[sci.electronics] INTERCONNECTS

bks@oz.berkeley.edu.UUCP (04/22/87)

i don't want to restart any old arguments, and therefore request that
any follow-ups to this article include ``INTERCONNECTS'' in the
subject, so the uninterested may easily avoid the follow-up.

in the beginning, i was skeptical about the claims made about
interconnects, but when i switched from an integrated amplifier to a
separate preamplifier, i built myself a pair of interconnects using
switchcraft plugs, and the monster interlink 4 stuff.  the completed
cables cost about $9 all tolled, but i had to do some work to get the
cable through the jack housing, and burning the insulation off the
wires was a bit troublesome.

anyway, i thought i could hear a difference between this interconnect,
and the ``comes with the component'' (regular) kind of interconnect,
but not really enough to be certain it wasn't psychosomatic.

yesterday, because i planned to do some rewiring in my preamp to solve
a hum problem, i also replaced the cable which connects my record
player to the preamp.  the difference which resulted from this change
is quite obvious; it now seems that the original cable was a low-pass
filter.  for instance, cymbal kinds of sound have more detail to them;
continuo harsichord attacks are much sharper; overall dynamic range at
least seems to be superior.

the cartridge is a grado 8mr; this connects to what appear to be solid copper
wires (no litz stuff, this is a stock ar arm).  the copper wire meets the
monster stuff at a terminal strip.  the other end of the monster stuff goes
to a plug, inserted in a jack, which connects to a 47K metal film resistor
via twisted 22 gauge teflon-insulated wire.

i would assume that standard measurements (resistance, impedance) between
regular and fancier interconnects do not differ, else we wouldn't argue about
these things.  does anyone have any evidence to the contrary?  has anyone
performed any measurements to take into account the level and impedance of
the source, which might be more meaningful measurements?


				brian

ark@alice.UUCP (04/23/87)

In article <18511@ucbvax.BERKELEY.EDU>, bks@oz.berkeley.edu.UUCP writes:
-> yesterday, because i planned to do some rewiring in my preamp to solve
-> a hum problem, i also replaced the cable which connects my record
-> player to the preamp.  the difference which resulted from this change
-> is quite obvious; it now seems that the original cable was a low-pass
-> filter.  for instance, cymbal kinds of sound have more detail to them;
-> continuo harsichord attacks are much sharper; overall dynamic range at
-> least seems to be superior.

There is no doubt that cables can make a substantial differenct
BETWEEN THE PHONO CARTRIDGE AND THE PREAMP.  You are dealing,
after all, with a 47K impedance on one end and a reactive
source at the other.  Cable capacitance is the crucial factor.

ross@raster.UUCP (04/23/87)

In article <18511@ucbvax.BERKELEY.EDU> bks@oz.berkeley.edu (Brian K. Shiratsuki) writes:

>anyway, i thought i could hear a difference between this interconnect,
>and the ``comes with the component'' (regular) kind of interconnect,
>but not really enough to be certain it wasn't psychosomatic.
...
>i would assume that standard measurements (resistance, impedance) between
>regular and fancier interconnects do not differ, else we wouldn't argue about
>these things.  does anyone have any evidence to the contrary?  has anyone
>performed any measurements to take into account the level and impedance of
>the source, which might be more meaningful measurements?
>
>				brian

I did some playing around with trying to measure differences between cables.
Using a function generator, a 'scope, and an HP gain-phase meter (shows
amplitude and phase differences between the two inputs), I connected the
function generator to one input of the HP, and the cable under test between
both inputs.  I also varied the terminating resistance, using 1k, 10k,
47k, and 100k resistors at the non-driven end of the cable.  I tried the
following cables:  El Cheapo from Radio Shack, Monster Cable Interlink 2,
Monster Cable Interlink 4, and "hardline" - essentially coax where the
shield is a solid pipe (not braid) and is used for interconnects at
gigahertz frequencies.  

The results?  NO MEASURABLE DIFFERENCES FROM DC TO 250K Hz.  The function
generator wouldn't go any higher - hey, it was cheap, and all I could
scrounge.

