If you have a Naim NDS what Ethernet wire do you use?

sjbabbey posted:

Mike, I have a total of 3 Netgear switches in my network. Do you think it's worthwhile upgrading the SMPS for all of them or simply for the connection to the NDS streamer.

Hard to say, depends on so many things, e.g. if they are on the same power circuit & the distance between them,  might be worth a punt on the NDS switch SMPS & see how that goes.

It was probably me that got the forum interested in the iFi iPower SMPS so I carry a bit of baggage with that.  I found it cleaned up the sound quite nicely,  not a huge change but it was audible.  BUT (yes its a but)  I need to come clean,  the audible SQ change was when I compared it side by side with the original Netgear SMPS & a MediGrade SMPS I was also using at the time when they were all on a trailing socket plugged into the wall.   My normal set up is a fixed 4 way socket thats not so easy to get to (hence the trailing socket test) but it's not an ordinary 4 way socket;  its fed from my UPS & that has an internal isolation transformer & a common mode choke,  plus I have the fixed 4 way socket cable, the UPS supply cable & each of the SMPS (phone, BB & Netgear) 12v output cables are all overloaded with ferrite clamps.   Swapping the iPower with the MediGrade & original Netgear on the ferrite loaded fixed 4 way did not bring so much of a change in SQ,  it was very subtle & still in favour of the iPower.  In all honesty although I did keep the iPower in my set up it was not really worth it,  the ferrite & the UPS isolation collectively suppress SMPS noise to a level that is hard to fault & it sorta killed the real need for the iPower. One reason I did keep the iPower was its very impressive voltage stability,  it was rock steady with both no load & a load that was just under the iPowers claimed full watt rating.   The Netgear & the MediGrade did not get close to the same stability,  but the Netgear was the better of the two (that surprised me). 

I heard a subtle but perceptible improvement with the iFi IPower on my Netgear switch but I am not sure if I did this substitution before or after I put ferrite chokes on SMPS power leads and on (both ends) of my ethernet cables. As it is relatively cheap compared to some of the ethernet cables mentioned here, the risk of not getting VFM is low IMHO.

One thing that is worth doing if you can is to physically and electrically separate the LAN gear (i.e. router, switch, NAS/Server/Unitiserve) from your Naim black boxes. I have my LAN gear on their own rack shut away in an airy cupboard, plugged in to a different ring main to the one my Naim black boxes are plugged in to. The two racks of equipment are on opposite sides of the room. This is all in an attempt to minimise the noise from the LAN gear and their SMPSs finding its way into the delicate electronics of the Naim streaming source and onwards.

Warning: This behaviour is probably a mixture of sensible precaution and obsessive hifi neurosis.

There are some interesting tips here - thanks chaps.

i might give the Netgear GS305 switch a go as I have the older version with the metal shroud. It's dead cheap and somehow seems a more elegant solution than a cheap connector on the end of the Vodka. After that I might try the iFi iPower, presumably, it's the 5 volt version that's needed?

Keith

The bigger Cisco 8 port switch recommended early in this thread for sound quality, is too big for where I can put it. Anyone recommend a good sounding switcher that is smaller but still sound good! I am using an inexpensive net gear one right now! 5 ports. 

nigelb posted:

I heard a subtle but perceptible improvement with the iFi IPower on my Netgear switch but I am not sure if I did this substitution before or after I put ferrite chokes on SMPS power leads and on (both ends) of my ethernet cables. As it is relatively cheap compared to some of the ethernet cables mentioned here, the risk of not getting VFM is low IMHO.

One thing that is worth doing if you can is to physically and electrically separate the LAN gear (i.e. router, switch, NAS/Server/Unitiserve) from your Naim black boxes. I have my LAN gear on their own rack shut away in an airy cupboard, plugged in to a different ring main to the one my Naim black boxes are plugged in to. The two racks of equipment are on opposite sides of the room. This is all in an attempt to minimise the noise from the LAN gear and their SMPSs finding its way into the delicate electronics of the Naim streaming source and onwards.

