J. G. is on the right tack, for instrumentation, think about which traces contain signals who's qualities you wish to preserve. Actually you should have done this at the schematic level and then decided where signals should be single ended or differential. Remember, that for quality, carefully matched impedence is very important, therefore thought shall not have signals caried by traces whose impedence changes by much. Events such as having a trace change thickness or pass over a gap in the ground plane matter a lot. Often, for matched impedence we keep critical signals on the two out side planes. Remember to caluculate the desired trace width and the spacing for differential using the thickness and dialectric constant of the plastic to the next layer. High end fab shops will ask you about impedence matching and help you in these calculations.
For high frequency digital signals we worry about ground bounce. Here the layer stack up is important and there is often a trade off between the number of layers and signal quality. I highly highly did I mention highly recomend reading the apps notes for the component vendors and looking for hints in how to do the layout. John Griessen wrote: > armdeveloper wrote: > > >> Does anyone have any strategies for placing components while still in >> the ratsnest stage that will make for a really nice compact auto routed >> board ? Or is it just trial and error ? I've always wondered this... >> > Think about grounding and shielding first, then about functional > divisions, then about how many layers and vias you want... > > I like a little book called Grounding and Shielding Techniques in > Instrumentation > by Morrison. > > Trial and error works. That's how humanity has progressed mostly. > > John Griessen > > _______________________________________________ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user