Now I have claimed to be able to hear a difference.  I couldn't measure
any difference, though it _was_ a quick and dirty test and I didn't
vary the output impedance of the source.  Anyone have any suggestions
for a more thorough test?

ross

markf@amc.UUCP (Mark Freeman) (04/24/87)

> In article <18511@ucbvax.BERKELEY.EDU> bks@oz.berkeley.edu (Brian K. Shiratsuki) writes:
> 
> ...
> Using a function generator, a 'scope, and an HP gain-phase meter (shows
> amplitude and phase differences between the two inputs), I connected the
> function generator to one input of the HP, and the cable under test between
> both inputs.  I also varied the terminating resistance, using 1k, 10k,
> 47k, and 100k resistors at the non-driven end of the cable.  I tried the
> ...
> ross


Don't just set the terminating resistance -- also set the appropriate source
impedance.  It might be difficult to match the characteristics of the cartrige.
A better test would be to use the cartrige and a test record.

A simpler test, however, should point out the most significant difference:
measure the capacitance!

-- 
				Mark S. Freeman
				Applied Microsystems, Inc.
				markf@amc

sjc@mips.UUCP (Steve Correll) (04/24/87)

In article <18511@ucbvax.BERKELEY.EDU>, bks@oz.berkeley.edu (Brian K. Shiratsuki) writes:
> ...i also replaced the cable which connects my record player to the
> preamp.  the difference which resulted from this change is quite obvious;
> it now seems that the original cable was a low-pass filter...
> 
> i would assume that standard measurements (resistance, impedance) between
> regular and fancier interconnects do not differ...

Some magnetic cartridges are fairly sensitive to load capacitance, and
cable capacitance does vary. (So does preamp input-stage capacitance.
A few years ago, the audio press noticed--apparently for the first
time--that you could make a cartridge's response vary by several dB at
15kHz by switching cables and preamps, and for a while it became
fashionable for preamps to provide user-adjustable input capacitance.
Rather than establish a standard, the audio industry has decided to
forget the problem and focus on CD dampers, turntable mats, litz-wound
speaker cables, etc. Excuse me; I editorialize.)

Amplifier and preamp designers are supposed to provide sufficiently low
source (output) impedance to tolerate variation in load capacitance. (The
EIA standard for testing audio preamps requires you to load the outputs
with capacitance to simulate cables.) Thus, changing cables at the
output of a good preamp seems unlikely to make an audible difference.

Cartridge designers have no such obligation.

> Using a function generator, a 'scope, and an HP gain-phase meter (shows
> amplitude and phase differences between the two inputs), I connected the
> function generator to one input of the HP, and the cable under test between
> both inputs.  I also varied the terminating resistance...
> The results?  NO MEASURABLE DIFFERENCES FROM DC TO 250K Hz...I didn't
> vary the output impedance of the source.  Anyone have any suggestions
> for a more thorough test?

Your experiment demonstrates nicely why the cable at the output of a
preamp doesn't matter.  What you're dealing with is a voltage
divider: the source impedance is on top and the load impedance (in
parallel with the cable capacitance) is on the bottom.  If the source
impedance is close to zero, as it is for a function generator (or audio
preamp), the value of the load capacitance has very little effect on
the voltage across the load.

If you repeat the experiment with a higher (and not purely resistive) source
impedance, perhaps by using a real phonograph cartridge playing a test record,
you may well see differences among cables.

-- 
...decwrl!mips!sjc						Steve Correll

bks@oz.berkeley.edu (Brian K. Shiratsuki) (04/24/87)

In article <196@raster.UUCP# ross@raster.UUCP (Ross Werner) writes:
#In article <18511@ucbvax.BERKELEY.EDU# bks@oz.berkeley.edu (Brian K. Shiratsuki) writes:
#
##anyway, i thought i could hear a difference between this interconnect,
##and the ``comes with the component'' (regular) kind of interconnect,
##but not really enough to be certain it wasn't psychosomatic.
#...
##i would assume that standard measurements (resistance, impedance) between
##regular and fancier interconnects do not differ, else we wouldn't argue about
##these things.  does anyone have any evidence to the contrary?  has anyone
##performed any measurements to take into account the level and impedance of
##the source, which might be more meaningful measurements?
##
##				brian
#
#I did some playing around with trying to measure differences between cables.
#Using a function generator, a 'scope, and an HP gain-phase meter (shows
#amplitude and phase differences between the two inputs)...
#...I also varied the terminating resistance, using 1k, 10k,
#47k, and 100k resistors at the non-driven end of the cable.  I tried the
#following cables:  El Cheapo from Radio Shack, Monster Cable Interlink 2,
#Monster Cable Interlink 4, and "hardline" - essentially coax where the
#shield is a solid pipe (not braid) and is used for interconnects at
#gigahertz frequencies.  
#
#The results?  NO MEASURABLE DIFFERENCES FROM DC TO 250K Hz.  The function
#generator wouldn't go any higher - hey, it was cheap, and all I could
#scrounge.
#
#...Anyone have any suggestions for a more thorough test?
#
#ross

