From: Stefan Jahn <stefan@gr...>  20070222 11:47:48

Am Mi, 21.02.2007, 19:51, schrieb Erasmo Alonso Iglesias: > Hello, Hi Erasmo! > I will comment the advantages of a PI/T topology in comparison with a L > topology... > On one hand, if we consider an output resistance, a L matching network and > an input resistance, the designer can't decide the bandwidth (or the Q > factor) of the matching, because Q factor using the L network is set only > by > the output resistance and the input resistance. On the other hand, if we > consider an output resistance, a PI/T matching network and an input > resistance, the designer can decide the bandwidth (or the Q factor) of the > matching, because Q factor using the PI/T network depends on the output > resistance, the input resistance and an arbitrary virtual resistance. > The PI/T matching network can be studied as a first L matching network, > connected among the output resistance and the virtual resistance, and a > second L matching network, connected among the virtual resistance and the > input resistance. Then, there are four types of PI/T matching networks: > Low > passLow pass, High passHigh Pass, Low PassHigh Pass and High PassLow > pass, according to the two types of the two L networks that form it. > I think it would be very useful to use a tool which generates a the PI/T > matching network from the central frequency, the bandwidth, the topology > (PI > or T) and the type (LPLP, HPHP, LPHP, HPLP). > I hope that you have understood my commentaries, greetings :) Thank you very much for your explanations. Makes perfect sense. I consider it a good thing as well... Probably Michael can put that into the GUI someday... Can you probably write a latexdocument describing the function and computation of those matching networks? Thanks, Stefan. 