OPTIMIZATION OF FREQUENCY FILTERS VIA VERTEX GRAPHS.pdfVIP

BIOLEK,D.: Optimization of frequency filters via VERTEX graphs. ISCAS2001 Sydney, Australia. 1/31 OPTIMIZATION OF FREQUENCY FILTERS VIA VERTEX GRAPHS D. Biolek, V. Biolkova*) Inst. of Telecommunications, Inst. of Radioelectronics*) Brno University of Technology, 612 00 Brno, Czech Republic dalibor.biolek@vabo.cz, biolkova@urel.fee.vutbr.cz http://www.vabo.cz/stranky/biolek BIOLEK,D.: Optimization of frequency filters via VERTEX graphs. ISCAS2001 Sydney, Australia. 2/31 ABSTRACT A method of systematic investigation of the influence of element parameters on the desired properties of a linear dynamical system –frequency filter in particular – is described. The couplings among the element parameters are examined by means of so-called VERTEX (or coupling) graph with the aim of algorithmic circuit optimization. One of the results is the partition of a set of parameters into disjoint sets. Within these sets, it is possible to vary the parameters independently to obtain the required circuit behavior. BIOLEK,D.: Optimization of frequency filters via VERTEX graphs. ISCAS2001 Sydney, Australia. 3/31 linear system V ... vector of voltages I ... vector of currents P ... vector of element parameters linear equations V, I nonlinear equations P (Signal) Flow Graphs VERTEX (coupling) graphs CAUSAL flow graphs basis transformation BIOLEK,D.: Optimization of frequency filters via VERTEX graphs. ISCAS2001 Sydney OPENING EXAMPLE (1): passive RCCR bandpass filter R1 160 R2 16kC1 100n C2 1n Vin Vout 2 21 sb s V V in out = , where sCRa μ16221 == sCRCRCRb μ16.322122111 =++= 2 21212 256 psCCRRb == et of 3 nonlinear 1 1 sb a ++ S, Australia. 4/31 equations with 4 unknown variables BIOLEK,D.: Optimization of frequency filters via VERTEX graphs. ISCAS2001 Sydney, Australia. 5/31 VERTEX graph R1 C1 R2 C2 b1 a1 b2 b2=R1C1R2C2 b1=R1C1 + R1C2 + R2C2 a1=R2C2 summing VERTEX simple VERTEX basis simple VERTEX BIOLEK,D.: Optimization of freq