Feasibility analysis and ANN-based tuning of a capacitorless bandpass biquad

A capacitorless all-OTA bandpass biquad is tuned by utilizing an artificial neural network (ANN) with updated training sets. The training set contains a few tens samples which is varying in experiment. The training set is selected from predefine bias points that are closing to the desired biquad requirement. A second-order bandpass requirement, centered at 406.2 MHz, is successfully tuned as a sample. Feasibility can be easily indicated by observing maximum error of the worst record in the initial training set. Experiments indicate the unnecessary of training set that contains over 10 records. In addition, feasibility is effectively estimated by threshold of 10% maximum error of the worst record in the training set as this value almost yields no type-I and type-II errors no matter how large the training set is.