An update on modeling semiconductor noise in SPICE:
The AF and KF parameters will work in diode and BJT SPICE models, but the results obtained can only be viewed using an AC noise analysis in SPICE. No noise signal will be seen using a transient analysis.
If you want to "see" noise in a transient analysis, the noise has to be generated as a source, i.e. from a random number generator, PWL source, or ABM source.
.99
It is clearly demonstratabe that pn junctions , when reversed biased create high noise ..every one of them. It is not what we might usually want but ..please a 10k resistor , a zener or a led or a transistor ,scope it ...see the noise ! we shunt it away with caps and filters to get rid of it. Those dirty ss devices than Steven pointed out are every where in our circiuts and are NOT what we want for pure frequency. Feedback is used to keep it down too. Simulators are useless as they assume we will design with convention in mind . That said, a noise generator, like petare has shared can be useful for finding resonance points in coils that we might other wise miss. But only one source per circuit and clean after that . we would not want distortion and noise in our frequency correction signal at the exact moment that we needed to use it would we? If we knew exactly waht we were doing we might be able to put up with delays and noise and regenerate it at the right phase,even if it was a bit late and messed up by the pn junctions...that is privided that they did not just sut down because of the field generated by the device. http://www.physorg.com/news162132202.html
« Last Edit: 2010-06-23, 05:50:11 by Lindsay »
|