simulator options for low power

[analogLow]

hi,

I am simulating a low power circuit (nanoWatt power consumption).

Anyone have any recommendations on spice integration methods for this low power range?

So far, it seems that the Backward-Euler method produces the most stable and realistic results.

Thanks,
analogLow :razz:

 

[se7ensong]

Hi,

I don't know the answer, but backward integration is always stable. I know some forward integration algorithms will causing non-convergence problem. But I don't know why it is particular important in low power range ... Could you please help me?

 

[dick_freebird]

I'd worry more about accurate modeling of capacitances
and leakages, than integration method. But I like Euler for
its lack of sawtoothing (better damped solution than TRAP).