Hspice options for Crystal Oscillator !!!

Hi,

Can anyone suggest me on the hspice options required for running closed loop crystal oscillator simulations ? ... The reason for asking the above question is that even though the negative resistance of the circuit is about 1.5 times the crystal resistance, I do not see any oscillations at the output. But if i make negative resistance to be 1.8 times, then i see oscillations at the output which increases my power consumption by 20%. I suspect my hspice options for the above increase in power . I am using the following hspice options
.option accurate post=1 gmindc=1e-12 method=trap puretp

Please suggest any methods to improve the simulator accuracy so that it produces oscillations even if my negative resistance is < 1.8 times that of crystal . Ideally oscillations should be produced when negative resistance = crystal resitance. To have a margin i have taken my negative resistance to be 1.5 times .


Thanks
Hari .

try to use
.opt delmax=time

set time value to be 1/20 of your expected frequency

Hi ankit,
Thanks a lot ....Its working ...

good to hear that it worked..

normally .option accurate or runlvl takes care of internal time steps but some times in oscillator it beceomes a problem.
.opt delmax works there, but it slows down simulations

Also, remember crystal oscillators can take a while to start. I cannot remember the equation but it is proportion to Q and the period of the oscillation. High Q 32kHz oscillators can take a second or more to start.

Keith.