http://www.stateye.org/developmentForum/ 2010-09-09T05:30:39+02:00 http://www.stateye.org/developmentForum/ http://www.stateye.org/developmentForum/lib/images/favicon.ico text/html 2008-07-26T21:31:14+02:00 http://www.stateye.org/developmentForum/doku.php?id=blogpage&rev=1217100674 Here's my new blog. ---------- You can even have a footer underneath your posts. text/html 2007-08-03T16:47:15+02:00 http://www.stateye.org/developmentForum/doku.php?id=documents&rev=1186152435 [Update as of 3rd August 2007] text/html 2007-07-14T21:58:54+02:00 http://www.stateye.org/developmentForum/doku.php?id=forum&rev=1184443134 I thought about setting up a majordomo email exploder, however, this is sometime difficult when the server doesn't belong to me. So in the meantime, I've decided to simple provide links here for emailing me or the group. Notice the automatic prefix in the subject. Please leave this in place to enable people to filter the emails. text/html 2007-07-02T19:35:20+02:00 http://www.stateye.org/developmentForum/doku.php?id=gui&rev=1183397720 * read legacy v4.2.2 xml scripts * remove transmit and receiver objects and combine parameters into a new equalisation object * create .py scripts based on final execution sequence requirements of engine text/html 2007-07-16T14:39:16+02:00 http://www.stateye.org/developmentForum/doku.php?id=inputstepapi&rev=1184589556 The standard/basic input to the stateye engine should be * A step response of the system under test (SUT) * A description of the sampling jitter * High frequency jitter (only bimodal uncorrelated described so far) * Medium band jitter (not included yet, as this is trivial) * A description of the equalisation of the SUT text/html 2007-07-23T21:43:12+02:00 http://www.stateye.org/developmentForum/doku.php?id=multistepinput&rev=1185219792 Stateye works on the principal of linear superposition. To this end the simplest step response of the transmitter would be a single response that can be used to generate all the required transmitter transitions, e.g. NRZ ; no emphasis 0 -> 1 1 -> 0 NRZ ; 2 tap FIR text/html 2009-05-29T11:14:59+02:00 http://www.stateye.org/developmentForum/doku.php?id=publications&rev=1243588499 Stateye Publications [Channel Compliance Testing Utilizing Novel Statistical Eye Methodology, DesignCon 2004] [Channel Compliance Testing Utilizing Novel Statistical Eye Methodology, Euro DesignCon 2004] [Statistical signal analysis (SSA) demystified, EETimes] text/html 2007-09-17T09:23:48+02:00 http://www.stateye.org/developmentForum/doku.php?id=python&rev=1190013828 Windows Under Windows I used the following. Simple install each one in the order below. And if you don't like Vi then you can always use your own editor ;-) * Python 2.5.1 Download * pywin download * iPython and pyreadline Download * numpy 1.0.1 Download * scipy 0.5.2 Download * matplotlib 0.90.1 Download * vim Download text/html 2008-05-05T16:21:56+02:00 http://www.stateye.org/developmentForum/doku.php?id=source&rev=1209997316 This page should be used for downloading of tested releases, or developements. The page should also be used for uploading proposed sources from developers. v4.2.2 released source v5 development source v5 development source (archive pre v5_080505) text/html 2007-09-14T11:19:17+02:00 http://www.stateye.org/developmentForum/doku.php?id=specification&rev=1189761557 Data Input Step input from scope * Initially a single step response time record will be requried * For more complex transmitter signals the use of a penrose matrix inversion can lead to multiple step responses representing the transmitter, however, I would suggest we leave this for now * The support of simple step responses is already supported in the XML definition language of v4.2.2. The support of multiple steps is open. * more details of this interface are discussed here text/html 2008-07-26T23:06:32+02:00 http://www.stateye.org/developmentForum/doku.php?id=start&rev=1217106392 Welcome to the Open Developers Forum for Stateye development. As you notice this a Wiki based environment, and I'm still experimenting with the best setup. You can request users now, however, to make sure we don't get any spammers, I must change your group afterwards to enable editing permission. text/html 2008-05-05T16:27:31+02:00 http://www.stateye.org/developmentForum/doku.php?id=testcases&rev=1209997651 v4 testcases v5 testcases v5 testcases (archived on 5th May 2008 v5 testcases (archived on 2nd August 2007 text/html 2007-07-30T08:07:35+02:00 http://www.stateye.org/developmentForum/doku.php?id=tum&rev=1185775655 Copy of Board Discussion concerning Jitter Distribution text/html 2007-07-02T12:37:45+02:00 http://www.stateye.org/developmentForum/doku.php?id=v4&rev=1183372665 The standard xml testcases used for testing v4.2.2 will be uploaded here. text/html 2008-06-22T21:48:52+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5.0.0&rev=1214164132 5th May 2008 Some serious development has taken place since the last release, but not all features targetted have been implemented. After spending large amounts of time implementing the receiver compliance capability of v5, time ran out to finish debugging the crosstalk and mode conversion features. These will be released within the next week, as it's simple a matter of debugging. Please refer to testcases examples of this release. text/html 2008-05-05T16:53:41+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5&rev=1209999221 5th May 2008 (Anthony) Based on code : v5_080505 The following [testcode] was used to verify the functionality of the v5 release. testcase.py provides a basic script interface for running an