Network hifi and mains pollution - underestimated problem?

In wi-fi mode there is not ground contamination.

If manufacturer does not take care of the ground isolation from ethernet line there is no chance you can do anything about it. I take it as granted that Naim/Linn do it properly (usually transformer decoupling would suffice).

To assist the RF rejection induced in ethernet line you may wish to install a ferrite choke around ethernet cable -- right at the entry point to a streamer. I took this advice from Simon and installed the one at KDS.

Another point is the digital electronics inside the streamers which is another source of contamination -- but normally decoupled in hi-end designs.

The major point (imho) is the dedicated power line for digital electronics which suffer from RF feedback and may contaminate the analogue power line. This is what Naim is doing through external PS and Linn is doing through the new range of Dynamic SMPS.  

For customer tweaks I can only suggest the following:

1. Use dedicated line for streamers and another one for amplifiers.

2. Use external PS when possible because they are normally meant to provide the additional stage of decoupling.

3. Use Naim Powerlines to power all gears.

4. Install the ferrite choke on the ethernet line.

5. When you see it all does not work -- drop a vinyl on LP12 and you're done.

Thanks for your thoughts, Ama.

I know that Wifi avoids ground problems. Maybe it has something to do with higher processor load why Wifi sounds worse IMO.

I think Naim/Linn make enourmous efforts to isolate ethernet from the hifi electronics but up to now it seems not to work perfect, at least in my case. I have very bad mains, the sound fluctuates significantly during the day. When I changed the phase (in Germany we have three phases per house) and established a dedicated spur on phase 3 the problem disappeared almost completely. But when I had an NDX for demo at home and hard wired it to my router (which is on the 'polluted' first phase), the problem with fluctuating sound quality appeared agan. Ferrite chokes didn't help either.

Different dedicated lines for PC and hifi gear might be an interesting option but I (as probably many others) have only one 'cleaner' phase.

You can install balanced main transformer.

I got this advice -- but I don't have enough passion to try it.

Actually my power is rock solid and clean. But both 500PS and 552 PS annoyingly hum.

Hi Oliver, this is certainly not a neglected area for many of us, just check my postings :-).

One thing you might want to consider is braided mains lead which can be quite good at taking the edge of spurious HF energy on the mains. I swear by it and can give a great performance boost, with out killing the sound like some mains filters.

Chokes on Ethernet UTP cable helps as AMA point out, and using linear powersupplies on quality routers and switches. You can also use PoE if you want to keep the switch powersupply well away physically frrom your local streaming switch

I am not surprised that wifi sounds worse, after aIl you are creating quite a strong RF field which almost certainly will produce low level intermods on high impedance audio circuitry.

Another area that isI often overlooked is RF coupling into audio interconnects, and the reflections created  at the terminations. This can upset driving electronics in terms of negative feedback lag. I use RF terminators in my main component interconnect (ndac/555ps) this creates a very natural organic sound. 

In short if you can control RFI your system will probably sound fantastic, organic and natural. Get it wrong and you wI'll hear  'digital' brightness and hardness, or lack of detail and spatial info and the sound will be tiresome and unnatural. 

BTW this is one reason why I seperate my network player from my DAC. I will be intriqued to hear the NDS, but is suspect there will be more pronounced TCP window interaction just like FLAC and WAV now. We will have to wait and hear.

Simon

Hi Simon, thanks for your interesting tweaks. I think I'll try the RF terminators in the next weeks. Regarding braided mains leads, it would be difficult to install them on my polluted first phase because I'd have to rip my standard mains leads out of the wall.

What really bothers me is that I didn't manage yet to decouple router and streamer properly without loss of sound quality. I had an HDX for demo at home and compared the sound from the internal HDD with different network setups and all sounded inferior to me.

1) I connected HDX via Wifi adaptor (Netgear WNCE 2001) with the router and it sounded as if the band was 2 meters further away.

2) I tried an inductor based galvanic isolator on the ethernet cable but it sounded dull and dead.

3) I even tried to decouple ethernet connection with two opto couplers (Ethernet > optical / optical > Ethernet) and, what I never expected, it also sounded dull and dead.

I have no idea why these modifcations, especially 2) and 3) which only concern ethernet transmission, can influence the sound. My only conclusion so far is that this whole network Hifi thing isn't fully understood yet.

Oliver

Oliver, intersting observations, especially 2). There is also the possibility that you have optimised your system to sound balanced to you with intermod distortion or brightness present through RF pollution.

That would be kind of like learning a musical instrument with a bad technique habbit. You will only be able to get so far before you plateau out. You would need unlearn the bad habit and temporarily step back so as to develop further.....

Simon, are you talking about the clip on ferrite type that you can pick up at Maplins for example? If they are am passing Maplins tomorrow I could pick some up and experiment.

Simon, the same thing does my dealer say. I got used to the 'polluted' sound and consider cleaner sound as dull. Well, that may be a possibility. On the other hand I liked the sound from the internal HDD of the HDX which I guess was not distorted so I hope that I can still distinguish good from bad. Did you try these decoupling methods? I would be very interested in other user experiences with isolators or opto couplers.

Spartacus - i am not sure what exactly Maplin sell, but it sounds like it could be similar, so worth a try. Clip then on at the source end and and one or two at the sink end. Experiment, and add until you hear no differnece.

Cables worth trying are twisted cables like UTP Ethernet leads  and shielded coax cables.

Oliver I have tried the following RF mitigation methods

1) Braided Belden Mainslead / Powerline

2) Chokes on UTP Ethernet Lead

3) Mutiple HF chokes on SPDIF lead and lead DC1 wound in a losse 4 turn coil - space between coils.

4) RF attenauators in the analogue interconnect between nDAC and preamp. (Mr Morecroft of DNM made the RF attenautors for me)

5) Ethernet lead not touching any other lead including mains leads 

BTW I beleive Galvanic isolator is more likely to stop low frequency/DC ground current ie ground loops  rather than RF currentls.