Indeed, a few years ago I used a USB stick as the constant (unchanging) benchmark to help fine tune my LAN, it was invaluable as without doubt our himan fallible memories & audio perceptions constantly change. Different cables do have differences, but its all very subtle & using the USB as a benchmark made it so much easier to continue over the days.I found (in my case & at that time) the biggest change was adding a switch to instead of routing via my (then) wireless hub.
I did exactly the same, except my cable came from CPC Farnell (my cable also cost just £5, but the two ferrites I also use cost more than the cable itself!).
And yes Mike, I also used USB sticky memory as a fixed reference point.
Disconnecting the lead on playout will also demonstrate the difference your lead is providing (on your new Naim streamer)… it really is so straightforward…
Not for me. I don’t have a new Naim streamer, but tried with Linn MSDM, Linn ADSM, Sonos Connect, Sonore μRendu & Mac mini. No difference. Tried switches from HP, TPLink, Cisco & NetGear. No difference. Whereas DACs, Amps & Speakers sound different. To me there’s better ways to spend a hifi budget than network switches & cables. Maybe an extreme view, but measurements I’ve seen are consistent with what I hear, That’s why I say ignore reviews and audiophile groups and listen carefully before you buy. What others hear or imagine they hear may not be what you hear.
Maybe, in the way that you personally perceive music, you’re just not that sensitive to the deleterious effects of RFI that some of the rest of us hear.
1 Like
Hi Xanthe, I’ve rethought my ethernet ferrite & removed them from the NAS & NDX branches & left only with one on the router to switch branch.
I’m questioning the function & effect of ferrite on ethernet. RJ45 (8P8C) ports include galvanic isolation as an integral part of the port. This isolation removes any direct electrical connection around the network & currents such as induced by RFI or EMI are blocked, so there is no common mode noise (current) on the network and there is nothing for the ferrite to absorb.
The counter argument is the galvanic isolation does not guarantee that no common mode at all might be present.
… I’m still thinking, looking for some www research on this, but my Mk-II audio receptors tell me there is no difference.
Hi Mike, I took the ferrites off for a period then did the test again and after that test I left them on (incidentally I removed the ferrites from the analogue cables as with the W2497 they made no difference between the 272 and the 300!).
My Mk 1 audio receptors revealed a different situation to yours (there can be many reasons for this both in our perception of music, and equipment and environmental factors).
Mike, can I refer you to the paper by Texas Instruments on reducing Fast Ethernet emissions in connected equipment.
You will see on an Ethernet segment common mode currents flow both ways between the MDI (Media Dependent Interfaces). This is potential cause of interference and ground and powerline modulation.
Now in some quality designs where the effects of Ethernet radiation needs to be reduced or interference is prevalent then ferrite beads are recommended internally to decouple physical interface components to reduce the effects of noise coupling. I understand an impedance of 100 to 2k ohms at 100 MHz is often considered adequate. However in my experience this is quite rare in consumer electronics, and I am not even sure Naim do this (though I am not certain of this), but a ferrite choke clamped around the Ethernet lead at each end of the lead can help reduce the effects of coupled noise.
The earth galvanic isolation of Ethernet connectors is for very low frequencies and DC (as with all galvanic isolation) and so will NOT form a barrier to higher common mode frequencies, but may attenuate due to hysteresis losses. However also remember the higher the frequency, as with higher RF, the common mode currents will likely capacitively couple across a galvanic transformer and so bypass it,
Now clearly the chokes are only going to have an effect if there are coupled common mode currents traversing the lead, and this will depend on the design and I suspect quality of the design of the connected Ethernet equipment at both ends of the lead. So the effects of such chokes will be implementation dependant.
Hi Simon, I’ve read that a while back, & circled it a few other times since, so thanks for the prompt, but I am aware.
I cannot determine at what numbers & what the low frequency attenuation/barrier curve(s) looks like, plus I suspect each isolator brand/design is different to another.
I do not have a significant amount of RFI/EMI noise that I can detect, my main concern in that area is SMPS switching noise - which is in the 50kHz to 3MHz range.
I have all my audio circuit related SMPS’s powered via my UPS, which has an internal isolation transformer & a C&D mode filter.
Additionally I have installed a number of LF specific 150kHz-5MHz & also 1MHz-500MHz ferrite clamps on both sides of the SMPS’s.
So the SMPS switching noise is pretty well suppressed to some degree, and that is why I’m rethinking & re-listening with/without ethernet ferrite.
NB: I have kept ferrite on the wireless hub branch as I’m unsure of the level of noise emanating from that.
Are there any updates to this covering 1000 Base-T, as this paper only covers 10/100 Base-T connections.
I understand that the NIC on the Naim end is 100 Base-T, but in many cases the switch it is connected to is 1000 Base-T (Gigabit)
Are you also following what UpTone (providers of the UltraCap LPS power supplies and ISOregen for USB) is planning in their EtherREGEN project - a new implementation of a 5-port switch with regulated power supply for minimal noise and also a fully isolated port for connection to Streamer/Network player.
Interesting this ‘moated’ Ethernet connection is only going to be 100 Base-T, whereas the other 4-ports on the other side of the isolation moat are to be Gigabit. There is also a STP cage for an optical module for those with an optical based connection.
Simon.
The principles are the same. with 1000BaseT all eight wires are used - so the potential for noise is slightly greater. With duplex 100BaseT then only four wires are used.
I think going optical addresses much at the MDI layer - I have mentioned that before to Naim…
The focus then would move to the digital stack and minimising network chatter.
I think this is a good place to ask for opinions. In my system I do not use Ethernet cable between my WiFi router (Asus RT-AC68U) and streamer Auralic Aries G1. Apart from it being very impractical and difficult to implement a hard wired connection in my rooms, Auralic believes WiFi is a better connection. I use 5ghz and Auralic is the only device on that band. At router side I have Synology NAS 214play hooked up to the router with AQ Vodka. Based on this setup any one have any thoughts on improvements ?
the only real potential improvement is a wired ethernet connection… however if wifi is fine and you get good signal strength and no drop outs then that it is fine…
However wifi can be totally acceptable and may even have some advantages over wired ethernet - as well as a few disadvantages - so swings and roundabouts - its what works for you. There is less to tweak for wifi - I have yet to see audiophile wifi antenna 
(now I have done it…)
2 Likes
You should be obsopherlated for suggesting such a thing !!!
Joking aside, I know someone who is working on just that, not so much ‘audiophool’ but attempting to make it work ‘better’. All harmless stuff but he should get out more & please stop trying to get me interested.
2 Likes
Simon and Mike, you need to install an audiophile air purifier to clean the air that the WiFi signal goes through otherwise the dust and pollutants will muddy the signal and you won’t get any inky blackness!
4 Likes
Night time listening only recommended.
G
I haven’t come across a recommendation for WiFi over wired from a manufacturer before, and that certainly isn’t Naim’s approach. Perhaps Auralic have accepted that some will expect WiFi to “just work” and tried to optimise their gear around it, which is fair enough. The problem is that WiFi is more prone to environmental influences than wired Ethernet, and if something changes at some point in the future, you are left tearing your hair out trying to diagnose and resolve it.
Having said that, wireless hardware for home use has improved considerably over the last couple of years, so if it works for you, that’s fine.
You’ll be stuck listening and enjoying music instead of going down an endless vortex of Ethernet switches and cables 
1 Like
What, pray, does obsopherlated mean?
Indecypherable obfuscation?
1 Like
… what Bluesfan sez, its twaddlespeek
2 Likes