hm, i wish i had such equipment to play with.  the load is identical
enough, though i probably also have a measureable amount of (grid to
cathode) capacitance.  the frequency range is more than adequate, since
my preamp doesn't pass dc, and the audible difference seems to be well
below the upper limit of my hearing.  did you consider using a real
cartridge as a source, playing some sort of record with test tones?
also, you didn't mention the amplitude of the function generator, but
i assume it was around 1mV?

this result may be more be more applicable to the pre to poweramp
interconnect (the audible difference here is much subtler, if it
exists).
Newsgroups: rec.audio,sci.electronics
Subject: Re: INTERCONNECTS
Expires: 
References: 
Sender: 
Reply-To: bks@oz.berkeley.edu (Brian K. Shiratsuki)
Followup-To: 
Distribution: 
Organization: University of California, Berkeley
Keywords: 

In article <196@raster.UUCP# ross@raster.UUCP (Ross Werner) writes:
#In article <18511@ucbvax.BERKELEY.EDU# bks@oz.berkeley.edu (Brian K. Shiratsuki) writes:
#
##anyway, i thought i could hear a difference between this interconnect,
##and the ``comes with the component'' (regular) kind of interconnect,
##but not really enough to be certain it wasn't psychosomatic.
#...
##i would assume that standard measurements (resistance, impedance) between
##regular and fancier interconnects do not differ, else we wouldn't argue about
##these things.  does anyone have any evidence to the contrary?  has anyone
##performed any measurements to take into account the level and impedance of
##the source, which might be more meaningful measurements?
##
##				brian
#
#I did some playing around with trying to measure differences between cables.
#Using a function generator, a 'scope, and an HP gain-phase meter (shows
#amplitude and phase differences between the two inputs)...
#...I also varied the terminating resistance, using 1k, 10k,
#47k, and 100k resistors at the non-driven end of the cable.  I tried the
#following cables:  El Cheapo from Radio Shack, Monster Cable Interlink 2,
#Monster Cable Interlink 4, and "hardline" - essentially coax where the
#shield is a solid pipe (not braid) and is used for interconnects at
#gigahertz frequencies.  
#
#The results?  NO MEASURABLE DIFFERENCES FROM DC TO 250K Hz.  The function
#generator wouldn't go any higher - hey, it was cheap, and all I could
#scrounge.
#
#...Anyone have any suggestions for a more thorough test?
#
#ross

hm, i wish i had such equipment to play with.  the load is identical
enough, though i probably also have a measureable amount of (grid to
cathode) capacitance.  the frequency range is more than adequate, since
my preamp doesn't pass dc, and the audible difference seems to be well
below the upper limit of my hearing.  did you consider using a real
cartridge as a source, playing some sort of record with test tones?
also, you didn't mention the amplitude of the function generator, but
i assume it was around 1mV?

this result may be more be more applicable to the pre to poweramp
interconnect (the audible difference here is much subtler, if it
exists).

				brian

grr@cbmvax.cbm.UUCP (George Robbins) (04/26/87)

In article <18511@ucbvax.BERKELEY.EDU> bks@oz.berkeley.edu (Brian K. Shiratsuki) writes:
>
>yesterday, because i planned to do some rewiring in my preamp to solve
>a hum problem, i also replaced the cable which connects my record
>player to the preamp.  the difference which resulted from this change
>is quite obvious; it now seems that the original cable was a low-pass
>filter.  for instance, cymbal kinds of sound have more detail to them;
>continuo harsichord attacks are much sharper; overall dynamic range at
>least seems to be superior.

This is the most sensitive place to cable characteristics, since the
cartridge has a very low-level and non-ideal output.  You should be
using a short, "low capacitance" cable between the cartridge and the
pre-amp.  In addition, if you cartridge is sufficiently exotic, you
may benefit from changing a few values in the pre-amp input circuitry.

Whether the cable has to be vanilla lo-cap or high-dollar ultra-xxx
cable depends on your pocketbook and ego.

The other area to check is you speaker cables.  If they look like
miniature zip-cord, then go to the local hardware store and get some
heavy duty zip-cord.  That much will make a difference...

