At 04:15 PM 9/29/2005 +0100, you wrote: >I'm asking about the third possibility. >In my case I've got CS8900A on board ( ethernet part ) and PIC18F8621. >PIC works grand but CS doesn't even start. Chip is OK. My guess is that I >placed >power circuit on PCB badly. My guess is that it's really something else, like a connection you missed or a duff crystal. Despite the care we typically take with layout and bypass capacitors, IME most digital stuff will work on the bench with pretty awful layout and with minimal or no bypassing. Bypassing is not the same as ensuring stability- many LDO regulators will oscillate without an output capacitor in the right capacitance range and with the right range of ESR. Follow those recommendations to the letter. >So my questions are: >- how thick should be power and gnd paths Thick as possible. ;-) Ground planes are always better than traces. The layout is also important-- minimize loop area, don't have long serpentine power and ground traces. Lay them out *first*. That's true even if you're using an autorouter. Really, layout is more important than thickness in many cases. A bad layout with a 100-mil wide trace can be a lot worse than a good layout with a 10 mil wide trace, especially if there is low-level or high-current stuff on the same board. Your board sounds especially non-critical. >- what decoupling capacitors should I use Ceramic, possibly with an electrolytic for more capacitance and damping. Or you could use a large ceramic with a small resistor in series. >- how should be decoupling capacitors placed Near the Vdd and GND pins of each LSI chip, or as the manufacturer recommends (one AD part I just used has a couple of dozen 0402 bypass caps for every chip, as recommended). >- how distance between paths influence signal integrity Not much at low frequencies. >Basically I'm looking for some reference about it. I did look on internet >and I >found some resources but some of them tell different things. It's kind of a black art. You can look at professionally designed boards and copy the techniques. I have found _Noise Reduction Techniques in Electronic Systems_ somewhat useful, but you can probably find the same information for free on the net. Try trolling through application notes, in particular. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com ->> Inexpensive test equipment & parts http://search.ebay.com/_W0QQsassZspeff -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist