Floating input/output Problem in post simulation

Hi all,

I met a floating nets problem during the post-simulation. The DRC and LVS are cleaned. But when I used calibre PEX, many warnings: floating input/output occurred in the calibreview schematic. All the rf_cap, rf_ind and rf_mos models and sone other unkown instances are floating, so I can't proceed to post simualtion. I have checked the floating nets reported, but all the connections are fine. Could anyone give me some suggestion? Thank you!

Seems like rules or view-switching setup may have gotten
lost.

Hi! Only one rule file is needed in the PEX setup window. So I just added the calibre.rcx file. I don't think i missed the rule file. But what is the view-switching setup?

You might be able to get rid of them with PEX NETLIST VIRTUAL CONNECT YES (it joins nets with the same name) but this is really only recommended for subsystems -- parasitic extraction is layout based, and if there is no physical connection you should probably be aware of it. Is LVS only passing because of VIRTUAL CONNECT statements?

View-switching: as netlisting traverses the design
hierarchy, it "switches" down into the first-found
view in the switch-view list, and keeps going until
it finds one (or more) of the valid stop-views (it
takes the first-appearing in the stop-view list).

For example your switch view list may be something
like

schematic cmos.sch cmos_sch

and your stop-view

symbol spectre veriloga

Now some of the stop-views may have nothing that
the Caliber netlist needs, so return null content and
then you get this sort of no connections nowhere
result. It's not uncommon to have to rejigger the
stop-view list so that the one your tool needs, appears
first.

What that view might be, I don't know. I think I used
Caliber on one project a few years back, and it just
worked. Somebody musta set it up right.