Cadence oscillator simulation problem

  • Thread starter Pleasedontreplytoemailadd
  • Start date
P

Pleasedontreplytoemailadd

Guest
hi,

i am simulating a crystal oscillator in cadence, which should ideally
oscillate at 1 MHz. i am finding a weird problem with the transient
simulation. say, when i simulate the circuit for <= 230 ms, it
oscillates perfectly, whereas when i simulate it for say 500 ms, the
simulation gets over very fast and the output shows a straight line
like a dc output.

Could some one enlighten abt the solution?

Thanx in advance
 
You will probably need to change your step size in your transient
simulation to a smaller value so that it increases accuracy.

Michael
 
On Tue, 15 Aug 2006 12:50:29 -0400, "michael_skoufis"
<skouf1974@nospam.yahoo.com> wrote:

You will probably need to change your step size in your transient
simulation to a smaller value so that it increases accuracy.

Michael
Indeed. maxstep is probably the way to control this. The default maxstep is
100th of the simulation interval - and so the timestep is probably relaxing so
much that it completely misses the oscillation.

Set maxstep to (say) 1/20th of the estimated oscillation period - that will
force at least 20 points per period. You may not need to be quite as tight as
that, but this can help the oscillator start up more quickly in the simulator.

Andrew.
--
Andrew Beckett
Principal European Technology Leader
Cadence Design Systems, UK.
 
Michael and Andrew,

thanx 4 the replies!

Andrew Beckett wrote:
On Tue, 15 Aug 2006 12:50:29 -0400, "michael_skoufis"
skouf1974@nospam.yahoo.com> wrote:

You will probably need to change your step size in your transient
simulation to a smaller value so that it increases accuracy.

Michael

Indeed. maxstep is probably the way to control this. The default maxstep is
100th of the simulation interval - and so the timestep is probably relaxing so
much that it completely misses the oscillation.

Set maxstep to (say) 1/20th of the estimated oscillation period - that will
force at least 20 points per period. You may not need to be quite as tight as
that, but this can help the oscillator start up more quickly in the simulator.

Andrew.
--
Andrew Beckett
Principal European Technology Leader
Cadence Design Systems, UK.
 

Welcome to EDABoard.com

Sponsor

Back
Top