100 mhz colpitts oscillator
i design a simple osc. 100 Mhz (colpitts oscillator) using orcad 9.2 using supply 9 V & transistor J2N5485 .
the circuit doesn't work at all except when i put initial Condition (IC) at the output but i notice a very strange something
all the following result for time 20 μSec ( about 2000 cycle from 100 Mhz)
1- when IC=9 volt
the o/p dc level = 7.2 volt
the o/p rms level = 1.4 volt
freq = 88 Mhz
2- when IC=7.2 volt
the o/p dc level = 6.3 volt
the o/p rms level = 0.94 volt
freq = 87 Mhz
3- when IC=6 volt
the o/p dc level = 5.52 volt
the o/p rms level = 0.516 volt
freq = 85 Mhz
Q1 why the simulation results changes alot with IC?
Q2 why No. Osc. At ALL if IC >10 volt or IC<4.9 volt ?
another problem when the sim. done using auto step the freq=88 Mhz
when i set the step (much lower than the auto) the freq=92.65 Mhz ? !
any tips ?
any lect about sim ?
thx all
bye
The simulator must do a curve fitting exercise to reach a suitable operating point value...
so the ic value should be something that the circuit actually attains at a certain point of time...
for the equation solver to converge....now .ic tell the simulator that at time=0 a particular node has a particular precharge value....now putting this value to max P-P value especially when its an oscillator is necessary to add desired unstability in bias point as noise would do...i.e make the system unstable....that eases the load on simulator to converge.... as i your case i guess you would bw obtaining damped oscillations that decay rapidly....
Can you provide a schematic diagram?
Hi,
I used Cadence to design a LC VCO. I had same problems with the Initial Conditions. For some it works, for some it doesn't. What I ended up using was just a pulsed current source that injected current into the oscillation node. something not too big, maybe like 100uA, for a very short pulse. That seemed to do the trick. Oscillator works.
-Sam
try putting the supply as a pulse (i.e. 0 >> Vdd, with some rise time and very large pulse width, to last through all ur simulation time)
, this will give some noise for te oscillation to start.
AA
Try to put Pulse source with only Th = 1Fs and all othe parameters =0 (Tr=Tf=period=0) sires with your DC suplly , and it will work