Hspice在跑tran時，time step的設定與影響？

doc

請問各位前輩...
  Y$ x. T! ?% c0 S% [- K一般hspice在跑 .tran時，大家會如何選取time step...
會取電路操作頻率的百分之幾？

  x5 q( U# |6 U5 X; ?  Z+ l小弟最近在模擬一個IO...
5 f/ l0 Y, R) X, t  K& R7 V8 H將time step設在0.1n以下，如0.05n或0.01n，模擬結果會產生震盪的現象...
output waveform會有一個高頻的小震幅上下抖動...
會導致這個模擬結果錯誤，LDO的輸出電壓飄到十幾伏(電壓源也才3.3V)，電流也飄掉...
試了一些option，結果還是一樣...

但把time step設成0.1n...結果就正常...
但又怕time step設不夠細...模擬結果不夠準確...

+ n( G% h, H0 ?7 c' T# I不知道有沒有什麼設定可以避免這個震盪的現象...
" G5 j' r8 R* M' n如果電路操作在5G...time step只是0.1n夠嗎？

hyseresis

How does the value of TSTEP affect my transient simulation?
9 {& e# c* z9 ITSTEP is specified in the transient analysis netlist command.
.TRAN TSTEP1 TSTOP1 TSTEP2 TSTOP2 ... TSTEPN TSTOPN
  G7 d5 h5 i% G+ M3 P! M5 UThe most common usage is a single TSTEP/TSTOP pair. For example:
4 \4 f# |. c& _.TRAN 0.1ps 100ns
TSTEP has a variety of effects on the operation of HSPICE.
( j9 u( y' Q% H- r1. When the RUNLVL algorithm is in use:
  - RUNLVL "in use" is defined as HSPICE version >= 2004.09 and OPTION RUNLVL > 0.
    For HSPICE version >= 2006.09, RUNLVL=3 by default.
  - The minimum timestep is determined by the RUNLVL algorithm. The RMIN option is ignored.
  z0 l* f; y( Y/ J* r( w: Z2 T  - The maximum timestep has a "soft" limit of RMAX*TSTEP. The RUNLVL algorithm
    allows the timestep to exceed RMAX*TSTEP as long as tighter Newton-Raphson
    tolerances are met (voltage, current, charge).The default value of RMAX is 5.
6 K# I; B5 S. I  _8 K4 g, `  - The DELMAX option can be used to override the RUNLVL algorithm's choice of
    maximum timestep. When DELMAX is used, TSTEP has no impact on simulation accuracy (except
    when used in conjunction with .OPTION INTERP, see #3 below).
* v: F6 B5 h/ Z, x: h; E5 h1 M2. When RUNLVL=0 (discouraged):
  - The minimum timestep is defined as TSTEP*RMIN. The default value of RMIN is 1e-9.
  - The maximum timestep is defined as TSTEP*RMAX. The default value of RMAX is 5.
* R  M1 _" D5 z  - These limits are "hard" limits.
  - If the simulation engine attempts to choose a timestep < TSTEP*RMIN, a
; v$ D9 H/ \- _  Y    Timestep Too Small Error will result.
  - The simulation engine is never allowed to take a larger timestep than
2 g2 G) O( |. P. e    TSTEP*RMAX.
Summary
With the introduction of the RUNLVL algorithm, the impact of TSTEP has been
$ i$ C. |/ Y6 J0 f; I2 Egreatly reduced to the point that it can be ignored except for niche applications
like INTERP. One should still be careful not to set TSTEP too small for long,
low frequency simulations in which the soft timestep upper limit of RMAX*TSTEP
! e) ^! [4 X" z. n7 ~: |9 ~may still force more timestep solutions than is necessary which will cause
excessive simulation time.

semico_ljj

step 可以是时钟频率的1/5~1/10，这样比较精确！