From: Larry May <larrymayster@gm...>  20110830 21:11:18

Hi, Thank you for the amazing work you've done on Qucs. I am just getting started with Qucs so the problem may be a misunderstanding on my part. I am doing a 1 port Sparameter simulation of an impedance transformation from 500 ohms to 20 ohms at 9 MHz. I am using a 2 element match, a series L and shunt C. I designed my component values, set my port1 source impedance to 20 Ohm and ran the simulation. To begin the match seemed to work as designed, dB(S[1,1]) was less than 30 dB. However if I use rtoz(S[1,1]) to calculate the complex input impedance the result is 50 + j10.3 Ohms. Not the good 20 Ohm match the S[1,1] indicates. If I change the source impedance to 50 ohms rtoz() calculates the input impedance to be 20 + j4.11 ohms, more like what I expect, and the dB(S[1,1]) rises to only 8 dB. Since the input impedance of the network is independant of the ports source impedance I expected rtoz()'s result to be indepedant of the source impedance as well. Am I misinterperting what rtoz() calculates? I suspect an assumption of 50 ohm source impedance was made in rtoz(). I am using Qucs 0.0.15 from the Ubuntu repositories. The file in question is attached. Thanks, Larry 