DComTalk.com

Robert Redelmeier <redelm@ev1.net.invalid> writes:

Tomi Holger Engdahl <then@solarflare.cs.hut.fi> wrote:
F = 1 / (2 * PI * R * C)

where F is the corner frequency in Hz, PI is 3.14, R is the
source's output impedance in Ohms, and C is the cable's
total capacitance in Farads. A first-order filter has a
slope of 6 dB per octave. This means that beyond the corner
frequency, the response will drop 6 dB for each doubling of
frequency. Generally it doesn't seem likely that you would
get detectable loss even at 20kHz unless you have one or
more of these conditions: unusually high source impedance
(many kilo-ohms), unusually high capacitance cable or
unusually long cable length (tens of meters).

Typical transmission characteristics of CAT5 wiring: DC
Resistance: 8.99 Ohms/100metres Capacitance: 13.5-17 pF/feet
(45-57 pf/meter)

So for a 100m Cat5 cable, at 8 Ohms source impedence, the
corner frequency is 3.9 MHz . I don't hear up that high.
We're talking about speakers on Cat5. A phono cartridge
might have some trouble :)

By the way where do you see 8 ohms source impedance on practical
audio systems ?
I have not seen that much.
The output impedance of a typical hifi amplifier built using transistors,
fets, ICs etc.. have output impedance that is considerably lower
than 8 ohms. The ideal amplifier has zero output impedance
(=pure voltage source). Practical audio amplifiers typically have
effective output impedance considerably less than one ohm.
So in normal hifi system speaker output the source impedance
is less than one ohm and the load impedance is tha 8 ohms.
With some tube amplifiers situations can be somewhat different,
those have typically considerably higher output impedance than
modern amplifiers built using other techniques. So the 8 ohms
source impedance assumptation might hold for some tube amplifiers,
and for other practical amplifiers the source impedance is
considerably lower than this 8 ohms.

With normal line level signals the source impedances are
typically in 30 ohms to 5 kohms range depeding on the equipment
used. Professional audio equipment have typically the source
impedance in 30 ohms to 600 ohms range. Normal hifi equipment
typically have source impedance from few hundred ohms to
few kilo-ohms..

The capacitance of a normal shielded audio interconnection
cables is typically 2-3 times higher than the capacitance
of the CAT5 wiring! One meter of typical shielded audio
interconnection cable (RCA cable) has typically capacitance
of around 100 picofarads.

Good to know. So capacitance is a non-issue for using
Cat5 in audio.

-- Robert

--
Tomi Engdahl (http://www.iki.fi/then/)
Take a look at my electronics web links and documents at
http://www.epanorama.net/

Robert Redelmeier wrote:

I thought the load coils were to retain signal strength (volume),
not frequency response.

Without loading, the frequency response follows an inverse exponetial curve.
With loading, it's essentially flat over the desired range, then drops like
a rock.

http://www.its.bldrdoc.gov/projects/dev ... _coil.html

Tomi Holger Engdahl <then@solarflare.cs.hut.fi> wrote:

Robert Redelmeier <redelm@ev1.net.invalid> writes:
Tomi Holger Engdahl <then@solarflare.cs.hut.fi> wrote:
F = 1 / (2 * PI * R * C)
So for a 100m Cat5 cable, at 8 Ohms source impedence, the
corner frequency is 3.9 MHz . I don't hear up that high.
We're talking about speakers on Cat5. A phono cartridge
might have some trouble :)

By the way where do you see 8 ohms source impedance on practical
audio systems ? I have not seen that much. The output impedance
of a typical hifi amplifier built using transistors, fets,
ICs etc.. have output impedance that is considerably lower
than 8 ohms. The ideal amplifier has zero output impedance

I was using 8 ohms as an upper bound to give the lowest
corner freq. Since impedence is less, cutoff goes even higher!

-- Robert

"Robert Redelmeier" <redelm@ev1.net.invalid> wrote in message
news:OnSOd.7376$D34.1625@newssvr12.news.prodigy.com...

Tomi Holger Engdahl <then@solarflare.cs.hut.fi> wrote:
F = 1 / (2 * PI * R * C)

where F is the corner frequency in Hz, PI is 3.14, R is the
source's output impedance in Ohms, and C is the cable's
total capacitance in Farads. A first-order filter has a
slope of 6 dB per octave. This means that beyond the corner
frequency, the response will drop 6 dB for each doubling of
frequency. Generally it doesn't seem likely that you would
get detectable loss even at 20kHz unless you have one or
more of these conditions: unusually high source impedance
(many kilo-ohms), unusually high capacitance cable or
unusually long cable length (tens of meters).

