If you have a Naim NDS what Ethernet wire do you use?
Posted by: musicfan51 on 27 September 2016
Do you use Cat5 or Cat7 or one of the upper end Ethernet cables like from Audioquest or Chord? I know many say it makes no difference, though I have read articles that says with higher end systems (I would think a Naim NDS with 555DR would be considered high end) it does make a difference. Love to hear what everyone uses and your opinions. And does it matter on what brand of Ethernet port you use also? Thank you.
charlesphoto posted:Ok, just ordered a used Cisco 2960 switch for $35 to test vs my Netgear FS105 with linear power supply.
Simon in Suffolk: question is what do I do with it when it arrives? I've never used a managed switch before, so any tips on settings? This will be used only for the audio gear.
Assuming the seller has reset the switch to factory default.. you just plug it in and use it just like any other switch... nice price by the way
if they haven't it may be a little more involved to clear it out... cross that bridge if we have to
musicfan51 posted:Mike-B posted:charlesphoto posted:Ok, just ordered a used Cisco 2960 switch for $35 to test vs my Netgear FS105 with linear power supply.
Good move IMO, waiting with bated breath on your report. The Netgear FS105 that I used in a comparison test a while back was the only noticeably inferior SQ. The other switches were all Gigabit (GS) from Netgear, TP-Link & Linksys & we could not detect any difference.
I look forward to hear what you have to say about SQ after installing your new switcher!
Well, in the meantime I did a quick re-arrange of what I have. I forgot that I had a never used D-Link 5 port switch kicking about. Hooked that up next to the microRendu and powered off the 5v rail on the HDPLEX (mR on the adjustable rail set at 7v). 2m (set in wall) of AQ Cinnamon from the Vortexbox to the new switch, AQ Cinnamon from a patch on the wall (cat6) to new switch, and finally Meicord Opal to the rendu.
Better sound than the Netgear FS105 for sure. More relaxed bigger soundstage all that. It's subtle for sure, but there. Prefer the Meicord as final leg to rendu, but both are good (AQ a bit brighter and leaner). Not sure just where I'll implement the Cisco, but for $35 it should be a relatively pain free experiment. My other FS105 is now maxed, so might be interesting to see how running all of the other traffic cleaner helps.
That's the thing, if you're running gear as $$$ as NDS (I'm not) then spend a few penny and upgrade your cabling and networking all around I say. And no need for metals sourced from meteorites. But good enough to meet the level of one's gear/wallet for sure. It makes a difference for not a lot of, comparatively, money.
charlesphoto posted:musicfan51 posted:Mike-B posted:charlesphoto posted:Ok, just ordered a used Cisco 2960 switch for $35 to test vs my Netgear FS105 with linear power supply.
Good move IMO, waiting with bated breath on your report. The Netgear FS105 that I used in a comparison test a while back was the only noticeably inferior SQ. The other switches were all Gigabit (GS) from Netgear, TP-Link & Linksys & we could not detect any difference.
I look forward to hear what you have to say about SQ after installing your new switcher!
Well, in the meantime I did a quick re-arrange of what I have. I forgot that I had a never used D-Link 5 port switch kicking about. Hooked that up next to the microRendu and powered off the 5v rail on the HDPLEX (mR on the adjustable rail set at 7v). 2m (set in wall) of AQ Cinnamon from the Vortexbox to the new switch, AQ Cinnamon from a patch on the wall (cat6) to new switch, and finally Meicord Opal to the rendu.
Better sound than the Netgear FS105 for sure. More relaxed bigger soundstage all that. It's subtle for sure, but there. Prefer the Meicord as final leg to rendu, but both are good (AQ a bit brighter and leaner). Not sure just where I'll implement the Cisco, but for $35 it should be a relatively pain free experiment. My other FS105 is now maxed, so might be interesting to see how running all of the other traffic cleaner helps.
That's the thing, if you're running gear as $$$ as NDS (I'm not) then spend a few penny and upgrade your cabling and networking all around I say. And no need for metals sourced from meteorites. But good enough to meet the level of one's gear/wallet for sure. It makes a difference for not a lot of, comparatively, money.
I definitely want to get a new switch box! Mine is older cheap Netgear. I agree with you , and what I am saying is if you have a high resolution system it's not bad thing to look at all the cabling in your system in case you have a weak link. You may find an improvement along the way!
And the changes might not be gob smacking like a new set of speakers, or bigger amp might be, but more measured in terms of fatigue (or lack of) imo.
You've confirmed my findings that the FS105 is not a good switch for audio. Like I said in my previous post it was the only one that was obvious to all (4 of us) as the only switch of the 4 tested that was different, the others (all gigabit) had diddly squat difference, probably all using the same/similar Broadcom chipset. Is your D-Link a gigabit ?? I think this will show DGS 'something' if so.
Simon-in-Suffolk posted:charlesphoto posted:Ok, just ordered a used Cisco 2960 switch for $35 to test vs my Netgear FS105 with linear power supply.
Simon in Suffolk: question is what do I do with it when it arrives? I've never used a managed switch before, so any tips on settings? This will be used only for the audio gear.
