Hi, I found it strange when I simulate the crystal oscillator with tran analysis in spectre. When I set the initial current of 1uA in the parameter of CDF of the inductor in the crystal model, I observed that the current of the inductor(sin wave) starts from 0 and begin to oscillate with 1mA amplitude, when I  set it of 1mA, it starts from 0 and begin to oscillate with 1A amplitude. Then the envelope begin to decay to a stable value about 900uA.  So, I am confused about relation ship between the initial current in the component parameter and the value at time 0, also the initial swing of the inductor current.  And how can I set the initial current in the inductor in order to make the crystal oscillator reach the stable state quickly? Thanks.

hi_Skill_11,

Here is what I do
Since you already where able to get to steady state once, plot the current wave form and record the value of 5 equally spaced points on a half period of the waveform starting at zero and ending at zero amps.  Then get a ipwl enter the 5 points and simulate it should be at steady state within a half period.

The Q of the xtal is large enough to cause many simulators with nominal settings to converge abnormally which is why you may have seen the 1A temporarily.

http://rfcooltools.com

raja.cedt wrote on Jan 5^th, 2012, 9:50am:

In spectre, the convergence aid seems to have no way to set initial current but voltage. Thanks.