Typical transmission characteristics of CAT5 wiring: DC
Resistance: 8.99 Ohms/100metres Capacitance: 13.5-17 pF/feet
(45-57 pf/meter)

So for a 100m Cat5 cable, at 8 Ohms source impedence, the
corner frequency is 3.9 MHz . I don't hear up that high.
We're talking about speakers on Cat5. A phono cartridge
might have some trouble :)

The capacitance of a normal shielded audio interconnection
cables is typically 2-3 times higher than the capacitance
of the CAT5 wiring! One meter of typical shielded audio
interconnection cable (RCA cable) has typically capacitance
of around 100 picofarads.

Good to know. So capacitance is a non-issue for using
Cat5 in audio.

-- Robert

I don't think that's what he was saying. He said that it depends on the
source impedance, length of cable, etc. For instance, the usual method
of running a PA system to a speaker dozens of yards away is to use thin
wire, could be cat5. So the PA puts out 70.7V line voltage. The
impedance is higher than 8 ohms, maybe 500 or 600 ohms. The matching
transformer at the speaker drops it down to 8 ohms. So the impedance of
the source isn't just 8 ohms, it's higher. And the length of a long
cable could affect the speaker's output, especially if you have a
tweeter for those freqs of 10k and above.

Another example is telephone line. You could run a POTS set over
hundreds of feet, maybe up to a few thousand, without any noticeable
effect at the higher freqs, 3 to 4kHz. But you get up to 5k or 10k
feet, and then you have to put loading coils on the lines to maintain
the quality of the telephone audio.

But what I've been talking about is that longer run of several hundred
yards. If instead I use a tube preamp to drive an amplifier, the output
impedance might be more than 10kohms. In that case, with the high Z, a
dozen yards (36 feet) of any cable, shielded, coax, cat5, or twisted
pair, might cause noticeable attenuation of the high freqs. That's why
the microphones for consumer use might be high Z, and will work fine
with a ten foot cord. But you put an extension cable on them and you
find them to be muffled. So with pro equipment, the mics are low Z, and
can have dozens of yards of extension cable without problem.

So this audio roadie guy gets an idea that he can make an adapter from
XLR to RJ-45, and run his microphones over his home netowrk. It should
work okay with low Z mics with balanced line, but the consumer mics
would need a transformer to drop the Z down to 150 ohms, and to balanced
line. Then at the other end, another to put it back to high Z
unbalanced line to go into his preamplifier. Putting the high Z
unbalanced mic signal directly over the cat5 would probably get him
nothing but HUMMMMMMMMM...

"James Knott" <james.knott@rogers.com> wrote in message
news:T6ydnYVTlcYvwZbfRVn-2g@rogers.com...

Robert Redelmeier wrote:

Al Dykes <adykes@panix.com> wrote:
isn't anyone concerned with high-end rolloff for
audio on UTP due to the capacitance ?

LOL! Do you think we have Golden Ears? Odd "capacitance"
that doesn't affect 100 MHz would somehow disturb 10kHz.


Actually, at least for long distances, cable characteristics are
important
at audio frequencies. For example the cable used to run telephone
service
to your home, has a significant rolloff within the voice band. The
phone
companies use loading coils to flatten the response within the desired
frequency range.

The roloff isn't limited to the voice band as you imply. It affects any
freq, but the longer the line, the lower the freq, until it's affecting
the voice freqs at thousands of feet.

"Robert Redelmeier" <redelm@ev1.net.invalid> wrote in message
news:DtSOd.7377$D34.6380@newssvr12.news.prodigy.com...

James Knott <james.knott@rogers.com> wrote:
Actually, at least for long distances, cable characteristics are
important at audio frequencies. For example the cable used to
run telephone service to your home, has a significant rolloff
within the voice band. The phone companies use loading coils
to flatten the response within the desired frequency range.

Load coils don't go on lines until 20kft (6km). At that distance,
I don't think the Cat0 line would have bandwidth (S/N=1) over 100 kHz.

We have 13kFt of telephone line between the CO and our campus. There
are two loading coils on many of those lines, and I've seen the tech use
one of those Progressive loading coil detectors and tune it, and it
shows two 'humps', indicating two coils. As far as I know, the loading
coils are approximately every 6kFt, not 6km. Almost all the lines had
them until they introduced DSL. But a few lines didn't have loading
coils because they were being used for T1 and other such services.

I thought the load coils were to retain signal strength (volume),
not frequency response.

