2 Replies Latest reply on Apr 11, 2011 7:47 PM by freelectron

    Parasitic Extraction problem

    freelectron

      hi, all.

      I came across parasitci extraciton problems when using calibre, hope someone can help me:

      The process i used provides deep nwell(DNW) for pusb isolation. So for some NFETs, I had S and B connected together to a different potential(net, suppose its name is Vx) from Gnd.

      When run calibre c+cc parasitic extracion, i found the parasitic lumped capacitors of  these NFETs, which are supposed to these isolated psub(net Vx), are to the whole pusb(net Gnd)~ Obviously, this is not correct.

      So, how can i extract the parasitic capacitors correctly?

       

      many thanks!

      Glen

        • 1. Re: Parasitic Extraction problem
          samantha_lizak

          Hi Glen-

           

          If I understand you correctly, the problem is that you have two ground regions but the netlist is showing just one.

           

          You don't mention the Calibre version you are using, but check the Calibre xRC User's Guide for the section "Modeling Multiple Ground Regions."  It has been available for a few years, so chances are good your version supports that.  (I'm not sure if Calibre xACT offers the same feature, as PEX Ground Layer is handled differently in the two programs.)

           

          -Sam.

           

           

          • 2. Re: Parasitic Extraction problem
            freelectron

            Hi, Sam.

             

            yse, it's a problem about different ground regions, and I've tried "Multiple Ground Regions" a few times at first. But it seemed not work, so i came here for some suggestions.

             

            And I have fixed this problem already, which is just caused by the wrong "ground layer" name~

             

            still, many thanks!

             

            Glen