In ADE-XL when I use calcVal in one test to refer to results from another test the simulation ordering seems to be right without me doing anything special. Either that or I have just been luck. I have just started to try switching over to Ocean-XL (because it is easier to re-use the moderately large amount of test setup on multiple similar designs) and I don't seem to be getting the correct orders. My sweep simulation seems to be running *before* my calibration simulation which of course makes the results look bad (they are uncalibrated). My sweep test has: desVar( "TRIM_CODE" "calcVal(\"Calibrated_TRIM_CODE\", \"calibration_test\")" ) I generated the ocean-XL code from ADE-XL. I'm using 6.1.6.500.14 Do I need to specify an order somewhere? Also if I need to specify an order, do I also need to specify a depencency? I have several tests and there are some dependencies as I've tried to indicate below (in Makefile syntax) sweep_test: calibration_test sweep_test2: calibration_test sweep_test3: sweep_testB: calibration_testB It would be helpful to have a "big picture" document for ocean-xl showing the relation between things like sessions, setupDB, tests, history, etc. A related question is if I have multiple sweep variables like: ocnxlSweepVar("VDD" "{Inclusion List}3.0 3.3 3.6{Inclusion List}") ocnxlSweepVar("sweep_temperature" "{From/To}Linear:-40:10:130{From/To}") how do I tell ADE-XL and Ocean-XL which of these variables should be the x-axis on plots and which should be a new waveform? For example, do I plot OUT vs sweep_temperature, one curve per VDD value or do I plot OUT vs VDD, one curve per temperature? This gets worse if I'm doing this over process corners and the corners have a variable. Maybe it has to do with which ocnxlSweepVar is listed first? Maybe it has to do with which variable has more points? Or maybe alphabetical? -Dan
↧