IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Cables orientation, RMAA results
Pio2001
post Sep 13 2004, 01:21
Post #1


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



There is a french audiophile forum where someone foolishly asked if cables had an orientation.
You know, do they sound different if you switch the source side with the
destination side. http://www.homecinema-fr.com/forum/viewtopic.php?t=29744347

21 pages and 3 monthes later, the discussion went on possible measurments.
So I did some RMAA measurments between the two possible ways of plugging a
line in-line out coaxial cable. I wouldn't talk about it here if I hadn't got some
significantly different results, according to the orientation of the cable.

The cable is an RG179bu, silver plated copper, 75 Ohms. Hand soldered with 2%
silver alloy, on basic teflon / golden plugs.

Picture of the opened plug.



The cable is covered with white tape, so as to preserve silver from light, since the
cable is a bit transparent. The audiophile myth about cable orientaton says that a
cable has normally no orientation, but slowly gets one as and when it is used. It
would be a burn-in phenomenon. This one was plugged in one direction for one
year, and reversed three days ago. It worked for several hours during these three
days. This orientation is the one called "Bon" (=good).

With RMAA 5.4, ( http://audio.rightmark.org/index_new.shtml ), I generated a 44100 Hz 16
bits test file. I burned it on CD, together with the calibration signal.
The CD player is a Yamaha CDX-860 from 1991. It plays CDR bit perfect. The
CDR was scanned by KProbe with a JLMS XJ-HD165H drive at full speed. The
maximum C1 error rate was 14.
The CD Players plays the CD. The cable is plugged between the fixed line output
of the CD player and the line input of the Sony DTC55ES DAT deck, that works in
48000 Hz 16 bits. The analog input level is set so that the -1 db RMAA calibration
signal is at -1 db. The digital output of the DAT deck is recorded bit-exact by
SoundForge into a 48 kHz 16 bits wav with a Marian Marc 2 souncard.

In order to evaluate the accuracy of the measurments, I repeated 10 times each recording.





We can see that there is one measurment that gives different results according to
the orientation of the cable : intermodulation distortion.

Here are the complete results for measurments 1 and 2 (I added the swept
frequencies with FrontPage).

http://perso.numericable.fr/~laguill2/sens...omparison01.htm
http://perso.numericable.fr/~laguill2/sens...omparison02.htm

We can see on all graphs a difference between the "bon" and the "mauvais"
orientation (Good / Bad) : in the good direction, odd harmonics of 50 Hz (150,
250, 350 etc until 950 Hz) are higher.
It means that in the good orientaton, the cable gets a 50 Hz signal, quite squared
(fundamental + odd harmonics), in a stronger way than in the bad orientation.
Here are zooms on the 1 and 2 intermodulations. The Intermodulation Distortion
(IMD) are the 6040 Hz and 7060 Hz peaks. The cursor displays the 7060 Hz values.





I don't know how the total is calculated, but the difference between the two
measurments is of the order of magnitude of 1 db. That here stands, if we convert
in %, for 1 significant digit. Thus the differences in the tables, that affect the
second significant digit, are too small to appear on these graphs. But we can zoom
for example the graph one :



And there is something weird here : the IMD is clearly inferior in the good
orientation than in the bad. Both at 6040 Hz and 7060 Hz. I don't see why RMAA
says it is the opposite ! It is probable that the IMD values in the table don't come
from the real IMD peaks, but from all the neighbourous noise that is visible
around. I don't know where these 20 Hz spaced peaks come from.

Whatever, there is a measured difference between the two pluggings.
Three factors have changed between them :

-The cable position (can be closer or farther to power supplies, or AC cords)
-The contact quality (possible additional fingerprints on the plugs, plugs more or less well plugged)
-The cable orientation.

The 20 recording were made with the cables in quite similar positions. Let's test
the influence of the cable position, trying different ones :
Position 1 : same as previously
Position 2 : The cables are inserted between the CD player and the tape deck.
Position 3 : the cables are pulled away from the machines (but there is still an AC
cord near them).



We can see that the position has no effect on the IMD value.

