Question about using RELV option in Hspice

a question about Hspice.

I meet a strange question.
In simulation when i use the option of '.relv=1e-7' i get completely different results from it with '.relv=1e-3'(the default value).
The analysis is transient.
The former work,but the latter doesnt work.
I dont understand.
Can aynbody explain it for me?
thx!

The explain about RELV in the Hspice user manual is list thereinafter.
The option RELV sets the relative error tolerance for voltage.When
voltages or currents exceed their absolute tolerances, the
RELV test is used to determine convergence. Increasing
RELV increases the relative error. In general, RELV
should be left at its default value. RELV controls
simulator charge conservation. For voltages, RELV is the
same as RELTOL. The default value equals 1e-3.

a question about Hspice.

Hi billchen,
Are you sure you set the RELV option with statement:
.relv=1e-7

Normally, the above line will be ignored by hspice, but it make sense in your netlist. I don't know why. Is there a mistake in the post? According to the hspice manual, the syntax should be ".option relv=1e-7".

Is your design a very-low-power design?

Re: a question about Hspice.

yes
I use the statement of .option relv=1e-7.

Re: a question about Hspice.

Is your design a very-low-power design?[/quote]

My design is a set-up dcdc.