## [Qucs-help] two ports in equation block

 [Qucs-help] two ports in equation block From: Chuck McGuire - 2007-06-15 00:11:21 ```I'm trying to de-embed a transistor. I would like to read in 3 sets of two port s-parameters. S-dut, S-pads-open, S-pads-short. Then 1) S_dut -> Y_dut, 2) S_open->Y_open, 3) Y_dut1 = Y_dut - Y_open, 4) Y_dut1 -> Z_dut1, 5) S_short -> Z_short 6) Z_dut_intrinsic = Z_dut1 - Z_short 7) S_dut_intrinsic -> S_naked Look at the nice transistor without parasitic pads It looks to me like I have only one set of S-matrix during simulation time. m=eye(2) will create a new matrix but I don't know how to assign m[1,1]=S[1,1] will not work. as the L-value sub-scripting doesn't seem to be accepted. thanks Chuck McGuire -- Chuck McGuire ```

 [Qucs-help] two ports in equation block From: Chuck McGuire - 2007-06-15 00:11:21 ```I'm trying to de-embed a transistor. I would like to read in 3 sets of two port s-parameters. S-dut, S-pads-open, S-pads-short. Then 1) S_dut -> Y_dut, 2) S_open->Y_open, 3) Y_dut1 = Y_dut - Y_open, 4) Y_dut1 -> Z_dut1, 5) S_short -> Z_short 6) Z_dut_intrinsic = Z_dut1 - Z_short 7) S_dut_intrinsic -> S_naked Look at the nice transistor without parasitic pads It looks to me like I have only one set of S-matrix during simulation time. m=eye(2) will create a new matrix but I don't know how to assign m[1,1]=S[1,1] will not work. as the L-value sub-scripting doesn't seem to be accepted. thanks Chuck McGuire -- Chuck McGuire ```
 Re: [Qucs-help] two ports in equation block From: Stefan Jahn - 2007-06-15 10:55:30 ```Am Fr, 15.06.2007, 02:11, schrieb Chuck McGuire: > I'm trying to de-embed a transistor. I would like to read in 3 sets of > two port s-parameters. S-dut, S-pads-open, S-pads-short. > Then > 1) S_dut -> Y_dut, > 2) S_open->Y_open, > 3) Y_dut1 = Y_dut - Y_open, > 4) Y_dut1 -> Z_dut1, > 5) S_short -> Z_short > 6) Z_dut_intrinsic = Z_dut1 - Z_short > 7) S_dut_intrinsic -> S_naked > > Look at the nice transistor without parasitic pads I see. The usual 2-step deembedding. > It looks to me like I have only one set of S-matrix during simulation > time. Correct. > m=eye(2) will create a new matrix but I don't know how to assign > m[1,1]=S[1,1] > will not work. as the L-value sub-scripting doesn't seem to be > accepted. Also correct. So currently you can't do that. Sorry for the inconvenience. What I plan to implemented someday is to 1) allow subscripts (e.g. S[1,1] = 1, A[2]=2, etc.) in assignments 2) allow input variables in equations from other datasets Point (2) will then allow you to store S_open, S_short and S_dut in three different datasets and compute the deembedding... Thanks for trying Qucs, Stefan. ```