Not logged in [
Login
-
Register
]
You Are Not Registered Or Not Logged In
Corsa Sport
»
Message Board
»
Off Day
»
House Day
»
Cinema Room
» Post Reply
Post Reply
Who Can Post?
All users can post new topics and all users can reply.
Icon:
Formatting Mode:
Normal
Advanced
Help
Andale Mono
Arial
Arial Black
Book Antiqua
Century Gothic
Comic Sans MS
Courier New
Georgia
Impact
Tahoma
Times New Roman
Trebuchet MS
Script MT Bold
Stencil
Verdana
Lucida Console
-2
-1
1
2
3
4
5
6
White
Black
Red
Yellow
Pink
Green
Orange
Purple
Blue
Beige
Brown
Teal
Navy
Maroon
LimeGreen
Message:
HTML is Off
Smilies are On
BB Code is On
[img] Code is On
[quote][i]Originally posted by Dom[/i] [quote][i]Originally posted by ed[/i] Digital signals are really susceptible to noise. Anyone own a DAB radio? I could understand if the debate was about why you shouldn't spend £100 in a 1m-2m cable, but £10 is acceptable in the grand scheme of things. [/quote] Been a while since I did anything (Hoeg's and Lauterbach's 'Digital Audio Broadcasting' is worth a read) with DAB, but if I remember correctly it'd uses/used convolutional UEP (unequal error protection) where you can vary the strength (aka code rate) of the error protection within an audio frame (each frame was split into four groups iirc - header info, code rate/scale info, audio samples and CRC checking). What caused the issue were the audio samples having the weakest error protection compared to the other groups, which resulted in a large intermittent area (audio drop-outs/muting) in reception between useable signal and no signal. Whereas most digital wireless systems tend to have a steep roll off between signal and no signal. That's the gist of it from what I can recall of my first year lecturers. [quote][i]Originally posted by Steve[/i] what i was getting at was digital signal (as far as a cable is concerned anyway) is fairly fault tolerant, as long as the signal carrying the info gets there at all it won’t be affected, only if part of a strand of that signal is lost completely will you notice :) im not arguing that shouldn’t spend 10 quid on the cable, i just wanted a discussion on the interference side of it [/quote] It's not as simple 'if the signal gets there then all is ok'; it's not. What you receive out of one end of a cable is, usually, not an exact copy of the original source. There will be differences (voltage/impedance); granted they'll be stupidly minute, but differences nonetheless. As said, one of the main issues with digital signal/transmission is clocking (there are a number of ways this is measured, usually it's edge-to-edge) of the signal between the D/A's and A/D's, where excess jitter/'clock-skew'/spikes can occur (usually) by induced interference; also there are other factors like temperature where an increase/decrease of temperature on the crystal/resonator causes a variation (aka temperature-drift( in the clock signal. Now a spike in the clock signal could be interpreted to sample an extra bit, so instead of a 32bit 'word' you end up with a 33bit 'word' or bits get shifted down so a significant bit (within a 'word') becomes a least-significant bit etc. And if you get rounding of a clock pulse it causes issue with measurement especially with edge-to-edge. There are a number of ways to manage this, but usually you'll have a start-frame/sync-bit that is sent prior to each 'word' which in essence readies/syncs the converter for the next word. You can also have differential clock lines, which I believe is used on HDMI (don't really know a huge amount about digital video transmission), where you have a number of clock lines (you'll also tend to have a clock ground/shield, usually grounded at one end which basically acts like a faraday cage around the clock lines) and you compare (how you compare depends on protocol) between them to attempt to cancel out variations/induced interference in the clock signal. You tend to find this on applications where you're clocking at stupid speeds - ie: GHz or 100's of MHz. And regarding clock signal transmission, this can be either via separate clock lines or the clock signal can be embedded within the data line eg - biphase mark (S/PDif aka AES3 protocol) or Manchester code (ethernet). Anyways this is the basics, The Sun/Daily Mail layman terms, of clocking anyways. You also have data side of it, but either way most digital protocols have a form(s) of error protection/correction to deal with issues like this. Edit - Apologies about the Paul_J essay and going waaaaayyy off topic. And feel free to correct any of this as it's been about 4/5 years since I’ve touched/read/looked at anything ADC/DAC/digital audio transmission/digital filter design related. [Edited on 26-02-2011 by Dom] [/quote]
Post Options:
Disable smileys?
Turn BBCode off?
Receive email notification of new replies?
This is a long topic, click
here
to review it.