Hey folks! I have two questions to ask. First, there's a rumor about a maximum continous width of 30um for all metals (or some metals). Is this true? If so, what are the rules for it? Second, how do I prevent automatic tile placement inside a specific circuit, or parts of it? @Tim Edwards

@Leonardo Gomes: There are usually stress rules for processes that prevent wide metal without slotting due to delamination effects. I could not find any such rule in the SkyWater DRC decks. It is possibly because their tools automatically insert slots where necessary during mask generation. Since I haven't been able to confirm that, I have added a maximum width rule to magic. You can consider it an "advisory" rule, but I think it is prudent when you have wide metal buses to slot them in the direction of current rather than depend on automatic tools to do it correctly. 2nd question: Use the

fillblock

layer to prevent generation of fill shapes. Use it sparingly, because too much fill blockage can cause the design to fail metal density checks.

I'll relay the information to my team. Can you just tell me what value of max W did you consider for this advisory rule? As for the fillblock layer, I intend to use it just to avoid random tiles inside (or below) transmission lines. Extracting gain at 60GHz is already hard enough without having to deal with extra random losses. By the way, what is the procedure to run density DRC checks using the sky130 PDK?

I used a 30um rule. I'm not sure if this is flagged outside of MiM caps. Obviously cutting slots in the bottom metal plate of a MiM cap would not be a good thing to do, but SkyWater doesn't have any limits on the MiM cap bottom plate size, either, so I applied the 30um limit. About fill block: That is a perfectly reasonable use of the

fillblock

layer.

@Alfonso Cortés fyi

After checking https://docs.google.com/spreadsheets/d/1oL6ldkQdLu-4FEQE0lX6BcgbqzYfNnd1XA8vERe0vpE/edit#gid=530166750 I saw reference to some maximum widths (rules m1.11, m2.11, m3.11 and m4.11), but only for the process with the CU flag (I'm guessing that's a copper-based version of the BEOL). Should we take these values for the aluminum BEOL too? @jeffdi @mkk @Tim 'mithro' Ansell @Tim Edwards