Composite vs. Component Cables
 

G-squared wrote:
Matthew L. Martin wrote:
Smarty wrote:
ZVR,


To bad neither of you got the original poster's question. The use of

coax for one and two meter lengths of video/audio cable is for
shielding, not for impedance.

Matthew

--
Thermodynamics and/or Golf for dummies: There is a game
You can't win
You can't break even
You can't get out of the
game

I used to think that too until I ran into an impedance mismatch on a
short 18" cable. Actually what Sony did inside the monitor was put the
terminator at the back panel with an 18" pigtail into the PCB. Changing
the 18" cable to transmission line and putting the terminator at the
end cleaned up the ringing in the video.

Truly a unique design, wouldn't you say?

Matthew

--
Thermodynamics and/or Golf for dummies: There is a game
You can't win
You can't break even
You can't get out of the game

Composite vs. Component Cables
 

"Jim Gilliland" wrote in message
...
R Sweeney wrote:
"Jim Gilliland" wrote in message

I'm also not sure why we're seeing any discussion of twisted pair. The
audio cables in a composite bundle are coaxial cables. They're just not
the right coaxial cables for video.

I am looking at an AV cable that came with an old VCR that I keep on the
computer to copy family tapes to DVD... the audio lines have 24 AWG
printed on them.

This is NOT a coax designation, but the description of a twisted pair.

Not necessarily. The center connector of a piece of coax can be labeled
that way as well, especially if the coax isn't designed to meet any
particular standard.

Then again, it may be twisted pair in the case of your specific cable.
It's not impossible to transfer audio over twisted pair, it's just not a
particularly good idea. But I'm sure there are a few manufacturers that
have cut that corner.

In any event, I sure wouldn't try to use those wires (the audio ones) for
any form of video.

What do you believe to be wrong with transfering audio over a twisted pair?

Composite vs. Component Cables
 

R Sweeney wrote:

What do you believe to be wrong with transfering audio over a twisted pair?

Twisted pair wiring is great for audio - as long as we're talking about balanced
signals. Unbalanced signals (such as those used in consumer electronics) would
need a transformer to be converted to a balanced line. We do exactly this
(convert via a transformer) sometimes in the studio. In the pro audio
environment, nearly all audio is carried via twisted pair. However, it's
usually shielded twisted pair.

I'm not saying that it's not possible to force an unbalanced signal through a
twisted pair, just that it is not ideal. It's not properly matched.

Composite vs. Component Cables
 

G-Squared understands transmission lines and impedance matching and his
explanation is correct. The ***TRUE*** impedance of an RCA connector is 10's
of megohms if not higher. It only begins to look like 75 ohms at extremely
high frequencies where its' dielectric insulator separating the center
conductor and the (normally grounded) outer barrel creates a capacitive
shunt. This dielectric capacitance, sitting as a lump of bulk Teflon or some
similar material, is a small value. I would guess maybe 10 or 20 picofarads,
but this is merely a ball-park estimate. For it to be a 75 ohm reactance at
100 MHz, at the very high end of the frequency range where hi def video
resides, it would have to be a vastly larger capacitance. Alternately, to
measure as a 75 ohm impedance, the connector would have to have a shunt
resistance of 75 ohms resistive were the impedance to be exclusively
resistive. This is indeed not the case whatsoever. Put an ohmmeter across an
RCA connector. You will measure above 10 megohms. Try it and see for
yourself.

The point is that the connector only looks like 75 ohms at very high or
ultra high frequencies, and this is not where an RCA cable is used.

The cable itself, nominally coax, will have a characteristic impedance of 52
or 75 ohms nominally (for RG-58u and RG-59u/RG-6 coax respectively). The RCA
connector creates an impedance mismatch, owing to it being substantially
different from the surrounding distributed impedance of the line. This is
correctly stated by G-Man as a "bump", and indeed effects the standing wave
ratio as well as the propagation delay, creating time / phase shifts and
waveform distortion. The degree to which this creates problems is not just
defined by the cable and connectors since the termination of the cable at
the receiving end, if mismatched, can and will cause one or more reflections
to be created on the line. This can distort the waveform even if the cable
and connector are "perfect".

All of these matters are indeed beyond the scope of the original poster's
question, and provided only in response to the comment by Matthew that RCA
connectors have no impedance. The poster's original question can best be
answered (as it already has been) by simply saying that composite cables are
seldom if ever well suited for component video used in high definition video
applications. Longer cables tend to illustrate the weakness more
dramatically for the reasons stated above, but even short cables are a
compromise using these interconnects.

