'Audiophile' Ethernet Cables.

[realhifi]

And now for something completely different...

 

What is the best way to install ethernet cable from one room to the next? My NAS drive can go in the office and I can either snake the cable through the heating duct work or through the power conduate where it parallels 120V AC Romex. My walls are atypical...non load bearing since the exposed beams are inside the house. The walls consist of styrofoam flanked by 3/4" plywood. The Romex runs through a pre-cut channel. I am tempted to run it through the latter, since I bet mice will chew up anything in the ductwork. I guess I could get a cat, but if it chewed up $$ cables, I will skin it.

 

Running parallel to 120 is a no no. Running it in cold air returns is fine and usually a heating duct is the last resort. Just be sure to silicon around the holes you drill for it when done. Heating ducts do not get hot enough to affect the cable.

[Jsmith]

Paul, I appreciate your thoughts.

 

The styrofoam is solid. There is no way. I pull the trim and run it along the floor then encase the cable in new trim along the support pillars of the house to keep it covered. I was going to run cheaper bulk Audioquest ethernet cable and spice them with the kit they provide.

 

If you have any other thoughts, let me know.

 

The Belden bonded cable is very good and you can buy reels for big runs. Just a thought.

[Mihnea]

And now for something completely different...

 

What is the best way to install ethernet cable from one room to the next? My NAS drive can go in the office and I can either snake the cable through the heating duct work or through the power conduate where it parallels 120V AC Romex. My walls are atypical...non load bearing since the exposed beams are inside the house. The walls consist of styrofoam flanked by 3/4" plywood. The Romex runs through a pre-cut channel. I am tempted to run it through the latter, since I bet mice will chew up anything in the ductwork. I guess I could get a cat, but if it chewed up $$ cables, I will skin it.

 

As far as I know you need to run Ethernet and power at least 10cm (4") apart in order to minimize interference. Of course, this was true for UTP (unshielded) Ethernet cables. If you use STP or FTP (shielded), you could run them even closer. Just make sure you have a good grounding at each end of the cable runs to really benefit from shielding. I saw several network cards broken due to using STP cords and not having an earthed power outlet.

[Paul R]

Assuming you want to put the NAS in your office to use with other equipment and to control noise, and you are using a Vortexbox player, basically forget the Ethernet cables and use either wireless or powerline technology. The Vortexbox will work fine that way so long as you take a little care setting it up.

 

As others have mentioned, you could run it in the heating ducts, but I would not do that. It isn't because the cables can get too hot, that is very unlikely. It is because the cable can provide a pathway for fire to spread in your home. Non plenum cable can be worse, generating noxious fumes and such during a fire event. And it may not be legal in your local area.

 

If you are absolutely opposed to wireless or other technology, hire an electrical contractor to run the wires and create a new cutout in the wall for them. You can also surface run the cables, using paintable channel.

 

Now this is assuming you are talking a fairly long run. If you are talking a meter of duct, no more, or duct open to both rooms, these concerns are not all that important.

 

Best- Paul

 

 

 

 

Paul, I appreciate your thoughts.

 

The styrofoam is solid. There is no way. I pull the trim and run it along the floor then encase the cable in new trim along the support pillars of the house to keep it covered. I was going to run cheaper bulk Audioquest ethernet cable and spice them with the kit they provide.

 

If you have any other thoughts, let me know.

[realhifi]

Powerline sucks. Heck, use wireless and leapfrog with AirPort Express's to extend your network. It is superior to Powerline. We have covered three story homes with basements in this way and have always been able to get great signals throughout the home. Of course ethernet is the best but if you are faced with crazy steps to get there I'd go wireless.

[Audio_ELF]

As far as I know you need to run Ethernet and power at least 10cm (4") apart in order to minimize interference.

Well maybe code says you should run Ethernet separate from power; but its not required for interference. In the UK it's common to run both in the same conduit around offices and there's no problem with interference. Of course UK is 50Hz 240v vs 60Hz 120v in USA so maybe that makes a difference.

 

Eloise

[sandyk]

Seeing that even desk top fluoros can induce noise into nearby 2 pair twisted telephone cables, why on earth would you want to do that ? Just because the Objectivists of the forum think that things like that don't matter, doesn't mean that others won't notice a difference.These codes are there for a reason, and have been long before the U.K's Postmaster General's Department became B.T. Australia followed the U.K. lead with Installation Circular No.4 more than 60 years ago with

minimum specified clearances between power cables and twisted pair telephone cables, even to requiring telephone cables to cross power cables at an angle of 90 degrees.

