Fibre is Ethernet unless you are referring to Toslink…. And FTTP doesn’t use copper twisted pair, it uses typically PON fibre. Twisted pair copper is FTTC, that is fibre to your nearby cab, and twisted pair from the cab to your house, using likely what would have been your phone line. Now nearly all house phone services are VoiP over the broadband link in the UK.
Sounds good, why don’t you think this is optimally ‘clean’ for audio.
Nope, made an assumption that on your comment of “interesting commercials” and “PEW” you were a service manager as they steered a very steady path of not stepping on account manager’s toes with commercials and know about PEWs, obviously wrongly 
Probably similar, a lot of network and mainly VoiP network design in CNI and secure environments.
lol sounds like we might be in similar lines of work 
. Though in architecture one is close to the commercials of the required outcome - and I encourage my designers to be more commercially aware.. Just like we have Secure by Design, I say we equally have commercial by design as well..
I haven’t focussed on VoIP in recent years…. If you exclude Teams…
Teams is not VoiP, it’s the devil’s bumhole 
1 Like
Hi Simon,
I hope you don’t mind answering another question. Do you like the higher noise suppression of Cat 5e generally or only for the cable between switch and streamer? We should get fiber soon and I want to establish a wired network at that time. Three Ethernet cables have to run approximately 25-30 meters parallel to the dedicated mains cable for my system, including the one for the streamer.
Now there is a question… this in the land of marginal affect. In t he days of my NDX2 I preferred Cat5e between switch and streamer - and had a fairly long run.
These days I separate my DAC from my streamer - NDX2 - and now all these considerations become irrelevant for me - thank goodness - as the effective decoupling (yes there is a minute amount of coupling still but not relevant to my senses) between DAC and streamer isolates all those network considerations.
20 to 30 metres is going to be fine - but running parallel to mains cable not so good - unless spaced out. There are actually structured wiring regulations on this ISTR - in the UK at least but can’t remember the required spacing.
What I would say though is not to get too hung up on it. If you have a recent streamer from Naim significant design effort went into decoupling the network interface from the delicate clocks and audio circuitry… I think if you use earthed network switches etc that will go a long way to keep common mode noise away.
This topic came up with length of cables in as far as given the choice of short or long cables to the streamer go long. If the choice between Cat 6 or Cat 5e between switch and streamer go Cat 5e.. but I wouldn’t go out of the way here and either is going to work fine… with these two aspects we are talking super subtle.
1 Like
Many thanks! There are regulations regarding mains and low voltage wiring over here in Germany, too. I have already spoken with the certified electrician who is going to do the final work about this and he gave the green light to what I have in mind. As much as I would like to separate the two, there is only a single path available for the ones mentioned.
1 Like
This is normally where shielded network cables would be used, however, you then run the risk of the streamer becoming the route to ground for all the common mode noise on the shielding, even if you have a grounded switch.
Another solution might be to find a small switch that is grounded and then locate that at the end of the run allowing you to use shielded cables alongside the mains. I don’t know if the forum favourite, the Cisco 2960 is grounded. A single (two if you’re worried about failure) cable will have more than enough bandwidth for music and 4K video running at 1Gbps.
2 Likes
The 2960 is grounded.
I would ensure that the shield on the Ethernet cable is connected at the switch end not at the streamer end. If necessary a simple RJ45 coupler could be used to break the ground connection at the streamer end.
1 Like
@Mulberry And my advice and experience suggest the exact opposite. Hearing for yourself is of course the only way to meaningfully put theories (these are aplenty!) to the test of practice.
Thanks for sharing! As I have to use the existing conduits, it will be a long cable to the streamer. Maybe a later remodeling will open up other options, I’m stuck for now.
A network switch or FMC installed for audio would typically make use of existing conduits or indeed in most cases existing copper - Cat 6 or whatever - or fiber connections.
If I were planning to run alongside power cables for 20-30m, I’d go fiber as there will be zero electrical interaction between the two; as @Simon-in-Suffolk says, running signal/data cables in the same conduit as power is not a good idea if sound quality is important.
Then I’d install a high quality (= low noise) fiber media converter to do the optical-electrical back conversion. The fiber connection is 100% noise-free so no concern here but the transceiver doing the back conversion is typically noisy in its own right which is one reason the theory of “perfect” fiber rarely translates to perfect sound.
If a streamer has an SFP port, most folk who do the comparison (many don’t bother, which is a missed sonic opportunity) find that doing the back conversion before rather than inside the streamer, and connecting to an RJ45 port on the streamer by a short (0.5m ideally) UNshielded Cat cable reaps sonic benefits. If you think about it, the last place you want to be doing noisy back conversion is actually inside the streamer!
Hope this helps, and god luck with your project.
Pedant flag raised… the fibre connection is only 100% noise free with respect to conducted common mode noise which is about stray analogue coupled voltages seeking the lowest impedance to earth, rather than anything network related.
Yes many cheap consumer/ audiophile switches use non earthed manis supplies and chassis and can be a creator rather than sink of common mode noise flowing through the metallic wires of twisted pair Ethernet leads etc… I say best use mains earthed switches to mitigate this.
However actual network related noise applies equally to fibre Ethernet and twisted pair Ethernet leads.
I agree on the benefits of using a fibre SFP module if a streamer has a transceiver port.. it will mitigate common mode noise… and if the other end is a quality device with low phase noise symbol clocking on the fibre interface, then some actual network related physical layer noise power is reduced too. I have previously suggested to Naim that they might consider using a transceiver port on their streamers… the consumer can then use a twisted pair SFP or fibre SFP to suit.
Symbol clock phase noise equally applies to twisted pair connections too.
Noooo… I won’t have it!! 
1 Like
It’s fine flag, underused in my opinion!
The fibre connection itself is 100% noise free with respect to all forms of noise. What happens with the devices at either end is important too of course and unfortunately this is overlooked by many.
For audio optimisation, the concern is not only noise being conducted from one device to another along the (metallic) conductor wires. It is also noise being conducted from one device to another along any (metallic) shield which is grounded to both metal plugs. And it is radiated noise which is picked up by conductors and/or shields and/or device cases.
Which streamers have you tried this on, comparing a fibre connection and optical-to-electrical conversion inside the streamer with a copper connection to the same streamer’ RJ45 port? My only experience, and it is secondhand not personal, is of a couple of Lumin streamers where copper connection or external conversion was preferred to internal conversion.
Best,
Nigel-in-Leicestershire!
Even this is not strictly true. Optical cables can pick up mechanical vibrations and this can cause measurable electrical noise when it is converted back to an electrical signal.
1 Like
Good catch. Some recent experiences comparing (metallic) cables which are identical except that one incorporates an anti-vibration layer suggest there is something meaningful going on here, which is not what I would have expected only a few years ago.
But I’ve not seen anyone do the same comparison with fiber; have you? I love a good theory as much as the next guy but I always like to see it put to the test of practice. It would mean comparing a stock fiber connection with one somehow modified to mitigate/reduce/manage/control/other-synonyms-are-available the mechanical vibrations of which you speak.
Thanks for any experiences you can share.
Only my own designs, as I said I am not aware of any commercial consumer hifi products that do this… but have suggested to Naim. I think the view might be that a,though there may be SQ improvements it may be deemed too ‘technical’ for the typical audiophile consumer.
The good things with SFPs is that they use the host clocking electronics for all symbol timing… so can scale well in this regard in to what they are plugged into.
I guess also these days with the latest wlan peotocols the same benefits can be had with wifi and top notch flow performance as well so might be reducing the need in the typical domestic environment.
Personally I have found in my home setup seperating the DAC from the streamer has the best audible performance in terms of decoupling network affects and noise considerations. The whole thing has become a non issue now.
1 Like
Yes this is the common mode noise… but in the limit is similar to noise picked up from the product metallic case work etc.as you suggest
So we need to put some context to this, and this is one of the reasons why I near,y an,ways recommend using earth grounded devices where low noise is required. Having a device reliant on a simple 12 or 9 volt DC jack is not going to cut it
Yes in some of my professional work of a few years ago.. this property of microphony in fibre cables has been exploited in security and anti tamper devices. It has been used in gas pipelines and security surveillance systems.
I had described this in more detail on this forum several years back.