analysis, and also a command line parameter alter interface for some of the basic more used parameters. An example of this command line interface can be seen in testall.py. It should be verified that python2.5 is included in your search path, other this file will fail. Either add the installation direct… text/html 2007-08-02T13:30:47+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5_070802&rev=1186054247 02.07.07 [Anthony] Test code [test.py] for use with stateye module This simple example uses an input pulse respone [Pulse response, post de-emphasis] to create the pdf stateye [pdf stateye] 14.07.07 [Anthony] Small test code for checking pre-distortion of step response with DFE [dfe.py] This simple code shows how the step response of a bandlimited channel is superimposed to give a pulse response [pulse response] Clearly the postcursors of the pulse response are non-zero and lead to inter… text/html 2008-05-05T16:27:50+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5_080505&rev=1209997670 02.07.07 14.07.07 19.07.07 20.07.07 26.07.07 27.07.07 02.08.07 27.08.07 31.08.07 07.09.07 08.09.07 17.09.07 10.12.07 text/html 2008-05-05T16:21:27+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5_pre080505&rev=1209997287 2nd July 2007 [Anthony] First core module code [stateye.py] 8th September 2007 [Anthony] [v5_070908_2] need impulse.csv to be found in the testcase directory 17th September 2007 [Anthony] [v5.070916.a] Full functioning stateye with simple script front end. Is currently setup for a SAS channel which unfortunately seems to be too big to upload. Simple change the filename, and parameters and run from iPython The expected results are in the testcase directory. text/html 2007-08-31T15:58:24+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070702&rev=1188568704 02.07.07 [Anthony] Test code [test.py] for use with stateye module This simple example uses an input pulse respone [Pulse response, post de-emphasis] to create the pdf stateye [pdf stateye] text/html 2007-08-31T15:59:47+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070719&rev=1188568787 19.07.07 [Anthony] First results for checking 8b10b functionality with the updated stateye module have been completed. We are still using the test case pulse response [pulse response]. The test scripts are only slightly modified to previous testcases but now include a different pre-processing function call text/html 2007-08-31T16:00:42+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070720&rev=1188568842 20.07.07 [Anthony] Using a real pulse response (yet to be plotted) the real difference between 8b10b and scrambled start to make itself apparent. text/html 2007-08-31T16:02:18+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070802&rev=1188568938 02.08.07 [Anthony] Stateye v5 and DFE Using the script xx the following step responses were generated 1) Take a simple example with no DFE and a 2 tap FIR with coeficients of [0.84 -0.16] python> rundfe(3,0,0.16) We see here how the original stateye (v4) would have calculated the statistics for the ISI using the pulse response (red). Please note the underlying steps required to build this pulse response are shown for reference (blue). Moving from left to right (x-axis 0-10) we start with a … text/html 2007-08-31T16:02:43+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070827&rev=1188568963 27.08.07 [Anthony] Full blown implementation of DFE and FIR After some quiet, mainly due to holiday. I've got the above theory now in the [main engine]. The code still need commenting and is not fool proof, but shows that it's working. Given a step response, we can see the difference for a [0.9, -0.1] de-emphasis eye with and without a 4 tap DFE. text/html 2007-08-31T17:04:24+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070831&rev=1188572664 31.08.07 [Anthony] Success in using previously working DFE theory on real example. In addition to the DFE theory being incorperated within the [v5 engine], the ability to post jitter the stateye has been included. There is still much commenting necessary in the code, but the initial results for a real example are looking very promising. text/html 2007-09-08T15:42:31+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070907&rev=1189258951 7th September 2007 ; [Anthony] Replaced by 8th September 2007 text/html 2007-09-08T15:49:45+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070908&rev=1189259385 8th September 2007 ; [Anthony] I've delete the 7th September entry as these results completely replace the previous ones. Some phenominal results showing how v5 seems to working with 8b10b, DFE and FIR simultaneously. Also we are seeing the clear advantage of 8b10b over scrambled data, and also 100% correlation with v4 for the scrambled cases. text/html 2007-09-17T09:43:17+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_070917&rev=1190014997 17th September 2007 [Anthony] Here's the results to the v5 release v5.070916.a. Given a touchstone file containing a differential transfer function , the step and pulse responses can be generated for entry into v5 . With no jitter and 8dB of de-emphasis, we get a Stateye . However, after applying a standard distribution , we get . text/html 2007-12-10T16:27:41+02:00 http://www.stateye.org/developmentForum/doku.php?id=v5testcases_071210&rev=1197300461 Together with the [v5.071210.a] release of the source code, there is a new script file necessary as well,[template_071210]. This file references s-parameters files openly available from the SAS website, . The appropriate s-parameter file should be commented in and out according to which example should be run