Warning: This behaviour is probably a mixture of sensible precaution and obsessive hifi neurosis.

Just the neurosis Nigel I seem to be suffering from the same though. Running my metwork equipment on a separate and dedicated circuit, with a UPS which acts as a mains and noise filter (in and out). Upgraded power supply for a NAS too.

Ahh - the joys of listening to some music

sjbabbey posted:

Mike, I have a total of 3 Netgear switches in my network. Do you think it's worthwhile upgrading the SMPS for all of them or simply for the connection to the NDS streamer.

In my experience the SMPS of the modern GS switches are better than earlier models - but I really don't think the key issue is so called 'RFI' and screening etc especially if you use chokes and keep wiring separated out -  unless the PSU is  noisy and modulating the clocks ... its more to do in my opinion with the jitter in the carrier clocks of the switch ports and accuracy of the chipsets with respect to queuing and timing.. Many consumer switches use similar Broadcom chipsets and the like. Higher grade devices can use custom ASICs  and that coupled with improved clock stability seems to help with SQ with respect to coupled intermodulation - just like using a higher spec  SPDIF transport.

Give it a try and see what you think - many would not use a cheap transport to feed their  high quality DAC? A switch (as opposed to a network hub)  is like a transport...

Simon

Adam Zielinski posted:

nigelb posted:

I heard a subtle but perceptible improvement with the iFi IPower on my Netgear switch but I am not sure if I did this substitution before or after I put ferrite chokes on SMPS power leads and on (both ends) of my ethernet cables. As it is relatively cheap compared to some of the ethernet cables mentioned here, the risk of not getting VFM is low IMHO.

One thing that is worth doing if you can is to physically and electrically separate the LAN gear (i.e. router, switch, NAS/Server/Unitiserve) from your Naim black boxes. I have my LAN gear on their own rack shut away in an airy cupboard, plugged in to a different ring main to the one my Naim black boxes are plugged in to. The two racks of equipment are on opposite sides of the room. This is all in an attempt to minimise the noise from the LAN gear and their SMPSs finding its way into the delicate electronics of the Naim streaming source and onwards.

Warning: This behaviour is probably a mixture of sensible precaution and obsessive hifi neurosis.

Just the neurosis Nigel I seem to be suffering from the same though. Running my metwork equipment on a separate and dedicated circuit, with a UPS which acts as a mains and noise filter (in and out). Upgraded power supply for a NAS too.

Ahh - the joys of listening to some music

 

Yes indeed. I keep telling myself to spend less time obsessing and more time listening to music. Ah, talking to oneself cannot be a good sign either! 

KRM,  Correct GS305 is 5V  .............  lets have a post on any up/downs if you go that way.  

Simon-in-Suffolk posted:

Many consumer switches use similar Broadcom chipsets and the like. Higher grade devices can use custom ASICs  and that coupled with improved clock stability seems to help with SQ with respect to coupled intermodulation

Give it a try and see what you think - many would not use a cheap transport to feed their  high quality DAC? A switch (as opposed to a network hub)  is like a transport..

This is probably the block stone Simon,   many people are more than happy to use their wireless router/hub as a switch & seem happy with whatever the router does even when they are not actually managing the data stream around the network.  The more 'aware' (searching the right word) insist on a pukka switch because of better data stream address mngt,  store-&-forward, buffer etc., this standard of switch seems to satisfy the vast majority, most of Naim it seems & from what I can tell 99.9% of Linn & other make users.   The problem I see is you get pretty  well most all the population more than satisfied with Broadcom (& similar) equipped switches at less than £20 (5 port),   but to move to the switches you are suggesting is x10+ times.   The sonic gains may be just as you say,  but without the required test equipment to verify this, most of us are left with our Mk-II audio receptors & the OS thats located in-between (in most cases).  Then the "whats the best" switch discussion falls back into the same mixing pot we have with "whats the best" ethernet cable.    I personally am very interested in progressing my system switch quality but am not willing to move into the cost of the switches you are advocating & I am sure many Naim'ites are the same, even if they have already paid a fortune for some boutique russian spirit soaked ethernet cables.   If however you do have a suggestion for something in the 5 port unmanaged category that fits your criteria  I am always very open & willing to give it a try.

musicfan51 posted:

The bigger Cisco 8 port switch recommended early in this thread for sound quality, is too big for where I can put it. Anyone recommend a good sounding switcher that is smaller but still sound good! I am using an inexpensive net gear one right now! 5 ports. 