Alex

[Jud]

Well maybe code says you should run Ethernet separate from power; but its not required for interference. In the UK it's common to run both in the same conduit around offices and there's no problem with interference. Of course UK is 50Hz 240v vs 60Hz 120v in USA so maybe that makes a difference.

 

Eloise

 

Right, but question is, interference for what purpose? Halting communication via the very robust Ethernet standards? Or inducing noise in the cable short of stopping communication, but not something you'd like to hear coming from the speakers of your nice audio system?

[Audio_ELF]

Seeing that even desk top fluoros can induce noise into nearby 2 pair twisted telephone cables, why on earth would you want to do that ? Just because the Objectivists of the forum think that things like that don't matter, doesn't mean that others won't notice a difference.These codes are there for a reason, and have been long before the U.K's Postmaster General's Department became B.T. Australia followed the U.K. lead with Installation Circular No.4 more than 60 years ago with minimum specified clearances between power cables and twisted pair telephone cables, even to requiring telephone cables to cross power cables at an angle of 90 degrees.

Alex

Twisted pair telephone cables carrying analogue telephone signal are not the same as twisted pair cables carrying Ethernet signal

 

Right, but question is, interference for what purpose? Halting communication via the very robust Ethernet standards? Or inducing noise in the cable short of stopping communication, but not something you'd like to hear coming from the speakers of your nice audio system?

But isn't the question how can any interference on the Ethernet cable get into the audio system?

 

Oh well, your milage may vary and you can spend your money where you like.

 

Eloise

 

PS. are you saying that in USA (and Australia) you don't use multi-use trunking around offices?

[sandyk]

Twisted pair telephone cables carrying analogue telephone signal are not the same as twisted pair cables carrying Ethernet signal

 

I didn't say they were.I am simply saying that what you are suggesting is not good practice, especially since the mains these days are far more polluted than back when those old standards evolved.

[Audio_ELF]

I didn't say they were.I am simply saying that what you are suggesting is not good practice, especially since the mains these days are far more polluted than back when those old standards evolved.

We'll have to agree to disagree...

[Paul R]

I'm not sure about code, but I am sure that it's not a great idea if you want a fast pretty much trouble free building network plant.

 

I've seen really weird things happen when UTP was laid parallel, and even had odd things happen with STP. Even more so with high speed (100BaseT, 1000BaseT, and faster) networks. Stupid, difficult to explain things, like a Gigabit port running only at 10mbs for no apparent reason.

 

Moved the cable away from an overhead fluorescent light fixture and the port jumped to a highly reliable gigabit speed.

 

I have no idea what running UTP parallel to power cables might do to audio - perhaps absolutely nothing - but it definitely can have strange and unwanted effects on networks.

 

Yours,

-Paul

 

 

 

Well maybe code says you should run Ethernet separate from power; but its not required for interference. In the UK it's common to run both in the same conduit around offices and there's no problem with interference. Of course UK is 50Hz 240v vs 60Hz 120v in USA so maybe that makes a difference.

 

Eloise

[Jud]

But isn't the question how can any interference on the Ethernet cable get into the audio system?

 

Oh well, your milage may vary and you can spend your money where you like.

 

Eloise

 

 

Hi, Eloise. My thinking is that what we are dealing with is an entire system that is electrically interconnected, and that electrical noise, RFI, etc., can get in from any point and then get to any other part of the system (analog amplification, DAC circuitry, etc.), often through the ground plane. People spend hundreds or thousands on power conditioning, but for some reason (perhaps it is that word "digital") think nothing of noise coming in through digital cabling, which is electrically connected and thus not magically immune to transmitting electrical noise. People like to mention that the Ethernet packets get transmitted just the same. Certainly they do. But the power comes through the mains just the same even if noise is transmitted along with it, yet for some reason people are concerned about noise through power cables but not through the Ethernet cabling.

 

Is this a terrible problem that people should be investing heavily in trying to solve? My sense is, probably not. It just puzzles me to see folks who have no problem taking extreme measures to keep noise from other sources out of their systems sort of assuming on principle that Ethernet cables are somehow not subject to the laws of electricity because the electrical impulses they transfer represent digital packets.

[Audio_ELF]

I'm not sure about code, but I am sure that it's not a great idea if you want a fast pretty much trouble free building network plant.

