On 18 Sep 2003 16:14:04 GMT, Don Pearce wrote:

On 18 Sep 2003 14:36:22 GMT, (Stewart Pinkerton)
wrote:

On Wed, 17 Sep 2003 18:09:09 GMT, (Bob-Stanton)
wrote:

(Stewart Pinkerton) wrote in message

Goertz
MI has very low impedance, around 6-8 ohms, and very high capacitance,
while the classic spaced construction cable, 'balanced' FM antenna
feeder, has 300 ohm impedance and very high inductance. Agreed that
we're talking a lot less than milliHenries here.

Actually, if the Goertz cable (or any other) is terminated in it's
characteristic impedance, it will not look capacitive or inductive to
the amplifier. It's terminal impedance will be a pure resistance.

Yes, but that isn't ever going to happen with a loudspeaker, except at
a very few specific frequencies.

True, but using a decent 8-ohm cable even with a real speaker will
yield an impedance (forgetting the speaker itself) that is very, very
close to a resistive 8 ohms.

No, it won't. Anyone who has dabbled in radio will tell you that you
need at least tolerably close load matching to get anything like a
proper resistive transmission line. Now, since the amplifier certainly
won't have anywhere close to 8 ohms source impedance, and the speaker
won't be anywhere near to an 8 ohm resistive load over the vast
majority of its working range, you are in fact back to the lumped
capacitance model.

It certainly will not be the highly
inductive load of a traditional twin-flex speaker cable. Of course
some amplifiers need that additional inductive load to stay stable,
but that is another story.

A reality check will indicate that a few dozen microHenries (as you'd
get for ten feet of 'zipcord') hardly constitutes a 'highly inductive
load', even at 20kHz.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

On 19 Sep 2003 16:32:15 GMT, (Stewart Pinkerton)
wrote:

On 18 Sep 2003 16:14:04 GMT, Don Pearce wrote:

On 18 Sep 2003 14:36:22 GMT, (Stewart Pinkerton)
wrote:

On Wed, 17 Sep 2003 18:09:09 GMT, (Bob-Stanton)
wrote:

(Stewart Pinkerton) wrote in message

Goertz
MI has very low impedance, around 6-8 ohms, and very high capacitance,
while the classic spaced construction cable, 'balanced' FM antenna
feeder, has 300 ohm impedance and very high inductance. Agreed that
we're talking a lot less than milliHenries here.

Actually, if the Goertz cable (or any other) is terminated in it's
characteristic impedance, it will not look capacitive or inductive to
the amplifier. It's terminal impedance will be a pure resistance.

Yes, but that isn't ever going to happen with a loudspeaker, except at
a very few specific frequencies.

True, but using a decent 8-ohm cable even with a real speaker will
yield an impedance (forgetting the speaker itself) that is very, very
close to a resistive 8 ohms.

No, it won't. Anyone who has dabbled in radio will tell you that you
need at least tolerably close load matching to get anything like a
proper resistive transmission line. Now, since the amplifier certainly
won't have anywhere close to 8 ohms source impedance, and the speaker
won't be anywhere near to an 8 ohm resistive load over the vast
majority of its working range, you are in fact back to the lumped
capacitance model.

I do mean close. With the line being a minute fraction of a wavelength
long, You would be very hard pressed to measure the speaker alone as
any different to the speaker plus line (once the line delay has been
normalised) even at the frequencies where the speaker is quite some
way from 8 ohms. Normal speaker cables, around a couple of hundred
ohms, can make a measurable difference - a fact attested to by the
instability of some amplifiers using them

What I am really trying to do here is dispel the myth that such cables
(like the Goertz) are in some way capacitive when driving a speaker.
The fact is they aren't.

It certainly will not be the highly
inductive load of a traditional twin-flex speaker cable. Of course
some amplifiers need that additional inductive load to stay stable,
but that is another story.

A reality check will indicate that a few dozen microHenries (as you'd
get for ten feet of 'zipcord') hardly constitutes a 'highly inductive
load', even at 20kHz.

Inductive enough to pull a poor amplifier back from the brink of
oscillation, though.

d

_____________________________

http://www.pearce.uk.com

On 19 Sep 2003 17:53:15 GMT, Don Pearce wrote:

On 19 Sep 2003 16:32:15 GMT, (Stewart Pinkerton)
wrote:

Anyone who has dabbled in radio will tell you that you
need at least tolerably close load matching to get anything like a
proper resistive transmission line. Now, since the amplifier certainly
won't have anywhere close to 8 ohms source impedance, and the speaker
won't be anywhere near to an 8 ohm resistive load over the vast
majority of its working range, you are in fact back to the lumped
capacitance model.

I do mean close. With the line being a minute fraction of a wavelength
long, You would be very hard pressed to measure the speaker alone as
any different to the speaker plus line (once the line delay has been
normalised) even at the frequencies where the speaker is quite some
way from 8 ohms.

Well yes, due to the 'line' being a very small fraction of a
wavelength, as you say.

Normal speaker cables, around a couple of hundred
ohms, can make a measurable difference - a fact attested to by the
instability of some amplifiers using them

That is due to the series inductance of the cable, not the
characteristic impedance per se. Stick a milliHenry inductor on the
front of some 8-ohm Goertz MI, and you get the same result.

What I am really trying to do here is dispel the myth that such cables
(like the Goertz) are in some way capacitive when driving a speaker.
The fact is they aren't.

The fact is, they are. Try hooking an original Naim NAP250 to some
Goertz MI, and watch it fry!

It certainly will not be the highly
inductive load of a traditional twin-flex speaker cable. Of course
some amplifiers need that additional inductive load to stay stable,
but that is another story.

A reality check will indicate that a few dozen microHenries (as you'd
get for ten feet of 'zipcord') hardly constitutes a 'highly inductive
load', even at 20kHz.

Inductive enough to pull a poor amplifier back from the brink of
oscillation, though.

Yes, as noted above.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

(Stewart Pinkerton) wrote in message ...
On 18 Sep 2003 16:14:04 GMT, Don Pearce wrote:

On 18 Sep 2003 14:36:22 GMT, (Stewart Pinkerton)
wrote:

On Wed, 17 Sep 2003 18:09:09 GMT, (Bob-Stanton)
wrote:

(Stewart Pinkerton) wrote in message



No, it won't. Anyone who has dabbled in radio will tell you that you
need at least tolerably close load matching to get anything like a
proper resistive transmission line.

That is because at RF, the cable is usually many wavelengths long. A
cable will act as an impedance transformer, if the load is other than
the cable's characteristic impedance.

As for audio, a 20 ft long cable is only 406 E-6 wavelengths long (at
20 kHz). This is far too short to cause an impedance transformation.
Therefore, the impedance seen by the amplifier will be nearly the same
as the impedance of the load (loudspeaker). A cable will add a tiny
amount of resistance and inductance, but that will usually be
insignificant.

Now, since the amplifier certainly
won't have anywhere close to 8 ohms source impedance, and the speaker
won't be anywhere near to an 8 ohm resistive load over the vast
majority of its working range, you are in fact back to the lumped
capacitance model.


Trying to analyze a transmission line, (even a short one like an audio
cable) by using lumped constants, will always result in wrong answers.

Bob Stanton