I have made two last recordings after that plugs are cleaned with 90° alcohol.
Still the bad orientation, position 1, then good orientation again.



This time the good orientation value is very close to the bad one. It doesn't match
the results previously got in the right orientation.

It seems thus that the main factor is the contact quality of the plugs.
In order to test this hypothesis, I'll have to make 20 other recordings, but
alterning the orientation, so as to see if the result varies randomly (because of the
quality of the contacts), or alternately (because of the orientation of the cable).
Now that I forgot which was the good and the bad orientations, I'm going to let the
cable burn in a little in a given orientation wink.gif
Go to the top of the page
+Quote Post
Axon
post Sep 13 2004, 01:55
Post #2





Group: Members (Donating)
Posts: 1985
Joined: 4-January 04
From: Austin, TX
Member No.: 10933



Actually I've never really heard of burn-in affecting the "directionality" of cables - I hear directionality a lot more with respect to shielding. It's considered customary practice with some manufacturers to not connect the ground/shield on one connector, allegedly to reduce ground loop issues, as the ground loop can no longer travel along the shield. Amazingly, some people have found some theoretical basis for this design in some RF books (the references escape me now). However disconnecting the shield effectively turns it into a low-grade radio antenna, and the potential to drive EM noise back into either the source or the destination component (depending on the orientation) becomes much greater.

Are you sure that the shield is connected on both sides?
Go to the top of the page
+Quote Post
Axon
post Sep 13 2004, 01:59
Post #3





Group: Members (Donating)
Posts: 1985
Joined: 4-January 04
From: Austin, TX
Member No.: 10933



Also note that there is a noise effect, mostly prevalent on microphone cables hooked up to high-gain preamps, where simply moving the cable around produces a significant voltage due to changing dielectric values in the insulator. This usually isn't an issue for interconnects, but if you're getting accuracies down to literally parts per million for IMD, it could certainly be something that mucks with things. As a control test, I'd definitely try bending the cable by a significant amount.
Go to the top of the page
+Quote Post
Pio2001
post Sep 13 2004, 02:43
Post #4


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



QUOTE (Axon @ Sep 13 2004, 02:55 AM)
Are you sure that the shield is connected on both sides?
*


Yes, sure. I just checked first with an ohm-meter (the cable is just coaxial), then visually, opening every plug.

QUOTE (Axon @ Sep 13 2004, 02:59 AM)
simply moving the cable around produces a significant voltage due to changing dielectric values in the insulator.
*


I know that moving the cable can produce a voltage. I've experienced it with a broked tape deck, moving the bare cable that used to lead to the playback head. I could actually hear my finger hit the cable in the line output ! I thought that it was the effect of the induced current caused by the acceleration of the cable in the earth magnetic field.

QUOTE (Axon @ Sep 13 2004, 02:59 AM)
I'd definitely try bending the cable by a significant amount.
*


You mean to apply a physical permanent stress the dielectric ? I'll try this next time.
Go to the top of the page
+Quote Post
Omion
post Sep 13 2004, 02:47
Post #5





Group: Developer
Posts: 432
Joined: 22-February 04
From: San Diego, CA
Member No.: 12180



What's the audiophile's explanation of this phenomenon? That the electrons will get bored with one orientation, and desire a change?

It's probably on the French thread Pio linked to, but I know absolutely no French.


--------------------
"We demand rigidly defined areas of doubt and uncertainty!" - Vroomfondel, H2G2
Go to the top of the page
+Quote Post
Pio2001
post Sep 13 2004, 02:53
Post #6


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



There is no technical audiophile explanation. Some talk about the direction in which the original copper threads are manufactured, but it doesn't fit with the burn-in hypothesis.
There are no blind tests that I know of either.
All that I've got is consistently different RMAA results when I unplug and replug my cable, everything else equal, and I'm trying to figure out why.
Go to the top of the page
+Quote Post
Yodule
post Sep 13 2004, 03:22
Post #7





Group: Members
Posts: 6
Joined: 28-May 03
Member No.: 6861



QUOTE (Axon @ Sep 12 2004, 08:55 PM)
Actually I've never really heard of burn-in affecting the "directionality" of cables - I hear directionality a lot more with respect to shielding. It's considered customary practice with some manufacturers to not connect the ground/shield on one connector, allegedly to reduce ground loop issues, as the ground loop can no longer travel along the shield. Amazingly, some people have found some theoretical basis for this design in some RF books (the references escape me now). However disconnecting the shield effectively turns it into a low-grade radio antenna, and the potential to drive EM noise back into either the source or the destination component (depending on the orientation) becomes much greater.