..........  the GS series from v4 is probably as good as a 5 port unmanaged switch as is available,  but I expect the difference in other makes in the same class is unmeasurably miniscule.

Would a Netgear GS605NA be better than my 6 year old Netgear switcher ? 

GraemeH posted:

Adam Zielinski posted:

Hungryhalibut posted:

Adam Zielinski posted:

NAS to switch: Chord C-Stream

Switch to wall sockets: standard CAT 6 cables

Wall socket to NDS: Chord Indigo Tuned ARAY streaming cable

Like most here I started with regular CAT 6 cables and tested various options - each one was an improvement. Hence the final selection.

Crumbs, that Indigo cable is expensive. Looking at the Chord website, it's interesting that it has cheap looking plastic connectors, whereas the AQ Vodka for example has very solid metal connectors. I wonder why this is - is the connector irrelevant and it's just the wire that makes the difference?

It does seem to be an oposit to the Audioquest Vodka, doesn't it? I thinik it's the quality of the cables and the shielding / screening that is critical here.

Chord seem to call this a 'streaming cable' as opposed to a regular 'network cable'. I don't know what the technical difference is, but it does sound different and marvelous.

My dealer was kind enough to loan me 4 or 5 different cables from the Chord range. I liked 3 of them: C-Stream, Signature and Indigo. 
A funny story: when I actually asked my dealer about the streaming cables, he looked at me as if I have finally and completely lost my mind. But something must have intrigued him - next time I popped in, he had a range of them on a rack, and his streamers were hooked up with at least C-Stream, Usually it's the dealer that pushes the envelope... 

...or smells an opportunity to tap a new (gullible - who, us?) market more like!

G

Fancy ethernet cables are truly the tin-foil hat/snake oil region of the hobby..

 

Fancy ethernet cables are truly the tin-foil hat/snake oil region of the hobby..

+1. I would think so too.

I see no harm in trying to see if improves the sound! Many say speaker wires don't make difference( they do) so I think I will try some Ethernet cables out and see what happens ! I think when you have a great revealing system it all matters! 

Dupree / Banzai

You make very sweeping statements, which I presume are firmly grounded in extensive tests you must have conducted. Would you mind sharing with the fellow Forum Members which streaming cables you have tested, under what conditions and what your observations were?

Thanks,

Adam

DUPREE posted:

.....

Take a step back, does anyone think a 10 grand ethernet cable, that meets the same spec as a 8 dollar one going to make the same difference as going from a NDX to NDS or other source improvement.

> On Sep 29, 2016, at 6:53 AM, Naim Audio Forums <alerts@hoop.la> wrote:
>

That depends if those specs are relevant for the mechanism that affects sound quality.

Since 'we' don't exactly know why there might be a difference, these current specs are most likely 'missing the point'.

it is known (by me from empirical evidence, by professional recording and production studios from technical knowledge) that LAN isolators that do proper galvanic isolation, do actually make a noticeable difference. So the regularly heared statement that 'ethernet is galvanically isolated by design' doesn't fly.

Mike-B posted:

Simon-in-Suffolk posted:

Many consumer switches use similar Broadcom chipsets and the like. Higher grade devices can use custom ASICs  and that coupled with improved clock stability seems to help with SQ with respect to coupled intermodulation

Give it a try and see what you think - many would not use a cheap transport to feed their  high quality DAC? A switch (as opposed to a network hub)  is like a transport..

