Hello Everyone (@Tim Edwards I think you might be able to answer this), I think I’ve gone slightly wrong in some point, but I’m not exactly sure where. I have a mixed signal design with the digital portions laid out using OpenLane. I have been using Magic to lay out all of the analog portions so I pulled the

/final/mag/xx.mag

file from OpenLane, however, upon exporting gds I find that the digital gds has some Klayout (not flagged in Magic) drc errors on the PSDM and NSDM layers. I tested just exporting only the digital portion gds (literally exporting the

/final/mag/xx.mag

file) from Magic and still got the same PSDM and NSDM layer DRC violations in Klayout. I then ran the same Klayout DRC check on the final outputted GDS from OpenLane and that is clean. So something is off somewhere. So, should I not use the

/final/mag/xx.mag

file? Is that not supposed to be used from OpenLane? Or is there perhaps a specific gds exportation setting I’m missing for Magic? The errors are few enough that I think I can manually fix them in Klayout, but I would like to know where I went wrong. Also, please forgive the uninformed question: what are the NSDM and PSDM layers? I don’t remember them from my Sold State class…. Thanks so much!

@Micah Tseng PSDM and NSDM are implant layers automatically created during magic gds out. There are situations when the hierarchical arrangement of cells results in DRC errors. There are multiple threads discussing this issue (search for PSDM). To see the actual data that magic will export, you can

cif see PSDM NSDM

(

feedback clear

to get rid of the markers). Tim offers a solution here https://open-source-silicon.slack.com/archives/C017HPHCMEY/p1647890613934859?thread_ts=1647872077.359559&cid=C017HPHCMEY

ah! that’s awesome! I missed this thread in searching before. Thanks a lot! And I apologize for the extra traffic.