Assuming the seller has reset the switch to factory default.. you just plug it in and use it just like any other switch... nice price by the way
if they haven't it may be a little more involved to clear it out... cross that bridge if we have to
Well, it says seller refurbished, from an IT clearance warehouse, with a short warranty, so theoretically it should be reset to default. Several on the market at this price; if it works out might invest in second to unload the rest of the network onto.
Mike-B posted:You've confirmed my findings that the FS105 is not a good switch for audio. Like I said in my previous post it was the only one that was obvious to all (4 of us) as the only switch of the 4 tested that was different, the others (all gigabit) had diddly squat difference, probably all using the same/similar Broadcom chipset. Is your D-Link a gigabit ?? I think this will show DGS 'something' if so.
Yes, its the DGS-105 which I believe is the latest. Metal case similar to the older netgear but appears just plastic on the inputs. Keep in mind it's also being run by clean power though an HDPLEX 100 into its own 20amp dedicated circuit. I still think its better despite that.
Re DGS-105 & my post "probably all using the same/similar Broadcom chipset." ........... no so, DGS-105 has a Realtek 8576N
The cable companies that I know of that makes upper end ethernet cables are Chord, Audioquest, and Nordost. Are there any others?
Hmm, the Netgear is a GS108 turns out. Still, any of these switches better than most combo modem/routers. I have a basic netgear modem, with a heftier supposedly low noise cheap smps brick powering that. Apple Extreme doing the wifi lifting separately. Separate circuits for hifi, network, and wifi.
Well I put Audioquest Cinnamon Ethernet cables in, along with a new netgear switch box. I immediately could tell a difference. Music sounded more natural, more relaxed. The brushes on the highhat the drummer was playing sounded more real for lack of any other term. Less brittle. Vocals sounded more natural & real & lifelike. My wife who was very skeptical that this was going to have any effect on the sound quality, kept saying over & over "This is Crazy" shaking her head! She heard the difference right off the bat, and found it hard to believe. It was so easy to perceive the difference she said this can not be so! But it was! So all I can say is it made a difference in my system! And it was not hard to perceive!
Very good.
As you have introduced two variables at the same time (a switch and a cable) the test is somewhat less valid.
For completneass sake could you revert back to your old LAN cable and see what heppens and let us know?
Thanks,
Adam
Today I asked the dealer about multiple earthing with Audioquest cables. The chap who knows for sure wasn't in, but we think the cables earth at one end so they don't cause earth loops and the worst case is that they act as aerials.
Keith
musicfan51 posted:HH, do you know what the Chord Ethernet cables sell for ?
I wonder if they are better? Audioquest is probably very close and a better
Value. You know with Naim making some of the best sounding
Streamers out on the market, I wonder if they would ever put out
An Ethernet cable?
since Vertere makes the Super Lumina cables from what I understand.......
since all the expensive Ethernet cables built by Audio Quest, Cardas, Vertere use Telegartner MFP8 RJ45 plugs, Telegartner makes patch cables or even custom......
https://www.telegaertner.com/e...y/?IdTreeGroup=14425
Audio Quest offers cheap highend custom cables
Telegartner RJ45 plugs $20ea USD
http://www.audioquest.com/cat_...rtner-male-connector
CAT700 Carbon $13.00ft USD
http://www.audioquest.com/cat_...spools/cat700-carbon
KRM posted:Today I asked the dealer about multiple earthing with Audioquest cables. The chap who knows for sure wasn't in, but we think the cables earth at one end so they don't cause earth loops and the worst case is that they act as aerials.
Keith
I doubt that Kieth, AQ & many other cable makers with paired conductors do have some cables with a floating screen - normally connected to return (-) signal at the source end & floating at the other; my own phono RCA's are this way. But it doesn't make sense with ethernet, the idea is to ground the screen at one point & carry the screen continuity over the rest of the network & components that are not ethernet port connected to ground, or to isolate them if they are. I've looked into AQ www blurb & see nothing about this, nothing mentioned in paragraph about cable directionality or in the paragraph on the screen where they say some confusing stuff about RF Interference from modulating the equipment’s ground reference requires AQ’s Noise-Dissipation System, so maybe they are doing something different. If this dealer guy says they do have a floating screen, then shame on them for not making this point clear, but I would still have doubts & would insist on testing a cable myself before believing it.
greekspec2 posted:since all the expensive Ethernet cables built by Audio Quest, Cardas, Vertere use Telegartner MFP8 RJ45 plugs,
Visual quality & robust construction aside, a problem with the Telegartner RJ45 plugs is they have short section of PCB where the ethernet pairs are run parallel to each other, this disrupts the twisted pair protocol & adds to near end cross talk (NEXT). It makes compliance to Cat7 ( 600MHz) difficult & they will probably only make it to Cat6A (500MHz) stnds. That is borne out in ARS www test of AQ Vodka ".... the cable did pass 6A standards .... the difficulty .... was near-end crosstalk. .... Vodka exhibited near-end crosstalk numbers that were borderline relative to (Cat6A 500MHz) spec .... only 0.6dB above the spec crosstalk limit" They part blamed the Telegartner's.