This is probably the block stone Simon,   many people are more than happy to use their wireless router/hub as a switch & seem happy with whatever the router does even when they are not actually managing the data stream around the network.  The more 'aware' (searching the right word) insist on a pukka switch because of better data stream address mngt,  store-&-forward, buffer etc., this standard of switch seems to satisfy the vast majority, most of Naim it seems & from what I can tell 99.9% of Linn & other make users.   The problem I see is you get pretty  well most all the population more than satisfied with Broadcom (& similar) equipped switches at less than £20 (5 port),   but to move to the switches you are suggesting is x10+ times.   The sonic gains may be just as you say,  but without the required test equipment to verify this, most of us are left with our Mk-II audio receptors & the OS thats located in-between (in most cases).  Then the "whats the best" switch discussion falls back into the same mixing pot we have with "whats the best" ethernet cable.    I personally am very interested in progressing my system switch quality but am not willing to move into the cost of the switches you are advocating & I am sure many Naim'ites are the same, even if they have already paid a fortune for some boutique russian spirit soaked ethernet cables.   If however you do have a suggestion for something in the 5 port unmanaged category that fits your criteria  I am always very open & willing to give it a try.

 

Mike, that is fine, many for example don't want the cost of going to a Melco for example for streaming and switching either.. for subtle refinement... there is nothing wrong with that.. but if you go to higher quality hardware in the audio system replay chain invariably it improves the overeall result..whether you think it's worth it or not is in the ear of the beholder. But if I am honest I scratch my head at someone buying a cheap consumer switch and connecting with an expensive Ethernet cable... it would to my mind be like connecting a Naim DC1 lead between a Sonos transport and the Naim DAC for example....

Consumer network equipment is built cheaply and typically has undemanding roles..I know of no custom Consumer switch or router switch port device.. as they all tend to use similar chipsets.. I can suggest however getting a secondhand Cisco 2960 catalyst switch with a basic firmware... these seem to be more affordable and appear on auction sites. If you have an IT friend you might be able to borrow one? In the professional world they are incredibly popular... other than that crack on with as you are ..... of course you might not hear a difference at all...

Aleg posted:

DUPREE posted:

.....

Take a step back, does anyone think a 10 grand ethernet cable, that meets the same spec as a 8 dollar one going to make the same difference as going from a NDX to NDS or other source improvement.

> On Sep 29, 2016, at 6:53 AM, Naim Audio Forums <alerts@hoop.la> wrote:
>

That depends if those specs are relevant for the mechanism that affects sound quality.

Since 'we' don't exactly know why there might be a difference, these current specs are most likely 'missing the point'.

it is known (by me from empirical evidence, by professional recording and production studios from technical knowledge) that LAN isolators that do proper galvanic isolation, do actually make a noticeable difference. So the regularly heared statement that 'ethernet is galvanically isolated by design' doesn't fly.

Aleg, I am afraid you might have been misinformed. By design and the the Ethernet physical standards, the physical switch ports are and have to be galvanically isolated.  Without looking up I can't remember the DC voltage standard Ethernet has to galvanically isolated by..but there also exists enhanced isolation for medical equipment which is isolated to a higher DC voltage for safety reasons.

in low noise environments fibre is sometimes used... and you might be referring to low pass noise filters on your electrical connections.. nothing potentially wrong as long as designed correctly but I have had poor results with such devices and they can upset the link synchronisation protocols for some equipment...but that is not the same as Galvanic isolation.

Simon

There was a discussion ongoing on the Chord Sarum ethernet cables a couple of years ago. The ASA jumped in and has forbidden certain advertisements because of ungrounded claims.

Since I came to know this, I use the Chord Sarum to test hifi dealers. Lets be positive and assume a quality difference of 1% (believe in, not proven - whatever 'quality' might be), some hifi dealers unashamedly tell me that these cables are miles better and my setup should deserve one.

I buy somewhere else, can't trust the guy.

I would vote for buying more expensive red wine, it has higher value for money than ethernet cables. It makes me happy and I believe in it - that's in the end the same reason why people buy these cables.

