Bruce, That is result of linear algebra. Any positive definite matrix could be decomposed into the product of its eigenvector matrices and eigen-value matrix. Those are the matrices you get from using [V,E] = eig(A) in many software like matlab. You can read more by reading book: "The mathematics of circuit analysis" by Guillemin or going through the online lecture by Gilbert Strang at http://ocw.mit.edu/courses/mathematics/18-06-linear-algebra-spring-2010/ Regards. Junwei -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Qin, Zhenshui (NSN - CN/Shanghai) Sent: Sunday, August 14, 2011 6:14 AM To: J_Dong@xxxxxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: De-embed fixture Hi Junwei, Can you help to explain more detail about how you get this equation? Af^-1 = V*(1/sqrt(E))*V^-1 (2) I am really a novice in this field, so please forgive me for asking such a basic question. Thanks a lot. Best regards, --------------------------------------------------------------------------------------------------- Bruce Qin 秦真水 Signal Integrity Engineer NWS ON RD HW Department Nokia Siemens Networks Technology (Beijing) Co.,Ltd.Shanghai Branch 诺基亚西门子通信技术(北京)有限公司上海分公司 9/F, Building 89, 1122 North Qin Zhou Road, Shanghai 200233, China 中国上海钦州北路1122号89号楼9楼 200233 Tel: +86-21-6085 5477 Email: zhenshui.qin@xxxxxxx --------------------------------------------------------------------------------------------------- -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of ext J_Dong@xxxxxxxxxxxxxxxx Sent: Friday, August 12, 2011 22:18 To: si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: De-embed fixture Bruce, That is a little tricky, but I think it is doable. Assume your fixtures are identical, preferable with female and male connector on each side (the same with DUT). Assume the measured 2-cascaded fixture network has s parameter of Sf2, whose abcd matrix could be found as Aa2. Let us still assume the overall s parameter for 3-cascaded network Sa, whose abcd matrix is found as Aa1. Also assume individual fixture has a matrix Af. We know: Aa1 = Af*Ax*Af Ax = Af^-1*Aa1*Af^-1 (1) where Af^-1 could be found from fixture measurement as: Aa2=Af*Af If the eigen-value and eigenvector matrices of Aa2 are E and V, from Aa2 = V*E*V^-1 solving the inverse of Af: Af^-1 = V*(1/sqrt(E))*V^-1 (2) Substituting (2) into (1) gives: Ax = V*(1/sqrt(E))*V^-1 * Aa1 * V*(1/sqrt(E))*V^-1 (3) Simply convert (3)into S parameter will give you the answer. Good luck. Junwei Dong -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Qin, Zhenshui (NSN - CN/Shanghai) Sent: Friday, August 12, 2011 5:36 AM To: J_Dong@xxxxxxxxxxxxxxxx; Ming Li; si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: De-embed fixture Hi Junwei, If what I can get is an S parameter file of 2 cascaded fixtures, let's call it Sf2, rather than Sf (the S parameter file of only one fixture), what can I do to de-embed it from my test results? Thanks. Bruce -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of ext J_Dong@xxxxxxxxxxxxxxxx Sent: Friday, August 05, 2011 00:26 To: 'Ming Li'; si-list@xxxxxxxxxxxxx Subject: [SI-LIST] Re: De-embed fixture Ming, Assume your measured S parameters are Sa (for all) and Sf (fixture itself), you can convert them to their individual ABCD matrices, say, Aa and Af. If your DUT has abcd of Ax, then from: Aa = A1*Ax*A1 you can solve Ax = A1^-1*Aa*A1^-1. After that, you can do an ABCD-to-S parameter convert. If I remember correctly, I think those functions are called s2abcd.m and abcd2s.m in matlab. You might need a little bit programming because the Snp file has frequencies in it, and I am not sure if matlab will take care of it automatically. Good luck. Junwei Dong -----Original Message----- From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On Behalf Of Ming Li Sent: Thursday, August 04, 2011 12:02 PM To: si-list@xxxxxxxxxxxxx Subject: [SI-LIST] De-embed fixture Hi, experts, I get two identical fixtures on both sides of my DUT. And I got the s4p data of the fixture itself and s4p data of the DUT cascading with the two fixtures. Could you please kindly let me know how I can de-embed the fixture and get the s4p data for the DUT? I know the math behind it but the data itself is too big to calculate. Thanks in advance. Regards, Ming Li ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List technical documents are available at: http://www.si-list.net List archives are viewable at: //www.freelists.org/archives/si-list Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu