quality ethernet cable?

Posted by: ridds on 13 June 2011

Hi to all at the forum. I am trying to set my new uniti up & was just wandering if the quality of ethernet cables matter at all when i purchase one to go direct to my nas drive?.   Any replys would be very helpfull.cheers.

Posted on: 13 June 2011 by aysil

Hi ridds,

This depends on how picky you want to be about sq. A couple of months ago, most would say no to your question and even ridicule you for asking; but recently, Simon found out even ferrite clamps around the ethernet cable would improve the sound. So, I would recommend not to go for the worst quality cable.

Posted on: 13 June 2011 by James L

Belden is what I'd opt for. 

 

And use the new cat-6 which has better specs (crosstalk etc) over the cat-5e.

 

 

Posted on: 14 June 2011 by Nagoya

This audio LAN CAT6a cable from Furutech costs around 150 pounds in Japan for 5 metres.


Furutech Cat6a LAN cable


The RJ45 connector is a non-magnetic body using a 24-carat gold-plated copper alloy. Speed is 10Gbps and the transmission bandwidth is 500MHz.I have no idea if it makes any difference, and I certainly won't be buying. But I'd thought I'd post in case someone has a little money to burn. Thanks!

Posted on: 14 June 2011 by Simon-in-Suffolk

Posted on: 14 June 2011 by lhau
Wow, I would hate it if that plastic connector lock breaks!!! Btw what is a ferrite clamp, how does it look like, how much is it and where can I find it?!
Posted on: 14 June 2011 by ridds

Thanks for your help. cheers.Ridds.

Posted on: 14 June 2011 by gmischol

Google Meicord Opal, costs around 70 Euros. I ordered and haven't regretted the choice.

Posted on: 14 June 2011 by Simon-in-Suffolk

Ridds, Lhau and everyone!! The ferrite clamp is a standard way for us radio chaps to stop RF travelling along the outside of a cable, or with balanced/twisted pair cables such as CAT5 cables across all of them, and the  RF we are talking about is usually generated by electronics and not the BBC.

The chokes are cheap and if you find it makes a difference you can add more until no difference is made - or if no difference is heard then you take off..

(Its the same concept you see around the mains leads on Plasma TVs and the like - and here the choke is often moulded into the cable

http://computer.howstuffworks.com/question352.htm

 

And here are some sample prices - but typically you are talking less than a pound to a couple of quid.

http://www.google.co.uk/products/catalog?hl=en&q=ferrite+clamps&rlz=&um=1&ie=UTF-8&tbm=shop&cid=12119951310093029267&sa=X&ei=wb73TeveNofDhAe9guyeDA&ved=0CDcQ8wIwAQ#

 

Just check the diameter of the cable first for the clamp.

 

BTW a more expensive cable isn't neccessarily going to be better at rejecting EMI.

 

Simon

 

Posted on: 30 June 2011 by Rockingdoc

Following Simon's post, I decided to experiment, and fitted 10 ferrite chokes along the cable from my router to the Squeezebox Touch. The improvement in sound quality was obvious.

Now I know my system lives in an RFI dirty area 'cos my Superline never tires of telling me so, but even so the result was astonishing. Wish I'd tried it before. Next, I suppose, should be the cable from the NAS to the router.

Posted on: 30 June 2011 by meissmar
My dealer is recommending a cable from audioquest which is not too expensive and should "sound" very good. They also say that if you take a good switch, then it will not affect the sound much, but the cable should be clearly audible. I will give it a try sometime with a borrowed cable over the weekend.

He also says that a good quality cat5 will Be better than taking a cheap cat6.
Posted on: 01 July 2011 by sbilotta
Originally Posted by Rockingdoc:

Following Simon's post, I decided to experiment, and fitted 10 ferrite chokes along the cable from my router to the Squeezebox Touch. The improvement in sound quality was obvious.

Now I know my system lives in an RFI dirty area 'cos my Superline never tires of telling me so, but even so the result was astonishing. Wish I'd tried it before. Next, I suppose, should be the cable from the NAS to the router.

Is there a limit to the amount of ferrite chokes that you can apply before negative sonic effects occur ?

I know that with power cords, too many ferrite chokes will dry up the sound.

 

Many thanks

Stefano

Posted on: 01 July 2011 by Harry

Simon. Thank you for the link. The product displayed in the link - will it fit an ordinary CAT5 cable, mine come out at about 6mm diameter? I am wary of ending up with a load of chokes which are too big for the cable.