As I say - it's pretty typical for offices (and even warehouse and factory installations) in UK to have a run of trunking with power and UTP Cat 5e in about 10cm high trunking, the two types separated by just a thin plastic wall. In such offices there's not issues with running gigabit speeds. This is practical experience rather than anything based on theory.

 

I will give you though that these are not audio systems...

 

Eloise

[Audio_ELF]

Is this a terrible problem that people should be investing heavily in trying to solve? My sense is, probably not. It just puzzles me to see folks who have no problem taking extreme measures to keep noise from other sources out of their systems sort of assuming on principle that Ethernet cables are somehow not subject to the laws of electricity because the electrical impulses they transfer represent digital packets.

I am relying on the isolation that Ethernet has designed into it as working. Yes they are subject to the laws of electricity - but they are also isolated systems. If you are concerned add hospital grade isolation.

 

I just think the money is better spend on other "problems".

 

Eloise

[Jud]

I am relying on the isolation that Ethernet has designed into it as working. Yes they are subject to the laws of electricity - but they are also isolated systems.

 

Sure, but virtually every other piece of the system is designed to be isolated as well through grounding, noise rejection circuitry, etc. Just about nothing is designed to let in every bit of electrical grunge. Yet people who are obsessive about keeping noise out of the rest of the system even though the components are designed to try to reject it don't feel that way about Ethernet cables or other digital components or cabling.

 

I just think the money is better spend on other "problems".

 

Eloise

 

I agree, and think I've gone on long enough (maybe too long, eh?) about what is really just the principle that one should ignore nothing, so I'll shut it now.

[Paul R]

Ah- that thin plastic wall actually does make a difference! Now I understand what you mean, and agree with your experience. I would personally separate them further, running two conduits, but that's arguable - some believe it would be overkill.

 

It's not all that uncommon to do what you describe anywhere I have been. It's when the cables are physically touching that you get the most problems. I have found installations that I had to redesign completely because Romex, Phone, Network, Coax for Cable TV, Broadband, 3270, and 5250- and even Fibre cables(!) - were zipcorded together! Yikes...

 

-Paul

 

 

As I say - it's pretty typical for offices (and even warehouse and factory installations) in UK to have a run of trunking with power and UTP Cat 5e in about 10cm high trunking, the two types separated by just a thin plastic wall. In such offices there's not issues with running gigabit speeds. This is practical experience rather than anything based on theory.

 

I will give you though that these are not audio systems...

 

Eloise

[VandyMan]

what is really just the principle that one should ignore nothing

 

Ignore everything, except the music. You will be much happier than worrying about noise on an Ethernet cable. ;-)

[Jud]

Ignore everything, except the music. You will be much happier than worrying about noise on an Ethernet cable. ;-)

 

And these widely varying points of view led us both to Vandersteens. :-)

[Mihnea]

The quality of Ethernet cables does matter if we want them to act as per standard and to do that after several years of use. Crappy cables (wires and connectors) can even fail a simple category test, or can do that after flexing them a bit.

 

The question is how far should one go with cable quality so that Ethernet standard is respected? Not that far IMHO. What works for computers (with no errors) will work for audio as well.

 

If using shielded cables, as I mentioned before, you need a good quality grounding or else you even risk harming one or both network adapters. The grounding at both ends needs to be measured by a qualified electrician and no "significant" (no idea what this should translate to) difference should exist between them (same potential).

 

If the above is OK, then you know external interference will not affect signal in transit. Of course, if noise does affect internally the source, then data sent will be already affected even if no additional harm will occur in transit to destination. At least this is my understanding.

[Bromo33333]

Galvanic isolation isn't always perfect - but most attempts at it are far preferable than not having it with noisy, nasty, signals from computer and computer equipment.

 

I have found much to my cost, that by the time you can hear it distinctly, it is well above the level at which it has influenced the sound due to EMI. Shielding is your friend!

 

WOuld I go and buy a bunch of kilobuck ethernet cables? Nope. But I sure would invest in shielded cables wherever possible. And I sure would select gear that was designed to reject that sort of noise. If I felt the Audioquest ethernet cables were good values and high quality, they'd be on the short list.

 

Galvanic isolation sure gets rid of the emi/rfi issue..... if your computer is decent and you are careful with keeping interference to a minimum... Ethernet cables make no difference...

[VandyMan]

I am afraid it's more to it than just listening to music. It's the sound. The qualities of the sound enhances listening experience. If I just wanted to hear the music I would go grab a $30 speakers with built-in audio in computer.