Are you sure that the shield is connected on both sides?
*


Normally, this kind of cable is shield + twisted pair : the shield is connected to one side only, one member of the twisted pair is the signal and the other member is the ground.

See for example http://www.jensen-transformers.com/as/as001.pdf
Go to the top of the page
+Quote Post
Axon
post Sep 13 2004, 03:36
Post #8





Group: Members (Donating)
Posts: 1985
Joined: 4-January 04
From: Austin, TX
Member No.: 10933



QUOTE (Pio2001 @ Sep 12 2004, 08:43 PM)
You mean to apply a physical permanent stress the dielectric ? I'll try this next time.
*

It doesn't need to be permanent - if the effect is there with just an orientation change, it should be OK to just put an extra twist or two in the cable before you plug it in.
Go to the top of the page
+Quote Post
Pio2001
post Sep 13 2004, 03:48
Post #9


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



Yes, I meant that it's the stress presence that is supposed to have an effect, not the stress modification, isn't it ?
Go to the top of the page
+Quote Post
Axon
post Sep 13 2004, 04:20
Post #10





Group: Members (Donating)
Posts: 1985
Joined: 4-January 04
From: Austin, TX
Member No.: 10933



QUOTE (Pio2001 @ Sep 12 2004, 09:48 PM)
Yes, I meant that it's the stress presence that is supposed to have an effect, not the stress modification, isn't it ?
*

Actually I think the theoretical effect is supposed to be based on a change in stress, not merely a different stress. Potentially if you stress the cable a bit and leave it alone, some internal slippage may occur for the next several minutes/hours as the bends in the cable become slightly more permanent. However, if this really were the observed effect, it would manifest itself as a large change for the first test, with the effect decaying over time.
Go to the top of the page
+Quote Post
boojum
post Sep 13 2004, 05:41
Post #11





Group: Members (Donating)
Posts: 819
Joined: 8-November 02
From: Astoria, OR
Member No.: 3727



Pio, I appreciate the exhaustive Lycée influenced search for the truth. But as a real skeptic I would like to see repeated tests of numerous cables before I believe that electrons pass one way through cuprous metals better than the other. cool.gif

Prends la Foy!


--------------------
Nov schmoz kapop.
Go to the top of the page
+Quote Post
Pio2001
post Sep 13 2004, 18:22
Post #12


Moderator


Group: Super Moderator
Posts: 3936
Joined: 29-September 01
Member No.: 73



For the time being, the measurement don't show that the cable orientation is the cause of the difference. I must test further.
As for other cables, I let people try with theirs. This way not only other cables would be tested, but also in different circumstances.
In the french thread, someone suggested that if the orientation was the cause, it might come from a difference in the plugs : in in one orientation the source plugs are loose while the destination ones are very tight, and if in the opposite direction all plugs are just right, the latter orientation might give better results.
Go to the top of the page
+Quote Post
Garf
post Sep 13 2004, 18:35
Post #13


Server Admin


Group: Admin
Posts: 4886
Joined: 24-September 01
Member No.: 13



I skimmed rather briefly through this, but my question is, isn't it possible that such a miniscule difference is, for example, a temperature rise in one end/connector, for example because you held it in your hand.
From what I see the difference is so small with so many other potential variables, that we can't say anything from this.

And is it me or did the crosstalk "improve" by wiring it the wrong way around? wink.gif

This post has been edited by Garf: Sep 13 2004, 18:36
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 21st October 2014 - 17:20