Skip to main content

in a thread on another bbs i made a comment after another said something about harmonics in a S/PDIF stream... i was of the understanding that S/PDIF was an asynchronous serial protocol... not a modulated clock... anybody clear this up for me?

Comments

Boswell Sat, 12/29/2007 - 02:43

S/PDIF is a synchronous protocol with an embedded clock. What that means is that the clock is not present directly, but the protocol is designed so that an S/PDIF receiver can reconstruct a clock from the transmitted bitstream.

What was the comment about harmonics? This can't have related to audio harmonics, as a lossless digital transmission medium such as S/PDIF does not alter the digital words that it transmits. It may have referred to harmonics of the electrical bitstream when running down lengths of unterminated coaxial cable. The capacitance of the cable acts as an R-C filter and attenuates the high frequencies from the transmitted waveform. When the high frequency attenuation is great enough, the receiver is unable to reconstruct the bitsream clock, the result being pops and dropout. This limits the practical cable length to a few metres. A correctly terminated coaxial cable maintains high frequency response, and so cable lengths can be much greater before other effects kick in to limit the useful length.

dementedchord Sat, 12/29/2007 - 08:56

hey bos thnx for the quick responce... and yes i knew it was synchronous... the guy caught me off guard and i thought about it a minute or two after i posted this and didnt go back and change that... and the embeded clock idea makes sense to me as well... but the comment on the harmonics unless i misunderstood him was that thesignal was switching from 1.5 to3 Mhz and somehow modulated such that the harmonics extended to the 30Mhz range and that was the data stream... my picture then was esentialy like a classD amplifier... when i had thought we were dealing with a relitively simple serial data stream... i guess i need to reread the thread... i gotta be missing something here...

Boswell Wed, 01/02/2008 - 05:30

Not like a class-D amplifier, no, but a square-wave nevertheless, with all its attendant harmonics. As I said before, if the transmitter-cable-receiver chain attenuates, filters or distorts the square-wave such that the bitstream cannot accurately be reconstructed by the receiver, you will get bad audio.

dementedchord Wed, 01/02/2008 - 12:58

thnx again for the respoce... i did get it sorted out for the most part... and understand the way it is technically a modulated signal... also i think the original issue i was having was more about a theorical ideal verses real world transmission difficulties... and some language as well... lets see what was that about two peoples separated by a common language????