 

You are taking my statement rather more seriously than intended. Note the wink at the end. Or the equipment listed in my signature. My "point" was just a reminder to not get stuck in the nit-picky details and miss the big picture.

[reverendo]

I have just taken the time to read through the whole thread, since I had some time I could spend on my hands.

after a lot of theorizing I actually didn't see anyone actually give us subjective results regarding different ethernet cables.

since now I'm in a niche market, actually using a DAC that works with UPnP with an ethernet port, I'm thinking of testing different cables to see if they 'sound' different to me.

At the moment this is how my setup looks like:

Win7 64bit PC > foobar2000 (with UPnPserver by bublleguuum) > Netgear R4500 (router) > Netgear WNCE4004 (wireless adapter) > Resolution Audio Cantata.

I'm still comparing how my current transport compares to UPnP, but intend to test how different if I hear differences in ethernet cables. I'll focus only on the cable between the wireless adapter and the Cantata.

ATM I'm using this ethernet cable from Rosewill. I also have some Cat6e and Cat5e cables lying around here from Microcenter or that came with the wireless adapter. Maybe I'll also grab one or two of the AQ ethernet cables from a local Magnolia which I can return after checking them out.

Now, I must say in advance, I hear differences between USB cables (http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/usb-cable-comparisons-14397/), so if you don't believe that USB cables into a DAC make a sonic difference, then any opinions or 'findings' from me will probably have no value to you anyhow.

I sincerely hope that ethernet cables don't make a difference, but I'm willing to check for myself.

Best regards

[macbrush]

As a network professional, I can't say whether ethernet cable makes any sonic differences, but they do come in different quality especially their RJ45. Normally a properly wired connection shouldn't produce any errors, but once in awhile there would be random I/O errors, CRC errors, or resets, very occasionally even to a point which triggering spanning-tree change path, or drop link from an etherchannel.

 

Although we do test all of our cables before deploying by a simple cable tester, we also put our cables in load test for a week making sure there is absolutely no errors before placing them in any critical points.

 

Maybe you should test the cables thoroughly before putting them to audition to avoid any technical issues affecting the results.

 

Just a suggestion.

 

Cheers

Kenneth

 

I have just taken the time to read through the whole thread, since I had some time I could spend on my hands.

after a lot of theorizing I actually didn't see anyone actually give us subjective results regarding different ethernet cables.

since now I'm in a niche market, actually using a DAC that works with UPnP with an ethernet port, I'm thinking of testing different cables to see if they 'sound' different to me.

At the moment this is how my setup looks like:

Win7 64bit PC > foobar2000 (with UPnPserver by bublleguuum) > Netgear R4500 (router) > Netgear WNCE4004 (wireless adapter) > Resolution Audio Cantata.

I'm still comparing how my current transport compares to UPnP, but intend to test how different if I hear differences in ethernet cables. I'll focus only on the cable between the wireless adapter and the Cantata.

ATM I'm using this ethernet cable from Rosewill. I also have some Cat6e and Cat5e cables lying around here from Microcenter or that came with the wireless adapter. Maybe I'll also grab one or two of the AQ ethernet cables from a local Magnolia which I can return after checking them out.

Now, I must say in advance, I hear differences between USB cables (http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/usb-cable-comparisons-14397/), so if you don't believe that USB cables into a DAC make a sonic difference, then any opinions or 'findings' from me will probably have no value to you anyhow.

I sincerely hope that ethernet cables don't make a difference, but I'm willing to check for myself.

Best regards

[Oystein]

Although this thread is already a bit old I want to respond.

 

TCP/IP protocol is designed to send packets of data without any errors.

In a datacenter where the network load is close to 100% many re-transmits due to check-sum errors I can imagine this will bottle neck the data throughput especially for "on the fly" processes as music playback.

 

When streaming music the (wired) LAN load is so small a few data packets to be sent again will not be an issue.

Only if you have bad quality RJ-45 connectors the data sent can get lost and many re-transmits are the result of that.

You will most likely lose connection to your router but you will otherwise notice reduced performance getting files from your NAS.

There are enough test programs to test LAN data througput.

 

To comment on any claims of "better" cables with less errors just think of your cheap UTP cables at home.

Did you ever worry about the amount of money that gets transmitted to your bank when entering a booking with electronic banking ??

I didn't think so.

 

TCP/IP is 100% error free in our systems.

As long as the buffers in our playback software are set correct we will never hear any difference using audio grade UTP cables.