VCO & Start up Circuit
I had initially used damped periodic vsin source to initialize the designed pMOS based VCO, the transient response are of satisfactory from 0-1.8V of tuning voltage, but due to its nature of damped periodic source, I had difficulities in executing the PSS analysis, I shifted to a pulse source initialization at the Vdd and a pulse current source in parallel to the LC tank, the transient response however is not of satisfactory from 0-1V of tuning voltage. but however there is variation of frequency with respect to the variation of the tuning voltage. Can anyone kindly comment on this?. Thanks in advance
Rgds
Dear HRKHARI,
I shall try to explain nature of your problems. When you turn ON OSC all C and L are empty. If we have condition that is Aβ>>1 OSC(no Aβ>=1 or little higher than 1 this is only in steady state ) oscillation will start slowly until it arrive to steady state oscillation. During this process output is changing from zero to the final amplitude. Oscillation frequency will be change also from frequency close to Fosc. How much time it will take to arrive to steady state conditions with amplitude and frequency depend from design and wanted specification. General rule is better OSC take much more time to arrive to steady state conditions than poor one. Time we are talking going from 10μsec to msec (it depend from Fosc). DC conditions is problem in SPICE design ,good oscillator must have good DC decoupling and it takes very long time to arrive to steady state region. Also sometimes this DC conditions unable start up OSC if they are not designed correctly .Little easier problem is when we have oscillation and we like to change F. Your way to detect OSC will be under ? because this is not natural way how it will happen in nature . Except this problem it is possible problem with CAD convergence.... I wish you good luck in design because it is real very hard task to obtain good correlation between reality and simulation in this field. With better oscillators problem is growing very much analysis take much more time and problems in design are growing.
GL XTASA
Did you put some initialization timle in PSS ? Or Could you start your simulation from when steady state arrives to full oscillation conditions are satisfied ?
If you link your design here , I may help you...