Mike-B posted:KRM posted:Today I asked the dealer about multiple earthing with Audioquest cables. The chap who knows for sure wasn't in, but we think the cables earth at one end so they don't cause earth loops and the worst case is that they act as aerials.
Keith
I doubt that Kieth, AQ & many other cable makers with paired conductors do have some cables with a floating screen - normally connected to return (-) signal at the source end & floating at the other; my own phono RCA's are this way. But it doesn't make sense with ethernet, the idea is to ground the screen at one point & carry the screen continuity over the rest of the network & components that are not ethernet port connected to ground, or to isolate them if they are. I've looked into AQ www blurb & see nothing about this, nothing mentioned in paragraph about cable directionality or in the paragraph on the screen where they say some confusing stuff about RF Interference from modulating the equipment’s ground reference requires AQ’s Noise-Dissipation System, so maybe they are doing something different. If this dealer guy says they do have a floating screen, then shame on them for not making this point clear, but I would still have doubts & would insist on testing a cable myself before believing it.
Hi Mike,
To be fair, I was talking to the wrong man - the right one is getting married. I did discuss this with the right one at the time the cables were installed, but can't remember why my setup was deemed ok (Unitiserve - Vodka - GS108 - Vodka - NDS). I will check when he returns.
Not that I would say this is definitive - but its quite a good reference to bust certain myths or at least call into question certain statements made on shielded cables such as needing to ground them at both ends - or any end for example..
http://www.pennington.net/arch...Shielded_Cabling.pdf
Simon-in-Suffolk posted:Not that I would say this is definitive - but its quite a good reference to bust certain myths or at least call into question certain statements made on shielded cables such as needing to ground them at both ends - or any end for example..
http://www.pennington.net/arch...Shielded_Cabling.pdf
We're so busted!
Well, I have read this thread with interest since the start and never really thought about the effect of Ethernet cables until I saw this.
Since installing my streaming solution (Synology Nas >> Netgear GS-105 switch >> NDS/555PS, all linked via home terminated QED 5e FTP cable), I have been using fairly standard cabling and a standard Belkin patch lead between the NDS and an Ethernet wall socket.
I did try a few ferrite chokes which brought a very small benefit but left everything else as it was...until I read this!
I figured that if there was a benefit to be had with a better patch lead, then I should hopefully notice a difference with the NDS, so decided, after a little research, to try a Chord C Stream. The theory being that if I heard a benefit with a cheaper 'purpose built' cable, I could try some more exotic cables when funds permit and I have more time to try all the alternatives.
Well, this afternoon, the Chord cable arrived. I sat down and listened to a couple of tracks with my current set up before swapping over to the Chord. I'm not going to try and provide a vast descriptive analysis here, but suffice to say that there was an immediate and noticeable benefit; in summary, less noise, more detail with less hash and a more refined sound.
I was sceptical if I'm honest, but curiosity has resulted in a very cost effective (£40) upgrade. To say I'm pleased at the result for the money is an understatement ��
Thank you to the OP and all those who posted their findings, it made interesting reading and lead to a very satisfactory and cheap upgrade!
Hi Simon, a bit theatrical but it does make for discussion. I'm quite sure their are many thousands of boutique STP ethernet installs around the world with more than one ground & I expect the owners are delighted with their cable investments. I prefer to stick to the work practices I know & to avoid any screen voltage PD (ground loops), but in most domestic (home audio) that is so small its not really worth worrying about - unless the ethernet ends are on different power circuits & one has a fault. !!!! Most of the major cmpy's install manuals say one ground point only is needed but don't go into too much length about avoiding multiple grounds other than warn about screens with PD <1v. So it seems its not worth loosing sleep over
But I have seen the light with my home system & am now all Cat-6
Some here are advocates of ferrite chokes, and I do use them where I still have ordinary Ethernet cables. Why wouldn't I? They are dead cheap and easy to install. However, I have to admit I never really heard them make much of a difference. The AQ cables, on the other hand, make a massive improvement and I'm suspect the same is true of the Chord cables.
Mike-B posted:Hi Simon, a bit theatrical but it does make for discussion. I'm quite sure their are many thousands of boutique STP ethernet installs around the world with more than one ground & I expect the owners are delighted with their cable investments. I prefer to stick to the work practices I know & to avoid any screen voltage PD (ground loops), but in most domestic (home audio) that is so small its not really worth worrying about - unless the ethernet ends are on different power circuits & one has a fault. !!!! Most of the major cmpy's install manuals say one ground point only is needed but don't go into too much length about avoiding multiple grounds other than warn about screens with PD <1v. So it seems its not worth loosing sleep over
But I have seen the light with my home system & am now all Cat-6
I'm getting confused - are you saying that the Lindy thingy that you suggested I use is not really necessary after all? I use it between the switch and the nas, at the nas end.
I think I am right in saying that cat6 is unscreened so you don't need to worry about multiple grounding points. As you use AQ cables which are screened it makes some sense to take the precaution with the Lindy isolating adaptor to take out one of the grounding points.