I'm trying out some Audioquest Cinnamon Ethernet cables and seeing if I hear a difference. They are returnable!  I want to hear for myself, that is always the best way.  Thanks to everyone for their input! I appreciate it. 

DUPREE posted:

As a hobby there is no harm in trying anything. Some may say standing on your head makes things sound better. Speaker cables while I believe to be an overstated impact; do have a measurable impact as they are part of the complete circuit. Ethernet is a transformer coupled digital transport; not only is the benefit or drawback of any in spec Cat6a cable is handled in the spec. Even if there were a difference, it would because of this totally unpredictable - the cheapest or most expensive cable might due to expectation bias sound the best. Everyone should buy what makes them happy, but be aware what has some merit on empirical basis. There are companies like audio quest who make a great living off of selling snake-oil for obscene money. Take a step back, does anyone think a 10 grand ethernet cable, that meets the same spec as a 8 dollar one going to make the same difference as going from a NDX to NDS or other source improvement.

> On Sep 29, 2016, at 6:53 AM, Naim Audio Forums <alerts@hoop.la> wrote:
>

There's nothing wrong with a healthy dose of scepticism but I'd suggest not writing something off as snake oil until you've tried it for yourself. I remember when someone on here (it may have been Harry) bought some Vodka cables and I was very negative about them, saying that they simply could not make a difference and that the owner was deluded. Then I realised that I hadn't tried them and therefore I had no sound basis on which to comment on them. The only way was to try some for myself so I bought a Cinnamon for £65 and had to eat my words. I now have Vodka which is much better again. 

My take on these cables, and any expensive cables come to that, is that they are about system optimisation, the icing on the cake if you like. Of course they will not turn an NDX into an NDS. But they will help whatever setup you have to perform at its best, just like posh speaker cables and interconnects. 

So I'd say to the cynics, try them yourselves, and then come back on here and tell us what you've discovered. You may find that you can't hear a difference, or don't like the difference. But at least you will know, rather than simply thinking that you know. 

Aleg - I looked it up - as I am in that sort of mood...

IEEE 802.3 40.60.1.1 100BaseT Isolation requirements: (This feeds into various safety regulation requirements)

But if anyone is interested in professional audio and networking or just audio engineering and research in general  - its is a pretty interesting area with some great activity and developments - you might want to consider joining the AES and join the community.

musicfan51 posted:

I see no harm in trying to see if improves the sound! Many say speaker wires don't make difference( they do) so I think I will try some Ethernet cables out and see what happens ! I think when you have a great revealing system it all matters! 

Just over a year ago I tested Cat5e, Audioquest Cinnamon, Meicord , Chord C-Stream and Indigo cables between NAS and Router: 

"The Cinnamon was better than a bog-standard Cat5. A Meicord was better still, though the bass was a bit bloated. A C-Stream had all the benefits of the Meicord but with a balanced bass, then the Chord Indigo was a bit of a wow moment. Very expensive but as I only needed 1.55m I was unable to resist."

I subsequently blind tested a friend and my long-suffering wife and the Indigo won hands down with them.

Edit: Just to say I went into this exercise rather on the side of bits are bits and would hear nothing different.

I'm sure people do find positive differences. Why don't the makers of these boutique Ethernet cables take one of the Network players that seem to work so much better with one of these cables in place and measure what is happening when they replace a bog standard Ethernet cable with the expensive version and publish the results to back up their claims. If it's changing the sound to such an extent, then it must be measurable and correlated to the SQ changes noted. 

Chord say - 

A lot of the work we’ve done over the past decade has been about reducing electrical interference to a minimum, the reason being that where typically twenty years ago the average house was relatively free of interference sources, it isn’t any more.  High frequency effective shields are important, particularly if you wish to hear music in the way in which you were meant to.  

Just saying this without some meaningful measurement of how effective this shielding is and its subsequent effect on the player seems a bit pointless and one of the reasons that so many people see a lot of these products as just snake oil for the gullible.