It changes the freq response from (view with courier font)

|
|*
| *
| *
| *
| *
| *
| *
0----------------------------

To
|
|* * * * *
| *
| *
| *
| *
| *
| *
0----------------------------

so that the midrange freqs where the intelligibility is contained are
not attenuated as much. But the highs up around 3 to 4kHz may be
attenuated.


> -- Robert

Watson A.Name - "Watt Sun, the Dark Remover" wrote:

Actually, at least for long distances, cable characteristics are
important
at audio frequencies.  For example the cable used to run telephone
service
to your home, has a significant rolloff within the voice band.  The
phone
companies use loading coils to flatten the response within the desired
frequency range.

The roloff isn't limited to the voice band as you imply.  It affects any
freq, but the longer the line, the lower the freq, until it's affecting
the voice freqs at thousands of feet.

Quite so But I don't think that was what I was implying. I simply tried
to express the idea in a few sentences. I tried to find a graph, that
showed the effect.

Many years ago, I studied those graphs in my work.

"Watson A.Name - \"Watt Sun, the Dark Remover\"" <NOSPAM@dslextreme.com> writes:

We have 13kFt of telephone line between the CO and our campus. There
are two loading coils on many of those lines, and I've seen the tech use
one of those Progressive loading coil detectors and tune it, and it
shows two 'humps', indicating two coils. As far as I know, the loading
coils are approximately every 6kFt, not 6km. Almost all the lines had
them until they introduced DSL.

SOP is every 6Kft, with the first coil being 3Kft from the CO.

That spec was adopted for {real} T1. There, the reason for the
initial 3 Kft was the CO was a really noisy place with all those
crossbar and step switches clacking away; so they wanted more noise
margin on that leg.

I've never grokked what good the 3Kft spec was with analog...
--
A host is a host from coast to coast.................wb8foz@nrk.com
& no one will talk to a host that's close........[v].(301) 56-LINUX
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433

James Knott <james.knott@rogers.com> writes:

That spec was adopted for {real} T1. There, the reason for the
initial 3 Kft was the CO was a really noisy place with all those
crossbar and step switches clacking away; so they wanted more noise
margin on that leg.

Or maybe, they wanted the loading coil in the middle of each 6K segment.

Hmm, a different way of thinking about it. You could well be correct.

In case it was not obvious; having T1 design meet the 3Kft/6Kft rule
means a crew can go to that point on the cable, and mount the repeater
on the same pole as the existing "can" with the coils.

Fewer cable openings, less work, lower cost. For all its faults,
Ma's engineering design was well thought out toward minimizing low
TCO/lifecycle costs; the opposite of our current obsession with
""lowest price""....



--
A host is a host from coast to coast.................wb8foz@nrk.com
& no one will talk to a host that's close........[v].(301) 56-LINUX
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433

David Lesher wrote:

That spec was adopted for {real} T1. There, the reason for the
initial 3 Kft was the CO was a really noisy place with all those
crossbar and step switches clacking away; so they wanted more noise
margin on that leg.

Or maybe, they wanted the loading coil in the middle of each 6K segment.

"David Lesher" <wb8foz@panix.com> wrote in message
news:cuqb81$4me$1@reader2.panix.com...

James Knott <james.knott@rogers.com> writes:


That spec was adopted for {real} T1. There, the reason for the
initial 3 Kft was the CO was a really noisy place with all those
crossbar and step switches clacking away; so they wanted more noise
margin on that leg.

Or maybe, they wanted the loading coil in the middle of each 6K
segment.


Hmm, a different way of thinking about it. You could well be correct.

In case it was not obvious; having T1 design meet the 3Kft/6Kft rule
means a crew can go to that point on the cable, and mount the repeater
on the same pole as the existing "can" with the coils.

Fewer cable openings, less work, lower cost. For all its faults,
Ma's engineering design was well thought out toward minimizing low
TCO/lifecycle costs; the opposite of our current obsession with
""lowest price""....

Yeah, true. Howwever, Ma Bell thought that telephone networks would
always be circuit switched, and never considered alternative
technologies such as cell phones, IP and DSL for the last mile.

Now SBC is going to buy AT&T, 'Ma Bell' being taken over by one of its
baby spinoffs. MCI, too.

> --

Thanks for all that info
Ro


"Ro Cathain" <ro_cathain@hotmail.com> wrote in message
news:5k4Od.47119$Z14.32375@news.indigo.ie...

I have Cat 5 and Coaxial cabling installed in my house.

I want to feed (a) an audio signal, (b) video signal, from a wirelessly
controlled sub-floor PC to wall connections.