Do take note that clear, cylindrical insulation has a certain
tendency to make it look like "monster" cables are heavier than they
really are.
-- 
George Robbins - now working for,	uucp: {ihnp4|seismo|rutgers}!cbmvax!grr
but no way officially representing	arpa: cbmvax!grr@seismo.css.GOV
Commodore, Engineering Department	fone: 215-431-9255 (only by moonlite)

mat@mtx5a.UUCP (m.terribile) (04/29/87)

> I did some playing around with trying to measure differences between cables.
> Using a function generator, a 'scope, and an HP gain-phase meter (shows
> amplitude and phase differences between the two inputs), I connected the
> function generator to one input of the HP, and the cable under test between
> both inputs.  I also varied the terminating resistance, using 1k, 10k,
> 47k, and 100k resistors at the non-driven end of the cable.  I tried the
> following cables:  El Cheapo from Radio Shack, Monster Cable Interlink 2,
> Monster Cable Interlink 4, and "hardline" - essentially coax where the
> shield is a solid pipe (not braid) and is used for interconnects at
> gigahertz frequencies.  
> 
> The results?  NO MEASURABLE DIFFERENCES FROM DC TO 250K Hz.  The function
> generator wouldn't go any higher - hey, it was cheap, and all I could
> scrounge.
> 
> Now I have claimed to be able to hear a difference.  I couldn't measure
> any difference, though it _was_ a quick and dirty test and I didn't
> vary the output impedance of the source.  Anyone have any suggestions
> for a more thorough test?
> 
> ross

1) Have you tried it with a 60K source and 40K load?

2) What happens when you put 120 pf in parallel with the load?

3) What functions did you drive the thing with, and can the PGM pick
   up phase lag/lead at different frequencies?
-- 

	from Mole End			Mark Terribile
		(scrape .. dig )	mtx5b!mat
					(Please mail to mtx5b!mat, NOT mtx5a!
						mat, or to mtx5a!mtx5b!mat)
					(mtx5b!mole-end!mat will also reach me)
    ,..      .,,       ,,,   ..,***_*.

ross@raster.UUCP (05/01/87)

In article <1770@mtx5a.UUCP> mat@mtx5a.UUCP (m.terribile) writes:
>> I did some playing around with trying to measure differences between cables.
>> Using a function generator, a 'scope, and an HP gain-phase meter (shows
>> amplitude and phase differences between the two inputs), I connected the
>> function generator to one input of the HP, and the cable under test between
>> both inputs.   [ ... ]
>> 
>> The results?  NO MEASURABLE DIFFERENCES FROM DC TO 250K Hz.  [more]

>
>1) Have you tried it with a 60K source and 40K load?

No.  I was attempting to simulate the pre-amp to amp interconnection.  My
pre-amp (PS Audio 4.5) has a class A buffer amp (i.e. low output impedance)
in the high-level mode that drives the outputs.  In "straightwire" mode,
the buffer amp is switched out and so the outputs are driven essentially
by a voltage divider (the volume control) and the output impedance would
vary with the position of the volume control.  (I haven't been able to
notice a difference in sound between the two modes, though).

I didn't try to simulate the cartidge/cable interactions since it is 
obvious that there will be a difference - the loading has a very
noticeable effect on cartidges.

>2) What happens when you put 120 pf in parallel with the load?

Have to try this.  But I would expect that there would be a roll
off due to the cap, not the cable.  Hmm....

>3) What functions did you drive the thing with, and can the PGM pick
>   up phase lag/lead at different frequencies?

Sine, square, triangle waves.  The HP has several frequency ranges - 
1 Hz to 1K Hz, 10 to 10K, 100 to 100K, etc. up to the many MHz.  It
automatically displays the gain from A to B in db and the phase shift
from A to B in degrees.

>	from Mole End			Mark Terribile

Thanks for the comments.  I'll let you know when I have a chance to
play some more.

Ross

larson@sri-unix.UUCP (05/07/87)

Recent discussions have considered why different cables will have
different sound when testing with a common drive seems to get the
same results at the other end.

One thing that was not clear, was whether the cables were both
connected to the source at the same time, and switched at the
load end, or whether they were fully swapped.  If they were both
connected to the source during the A/B test, they will both be
loading the source all the time.

I personally consider the reactive load on the cartridge to be the
most likely effect.  Perhaps a suitable recording could be found to
allow measuring the cartridge output while it's load is artifically
varied.

	Alan