I'm more after precise information to further refine the model I'm using, that's all. And, yes, the substrate plays an important role in determining the losses of basically any device at millimeter waves (even microstrips, because the bottom metals are thinner than the skin depth, and thus doesn't completely shield the signal strip from the substrate). From what Tim Edwards told me, Rsh for the P substrate is 4.4 kOhm/square. For a 750um wafer, that gives 0.3 S/m. Low, but that's actually a good thing for RF :^) Thread in Slack Conversation