So, I anticipate using 4 of the Cat 5 wires to feed an audio signal to
hifi
equipment. (ok quality?)

Then, should I use the other cables of the CAT 5 to make up a scart
connection to the TV... or should I use the co-axial cable for best
quality???

Cheers.

Al Dykes wrote:

(someone wrote)

I want to feed (a) an audio signal, (b) video signal, from a wirelessly
controlled sub-floor PC to wall connections.

So, I anticipate using 4 of the Cat 5 wires to feed an audio signal to
hifi equipment. (ok quality?)

(snip)

isn't anyone concerned with high-end rolloff for audio on UTP due to
the capacitance ?

Professional equipment uses balanced low impedance for audio
signals to avoid that problem. Transformers on each end.
I believe it is usually shielded twisted pair, but UTP might work.

For speakers UTP ad too much resistance.

Again transformers should work, but this time the low impedance
side for the speaker/amp and high for the UTP.

The usual standard uses a transformer called 70.7 volt on one
side, and 4/8/16 ohm speaker terminals on the other. Of course
the actual voltage depends on the signal going in.

Those should work, though I have never tried it. They are commonly
used for PA systems in schools and such. Otherwise, home intercoms
usually use higher impedance, maybe 40 ohm, speakers.

-- glen

Robert Redelmeier wrote:

Al Dykes <adykes@panix.com> wrote:

isn't anyone concerned with high-end rolloff for
audio on UTP due to the capacitance ?

LOL! Do you think we have Golden Ears? Odd "capacitance"
that doesn't affect 100 MHz would somehow disturb 10kHz.

If you drive it with a 100 ohm source, then yes, but most line
outputs are 1000 ohms or so. 15pF/ft and 300ft is 4500pf.
4500pF*1000ohms is 4.5us, f=1/(2pi*4.5us)=100kHz, will have
some rolloff at 20kHz.

Telephone wiring uses 600 ohm source impedance and longer
wires. The parallel capacitance is balanced by series inductors
spaced out along the cable.

-- glen

In article <2LudnVbmRs6nyoneRVn-1A@rogers.com>,
James Knott <james.knott@rogers.com> wrote:

glen herrmannsfeldt wrote:

Robert Redelmeier wrote:

Al Dykes <adykes@panix.com> wrote:

isn't anyone concerned with high-end rolloff for
audio on UTP due to the capacitance ?

LOL! Do you think we have Golden Ears? Odd "capacitance"
that doesn't affect 100 MHz would somehow disturb 10kHz.

If you drive it with a 100 ohm source, then yes, but most line
outputs are 1000 ohms or so. 15pF/ft and 300ft is 4500pf.
4500pF*1000ohms is 4.5us, f=1/(2pi*4.5us)=100kHz, will have
some rolloff at 20kHz.

Telephone wiring uses 600 ohm source impedance and longer
wires. The parallel capacitance is balanced by series inductors
spaced out along the cable.

Actually, telephone cable has appreciable rolloff within the voice
frequencies, which is why loading coils are used when going significant
distances. The coils flatten the attenuation within the desired bandwidth,
at the expense of higher frequencies. However I doubt that roll off would
be significant at the distances used in a home.

As for those "Golden Ears" types, don't you know that their ears are so
good, that they can hear things that are physically impossible to hear? A
few years ago, there was a columnist (Larry Klien IIRC) in Radio
Electronics magazine, who'd occasionally take aim at those Golden Ear
types. Seems some of them can listen to two different audio systems, in
different environments at different times and tell which one sounds best,
but they also reject A-B tests as invalid. They're also the idiots who
tend to buy stuff such as Monster cables and other such junk.

Jeeesh, this thread came back from the dead. When people ask about
using telco wire for HiFi speakers (amps into a low impedance) is
making a mistake and it's possible to show rolloff, IMO. (I use 12 ga
stranded power tool extension cord wire cut to length. It comes in
pretty colors and handles nicely and is cheap.)

As for "golden ears", I find that if I'm equipment shopping, when I
carry a couple records (it's been *that long*), I can go system to
system or store to store and hear different things, but for me it's a
skill that needs to be practiced. Once I've bought some kit I just
enjoy it and cease to listen diagnostically. I've never done AB tests.

I have blood relatives that have perfect pitch and can identify
specific musical instruments on classical recordings.

I dropped my Golden Ear magazine subscriptions years ago. I once told
the editor for one of them that I thought that spending the money on
*msic* was mutch more satisfying. He wasn't happy.






--
a d y k e s @ p a n i x . c o m

Don't blame me. I voted for Gore.