> While discussing with my friend, he told me that the PCB trace > impedance should be maintained as 50 Ohms (for differential=3D > 100 and for some RDRAM buses 27ohms). Why 50 ohms? There are many parts to this question. Not really related to what you are asking, is why does so much test equipment use 50 ohms, and why are so many coax cables 50 ohms. Some of the previous replies answer this. Another reference that I have pointed people to for years (decades, really), is in the Trompeter catalog. A shortened version of their description is on the first page of http://www.trompeter.com/assets/product/PDF/wiring.pdf I believe that 50 ohm cables came first (to match antennas), and 50 ohm test equipment followed. Now, why does so much PCB wiring also use 50 ohms? A lot of reasons. When it comes to PC boards, the short answer is that there is no magic to 50 ohms. Unless you have a standardized signal like Ethernet or SCSI, you DON'T necessarily need to have 50 (or 100 or 27) ohm traces. But you DO often need to know what your trace impedances are. Depending on risetimes and trace lengths, some terminations may be required, and then you should know what are your driver impedances, load impedances, and board trace impedances, and what their mismatch does to your signals. It may be that the drivers were optimized for driving 50 (or 70 or whatever) ohm traces, and then you ought to target that value (whatever it is), when you make your PC boards. If you're doing, say, Ethernet, then you should maintain its cable impedance (50 ohms S.E. or 100 differential) on your PC boards, for those signals connected to the cable. And there is some PCB "history." ECL is an older but very fast logic family. It often requires terminated lines, even on moderately short traces. Many ECL output drivers were designed to drive 50 ohm terminated loads, which also happens to be ideal for driving, or being driven by, standard test equipment. The need to interface with test equipment continues to this day, and can be one reason for sticking with 50 ohm traces. If you ignore all the history, had no standards to adhere to, and no test equipment to interface with, you could choose your own impedance. You would probably end up with in the vicinity of 30-100 ohms anyway, because that's what you get with typical trace widths and materials on multi-layer boards. Designing everything for lower impedances (50 ohms or lower) requires somewhat more switching power, bigger transistors, but is better in the face of parasitic capacitances. There may be some advantage too in electrical crosstalk immunity. On the downside, lower impedances might be worse for magnetic crosstalk, and more di/dt noise. Regards, Andy ------------------------------------------------------------------ To unsubscribe from si-list: si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field or to administer your membership from a web page, go to: //www.freelists.org/webpage/si-list For help: si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field List archives are viewable at: //www.freelists.org/archives/si-list or at our remote archives: http://groups.yahoo.com/group/si-list/messages Old (prior to June 6, 2001) list archives are viewable at: http://www.qsl.net/wb6tpu