the Cardiff behavioural model is to correctly select the required mixing terms to have an accurate model. Initial work has been focused on determining the correct, phase polynomial coefficients of the Cardiff Model. The first technique presented in this thesis utilises two-tone measurements and the Fast Fourier Transform (FFT) to observe the mixing order from the resultant intermodulation (IMD) products, which are directly associated with the Cardiff Model coefficients. Employing the IFFT, the selected tones have been transferred to the time-domain travelling-wave. This allowed for b-wave analysis and load modulation coverage to be seen on the Smith Chart. The resultant load-modulation from the two-tone measurements has been used as a target for CW impedance measurements to verify whether the identification holds for the CW domain. The result shows there is an VII offset between the CW measurements and the two-tone measurements for larger annuli as the α₂₁ exhibits magnitude and phase variation. Nevertheless, the maximum model order, obtained from the two-tone measurements, is acceptable for CW measurements when NMSE < - 40dBm. The second technique in this thesis demonstrates the model complexity identification for both magnitude and phase related mixing terms. It utilizes a multi-tone measurements approach. Load-pull measurements are performed using an engineered multi-tone active load-pull excitation, A21₂₁(t), involving a carrier phase and amplitude modulation components. The tone frequencies are set to a prime number, which, enables clear identification of model complexity of the device response B₂₁(t), by avoiding spectral overlap of the associated mixing terms. This is considered an essential step in correctly and accurately modelling the device and the ability to visualize all the intermodulation (IMD) products. Formulating the Cardiff behavioural model in the frequency domain, with this selected multi-tone stimulus, allowed for the first time the direct extraction of the model coefficients. Additionally, formulating the model coefficient least mean square (LMS) extraction process using the FFT transformed data provides for a very robust, direct extraction, solution for model coefficient determination. Utilising the FFT to optimise least mean square (LMS) algorithm for model extraction gives a very robust, direct extraction, solution for model coefficient determination.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.787974  DOI: Not available
Share: