On Tue, Jun 5, 2012 at 6:54 PM, V G wrote: > > 1. Consider a parallel LC circuit where the bottom is tied to ground and > the top is tied to a mixed-frequency signal source. I understand that the > parallel LC circuit will shunt all frequencies to ground other than those > around its resonant frequency. After the signal source is removed, the > circuit will eventually lose its energy. How do I make an oscillator out = of > such an LC circuit that self-starts (that is, I can simulate in LTSPICE > which doesn't take into account external noise coming into an oscillator > circuit, which is required for some oscillator designs)? I've looked at L= C > oscillator circuit types like the Hartley oscillator, but I don't see how > this can self start in theory under ideal conditions. > Simulating oscillators in SPICE can be a little tricky, but if you are aware of potential problems you can tweak stuff as necessary to simulate the real world more accurately. For instance with external noise (or lack of), you can simply add a small "start up" voltage briefly to get things started, then set it to 0V so it does not interfere with the circuit (e.g. set a source up for something like 100 cycles/10mV at required frequency) However this is rarely necessary (unless things are really slow getting going such as in some crystal simulations) as selecting "start external DC supply voltages a 0V" usually does the trick. IIRC there are a few example oscillator circuits in /examples/educational. Pretty sure there is a Hartley example, plus Colpitts, Wien Bridge, etc. I'd have a look at these (and their settings) then try your own out. Sometimes you need to do stuff like change the maximum timestep, skip initial DC operating point condition, Integration method etc to make it work, but I've found LTSpice generally works pretty well (compared to some other SPICEs I've tried) with such circuits. --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .