AudioQuest Ethernet Cables

Eletrical - yes, but I have 2 problems with wi-fi streaming to my ND5XS:

1. I don't like the idea of that little antenna transmitting RF in the back of the rack, just where all the cables are...

2. It adds another layer of processing. A bit like the difference between coax and fiber optic for S/PDIF.

I listened to both and I think direct network connection sounded better, but I can't be sure, because I have a psychological problem with it (see 1. above) which must color my perception 

Please create own PSU/Isolation topic & do not mix it into AQ or ethernet cables.

Ah, off-topic Police?

It's only human to go from one topic to another, than return... In fact I noticed that the more intelligent the individuals, the more they jump between topics. You should see what's happening at MENSA meetings - impossible to keep the conversation to the point for longer than 1 minute

Yes quite agree. Every little helps. 

Originally Posted by 40 below:

It was from a well known international supplier of specialist power supplies

 

 

So who is it??????

I suspect it's an Israeli gentleman - Eddie Bardot or something similar 

The experience I have had with Audioquest ethernet (Pearl) and USB (Forest, Cinnamon, and Carbon) lead me to believe that this company should always be on the short list for demos.  

In respect of a linear PSUs, search fairly recent forum posts PSUs for the unitiserve.....

You will not ever fully understand the real performance of your streamer before you change all the internal wires in the router to AQ vodka, and that comes from my five years experience in the matter.

An AudioQuest Cinnamon 1.5m Ethernet cable will be winging it's way to me shortly to replace whatever I have plugged into the back of my NDX. What I am anticipating is that it will make absolutely no difference, as the bits are bits logic makes a lot of sense. However, after reading all the positive comments on here, I though I'd give one a shot for myself! Worst case scenario is I will sell it on, and it's not a great outlay in this context.

Tog

That's exactly the approach I took. It'll be interesting to hear what you think of it.

Hi Allen, Ethernet and Fast Ethernet noise is rather well understood in the land of digital hardware design but does seem to cause some mystique in hifi circles.

Anyway essentially an Ethernet connection will exhibit to some extent common mode interference. That is unwanted energy present across the 4 differential pairs in the Ethernet lead. A LAN isolator works by usually providing an effective low pass filter that attenuates the high frequency common mode interference away from the chassis.

This interference will couple into the receiving equipment earth and chassis.

There are mitigations but almost difficult if not impossible to reduce completely

This white paper by Texas Instruments is a good guide and what interference is caused by Ethernet leads and what design considerations can be implemented to reduce noise contamination.

Because Ethernet can be so relatively noisy in terms of EMI and EMI coupling, its not at all suprising that  different Ethernet leads and their interfaces can 'tune' the interference and therefore 'sound' different. But you don't need IMO to use expensive boutique  Ethernet leads to modify the EMI.

http://www.ti.com/lit/an/snla107a/snla107a.pdf

PS This has noting to do with the (digital) data being encoded across the differential pairs - which I think is where some people not familiar with the engineering pricniples get confused.

Regards

Simon

Great question Gajdzin I was wondering the same thing.

Gajdzin, it is the link nearest the audio equipment that matters.

i'd say best keep clear of cheap switches unless you are using isolators, but a) should be fine, but do watch out for cheap noisy SMPSs and try and keep switches/routers/Ethernet patch leads well away from your hifi leads.

BTW switches deal in frames not packets, unless you have an advanced managed switch they are oblivious to packets. They will rebuild the network frame and send it out on one or more switch ports.

finally I recommend commercial grade Cat5e (as opposed to Cat 5) which is really the recommended cable for fast Ethernet.

Simon

Good advice, my house is wired in CAT5 because that was all the rage 10 years ago, when my internet link was 10mbps (its 120mbps now and I'm upgrading to 250mbps next month)

But assuming I buy an inexpensive switch and a separate inline (not switching mode) power supply for it, do you think it makes sense to buy 75cm of the best AQ cable I can afford for that last link to ND5XS, rather than no switch and 3 or 4 meters of a cheaper AQ cable?

Probably a very dumb question(!) but would the same advice around only the nearest link being important apply in the case of a junction box in the wall?

Do not buy cheapest possible switch e.g. 20€, if you spend 60€ you can get a proper more reliable metal case HP etc switch.  My friend just returned cheap zyxel router/switch due it need reboots, limits some ports pass through etc.

It is the last connection what counts, so 75cm recommendation from here. In my case AG Vodka made a clear difference, equal to HiLine. Sound is more warm, dynamic & instead of highlighting independent instruments whole presentation comes more natural. I do understand if someone likes just normal cat6 more. Of cource change depends quite much from system e.g. Speakers could limit. 

I currently use 2 TP-Link switches, an 8-port DGS-108 (30 Euro) and a 4-port version of the same (it was 23 Euro or so). All-metal chassis, good brand, very good reviews on the internet.

I could add to that a decent in-line power supply for ca. 18 Euro in place of the SMPS that comes with the D-Link and voila: I can buy the shortest possible run of the highest quality AQ ethernet cable I can afford... That's my original idea at least, still waiting for someone to bomb it as wrong, who knows

Why is metal chassis important?

Less RFI, and, if the chassis is ferromagnetic, probably less EMI that Allen was writing about. The cable can be shielded, but what of it if the box radiates like crazy and it's next to the streamer (as in my scenario, where we spend the money on the shortest and top quality cable we can afford rather than a longer run of lower quality cable). The D-Link gigabit switch I mentioned earlier also comes in a pretty plastic enclosure (DGS-1008d), all other parameters are the same. I had that and replaced it with the metal version (DGS-108).

If I owned an NDS I know I would 'upgrade' all cables. Peace of mind would be paramount over 'can you hear it' after such a hefty investment IMO.

Gajdzin, I think are in the land of trial and error here. I prefer not to use very short Ethernet patch leads when I am concerned about noise - unless I am using a switch that senses the patch lead length and lowers the drive voltage. I use a 8m run of form a Cisco 2960 switch to my NDX. The final 3 metres are shielded. The other 5 metres are flat Ethernet cable that goes under the carpet and through a door frame. I simply connect the two cables together.

Allen - let us know how you get on with the wireless. I would be surprised if it sounded better - as the NDS is working harder with quite noisy electronics to drive the wifi - but you never know until you try as you say.

Simon

I agree, Simon Still, the cheap D-Link switches in metal enclosures do in fact adjust their power consumption to the cable length. And I found a good source of in-line, transformer-based regulated power supplies in wall-wart format for about 7 Euro. But then again I may not bother and just to be safe buy 3 meters of Cinnamon (distance from ND5XS to the nearest router) rather than 0,75m of Vodka + the extra switch... Still in doubt.

The reason I'm experimenting with these things is I'm not 100% happy with the SQ I get when streaming my ripped CDs from NAS to ND5XS. The soundstage is smaller and less euphonic than straight from the CD5XS. Who knows, maybe it's using cheap CAT5 cable throughout the system, all the way to ND?