I'll be surprised if anyone can justify the xilinx definition (explanation) of pre-emphasis in it's entirety. Certainly this type of signal "pre-emphasis" (pre-compensation) can improve hi freq losses if properly done. What it does to phase and ISI may be another story. The results will show up in the eye pattern of a digital stream. What is happening, it appears to me, is that hi-freqs are being emphasized in an attempt to overcome a Lossy Low Pass Filter (LPF) characteristic of a transmission line. Note: LPF's have "dispersion" So, call it pre-emphasis -- but it is more like what is generally call pre-compensation or pre-distortion. This is not, of course, the analog signal pre-emphasis concept where the notion of de-emphasis is employed on the receiving end - to improve overall Signal to Noise ratio. To say this compensates (overcomes?) LPF dispersion may be correct if it sufficiently pre-compensates or pre-distorts in such a way that the result after the Filter (transmission line) is more like the original signal -- in an important way. Good Luck dong this as a general case. But it is possible in specific cases. The eye diagram will tell the story. Additionally, I have not seen a case where dispersion was a problem in PCB traces. If you or anyone find such a case, let me know please as I have been searching for this. Let's see what other have to say. Sincerely, JC Pat Darden JC Darden Consulting At 09:39 AM 7/8/02 +1000, you wrote: >There is something about that Xilinx text (it can be found at=20 >http://www.xilinx.com/publications/products/v2pro/handbook/ug012_glossary.p= >df) which left me thinking. > >The thing that puzzles me is that they specifically mention dispersion. No= >w, I understand how pre-ephasis will help in the presence of all the othe= >r effects mentioned, but how can pre-emphasis "magically" counter dispers= >ion? > >Could somebody please share his/her insight into this? > >Thanks, > >Sandor > >--- >Sandor Daranyi >Senior Design Engineer >Aristocrat Technologies Australia Pty Limited > > > > -----Original Message----- > > From: DoraBabu - CTD, Chennai. [mailto:dbabu@xxxxxxxxxxxxxxx] > > Sent: Saturday, 6 July 2002 2:27 > > To: lincj@xxxxxxxxxxxxxxxxxxx; hpc@xxxxxxxxxx > > Cc: SI_LIST (E-mail) > > Subject: [SI-LIST] Re: TO define the "Preemphasis" > >=20 > >=20 > >=20 > >=20 > > Got it from Xilinx Signal Integrity glossary :=20 > >=20 > > Pre emphasis is magnitude boosting of high frequency spectral=20 > > components > > before launching the signal (wave) onto the Transmission=20 > > Line. Transmission > > Lines embedded in most standard PCB materials (FR4, Rogers=20 > > 43xx, Nelco and > > Rogers) suffer varying degrees of dispersion and loss in the=20 > > 1 gigahertz > > spectrum. This is mostly due to conductance losses (leakage=20 > > from the copper > > trace to any other conducting structure) and Skin Effect.=20 > > Dispersion is a > > phenomenon whereby spectral components travel at different=20 > > velocities. The > > waveform looks smeared when it arrives at the receiver.=20 > > Both of these "characteristics" play into a diminished and=20 > > poorly received > > signal. By boosting the high freq. spectral components, the=20 > > magnitude of > > these components can be diminished as the wave travels through the > > Transmission Line, but since it starts out larger than the=20 > > lower frequency > > components, the composite signal arrives at the receiver=20 > > looking the way it > > was intended.=20 > >=20 > > Pre Emphasis is done by simply increasing the maximum amplitude of the > > signal for one bit period. If the signal is 1 bit in > > duration, the amplitude > > is allowed to rise to a value which is some percentage > > greater in magnitude. > > At this point, if the signal is to stay at the same logic > > state, the driver > > sends a decreased magnitude signal, or nominal logic level. > > Every time a > > transition occurrs, the greater magnitude level is used. For > > all times after > > this that the same level is to be transmitted, the nominal > > magnitude is > > used. > > > > --Regards > > Dorababu > > > > -----Original Message----- > > From: Miller [mailto:lincj@xxxxxxxxxxxxxxxxxxx] > > Sent: Friday, July 05, 2002 5:38 PM > > To: hpc@xxxxxxxxxx > > Cc: SI_LIST (E-mail) > > Subject: [SI-LIST] Re: TO define the "Preemphasis" > > > > > > > > You can also check xilinx's Multi-Gigahertz Signaling FAQ > > http://www.xilinx.com/xlnx/xil_prodcat_product.jsp?title=3Dsi_gigfaq > > > > Best Regards > > > > Miller Lin > > > > > > >------------------------------------------------------------------ >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 > ------------------------------------------------------------------ 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