The primary reason RCA connectors are being used is that they are cheap, and
the differences in performance are not directly seen unless a comparison is
made to correctly designed cables.

Smarty


"G-squared" wrote in message
oups.com...

Richard Harison wrote:
Is it really the connector that has a natural impedance--or is it the
cable
itself with a natural capacitance (reactance)?

snip

Your understanding appears to be more in the theoretical than real
realm. Granted, an RCA connector is not a 75 ohm connector, if its such
a problem, why are they used on nearly all consumer video units?
Answer: it has very little inpact on the performance for 1 or 2
connectors. Whoever said an RCA connectors impedance in 10's of megohms
is also WAY off the mark. No, I DON'T know what it is but I surely know
what it isn't.

BNC connctors used to be 50 ohm, not 75, until it became an issue with
135 MHz serial digital video. The 50 ohm connectors were used for many
years in broadcast. So mis-matches were common on baseband video and
not terribly troublesome.

The question was about component/composite cables. The audio cables are
rarely if ever 75 ohm cable and is to be avoided for video though 75
ohm cable is OK for the analog audio. The component cables should all
be 75 ohm and the same length and TYPE as not all 75 ohm cable has the
same transmission velocity. If you can measure the velocity and adjust
the lengths so all the signals stay 'in time' you could mix cable
types with no problems.

GG



The inductance and capacitance is continuously distributed along the
length of the cable. The 'impedance' is where the reactive components
effectively cancel and it becomes resistive. The connector is only a
tiny part of the cable and just puts a 'bump' in the electical road
which is why you can get away with a less than perfect connector
particularly at the low frequencies of baseband video. The improper
impedance of the connector causes a small reflection in the cable i.e.
the VSWR is no longer 1. Signal-wise, it's identical to multipath. In
baseband analog video it shows up as smearing or ringing of the
transition depending on the severity of the reflection.

GG

Composite vs. Component Cables
 

R Sweeney wrote:
"Jim Gilliland" wrote in message
...
R Sweeney wrote:
"Jim Gilliland" wrote in message

I'm also not sure why we're seeing any discussion of twisted
pair. The
audio cables in a composite bundle are coaxial cables. They're
just not
the right coaxial cables for video.

I am looking at an AV cable that came with an old VCR that I keep
on the
computer to copy family tapes to DVD... the audio lines have 24
AWG
printed on them.

This is NOT a coax designation, but the description of a twisted
pair.

Not necessarily. The center connector of a piece of coax can be
labeled
that way as well, especially if the coax isn't designed to meet
any
particular standard.

Then again, it may be twisted pair in the case of your specific
cable.
It's not impossible to transfer audio over twisted pair, it's just
not a
particularly good idea. But I'm sure there are a few
manufacturers that
have cut that corner.

In any event, I sure wouldn't try to use those wires (the audio
ones) for
any form of video.

What do you believe to be wrong with transfering audio over a
twisted pair?

Shielded twisted pair is the norm in the broadcast industry running
balanced line. The idea is, the noise is induced on the center
conductor(s) vs the shield. With the noise on both audio lines IN phase
but the desired signal OUT of phase, its reasonably simple to use
either a transformer or better yet (to me), an Opamp differential input
which will add the desired signal and subtract the noise.

Unshielded twisted pair could work but becomes more interesting
particularly for an Opamp input as there is no ground reference via the
shield. It's never a good idea to rely on the power company for the
ground reference with small signals such as audio and it's certain
failure for video. It's relatively trivial for a transformer input
audio as there is no ground reference needed but it's better to use the
shielded twisted pair - as in fewer headaches.

GG

Composite vs. Component Cables
 

Thus XLR connectors having 3 pins!

--
All the Best
Richard Harison
"Jim Gilliland" wrote in message
...
R Sweeney wrote:

What do you believe to be wrong with transfering audio over a twisted pair?

Twisted pair wiring is great for audio - as long as we're talking about
balanced signals. Unbalanced signals (such as those used in consumer
electronics) would need a transformer to be converted to a balanced line. We
do exactly this (convert via a transformer) sometimes in the studio. In the
pro audio environment, nearly all audio is carried via twisted pair. However,
it's usually shielded twisted pair.

I'm not saying that it's not possible to force an unbalanced signal through a
twisted pair, just that it is not ideal. It's not properly matched.



----== Posted via Newsfeeds.Com - Unlimited-Unrestricted-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----