I got this warning when running a spectre (7.2) simulation on my design:
" Warning from spectre in `mimcap':`I6.I1.C44', in `Cascode':`I6.I1', in `DiffAmp':`I6', during hierarchy flattening
WARNING (CMI-2426): I6.I1.C44.R0: `r' = -162.911 mOhm is negative. "
any idea about what could cause "r" to be negative? Can this warning cause convergence problems or result in inaccurate results ?
You'd probably have to look into the mimcap model to see how the resistor value for R0 is computed. It's presumably a subckt model in your model files.
In reply to Andrew Beckett:
Thanks for the reply, you right ! This negative resistance was due to maximum capacitance length . R0 become negative when the length in over 40um. Then I separate the mimcap into small ones (as it is done in the layout) to avoid this warning.