 

Thanks

Posted on: 01 July 2011 by Rockingdoc

My NAS+router and SB Touch are in separate rooms and the connecting Cat6 cable runs across the garage ceiling, so 10 chokes didn't seem too mad. This isn't the same as removing RFI from the mains, and I can't see any theoretical reason for degredation with more chokes, but science theory doesn't always play a large role in the hi-fi domain, e.g. directional cables.

 

p.s. 6mm chokes are ideal for standard Cat5/Cat6

Posted on: 01 July 2011 by Noogle

Are the ferrite beads preventing data corruption?  I thought Ethernet had error correction built in?

Posted on: 01 July 2011 by Harry

Thanks Doc.

Posted on: 01 July 2011 by Simon-in-Suffolk

Harry, Stefano, everyone..

The clamps in the link were for 10mm diameter cable so if your cat5 is 6mm it will be fine, you might want to put a blob of blu tak to hold it firm. Alternatively you can search for a 6mm clamp - there are are bound to be some somewhere.

There is no limit to the number of clamps you can put on, but add until no difference is heard.

If adding inductance to the outside of a mains lead dries up the sound, then I guess your have tuned your system to accommodate the RF distortion (albeit very slight) which can be heard as a brighter sharper sound which you might prefer - there is no absolute answer here. However the relationship of RF to the affects casued are not always straightforward. One example is RF contimainated into a source will travel along an interconnect and may reflect back across a interconnect or casue standing waves. This can casue distortion - albeit very slight in output amplifiers, expecially ones using negative feedback that effectively trying to compensate badly for the RF variation in the signal. Again this can be heard a sharpness or brightness that some may actually prefer.. it just I dont :-) - and also it becomes unpredicatable and you can start finding interconnects start having a major difference on the sound if RF contaminated, as well as digital interconnects, USB and mains leads for that matter.

 

Finally the RF on ethernet cables is common mode and has no relatioship to the physical signal levels between the signal wires, therefore has no bearing on the frames transmitted. For note at the ethernet level there is no error correction, errored frames are thrown away, and a higher level transport windowing system within TCP stacks of say the NDX and uPNP server will usually force a transport packet which is encapsulated by an Ethernet frame to be resent.

 

Simon

 

Posted on: 01 July 2011 by Noogle

A counter-effect of putting chokes on Ethernet cables would be to increase signal rise times and reduce noise immunity of the transceivers at each end of the cable.

Posted on: 01 July 2011 by Simon-in-Suffolk

Hi - which signal rise time are you referring to? Ethernet uses twisted pairs, so the inductance is 'cancelled' on the return twist. It would casue an issue if you just put indiuctance on one of the twists however by unravelling the cable. 

Simon

 

Posted on: 01 July 2011 by garyi
Originally Posted by Rockingdoc:

Following Simon's post, I decided to experiment, and fitted 10 ferrite chokes along the cable from my router to the Squeezebox Touch. The improvement in sound quality was obvious.

Now I know my system lives in an RFI dirty area 'cos my Superline never tires of telling me so, but even so the result was astonishing. Wish I'd tried it before. Next, I suppose, should be the cable from the NAS to the router.

Hahhaahhaa, your a good wizz Rockingdoc.

Posted on: 01 July 2011 by Noogle
Originally Posted by Simon-in-Suffolk:

Hi - which signal rise time are you referring to? Ethernet uses twisted pairs, so the inductance is 'cancelled' on the return twist. It would casue an issue if you just put indiuctance on one of the twists however by unravelling the cable. 

Simon

 

So doesn't having twisted pairs nullify the RFI effects?

Posted on: 01 July 2011 by Harry

That's great Simon. Many thanks.

Posted on: 01 July 2011 by Richwleeds
Anyone remember Peter Belt?
Posted on: 01 July 2011 by Noogle

He gave an invited talk at a company I once worked for.  Mad as a box of frogs.

 

Posted on: 01 July 2011 by Simon-in-Suffolk
Noogle, no because the RFI is common mode, ie carrying down the pairs in phase, and is relative to earth. You are thinking about the RF current carried as data signals within the Ethernet able, and yes the electro magnetic affects of the signal itself are cancelled out. This is why twisted pairs are used for telephone wires, Ethernet cables etc. Twisted pairs are often beneficial to shielded cable but require the signal to be balanced. If the signal is unbalanced you have to use shielded or coax cable.
These techniques are basic electronic/electrical engineering principles and certainly not unique to audio equipment.
Simon
Posted on: 01 July 2011 by Noogle

Does the RF corrupt the data or just create high frequency noise in the receiving equipment?