HD Music Streaming over Powerline Adapter

Actually the noise from switch mode power supplies is typically higher than the signals used for PLC. One of the problems with PLC is the interference caused by all those rubbish chinese power supplies. My advice is go for mains passthrough homeplug units and then ensure that EVERYTHING that uses a switching PSU (and that includes CFL and LED lamps) is connected to the filtered mains passthrough socket on the homeplug device. This stops the PLC signals getting to the power supplies and, more importantly, stops the noise from the power supplies getting onto the mains.

Power line connections are weak links. You could consider what I have done, just install a 2nd NAS an LAN close to your other source.

it's a bit more expensive, but very good.

Bert exactly.

Solwisesteve, based on my experience of lab measurements, discussions in the past with Ofcom and EM compliance professionals and my own professional experience we will have to agree to disagree on that one unless your SMPS has gone faulty.

Simon

Agreed... tbh I use PLC technology in my own house but I find having the NAS close to what ever's using the data to be the preferred option.

erm.... You come to conclusions not based on anything.  For example " if using PLAs I certainly would not worry about using SMPS on or near your audio equipment, as the noise from PLA almost certainly will be significantly greater."  Will it? Why?  Who mentioned switched power supplies, why the comparison?

You seem to be worried about the affects of HF from PLAs on SQ but totally ignore affects of HF from other sources.  Have you ever hooked up a scope to your mains supply and then compared to what is on the other side of the PSU of a piece of HiFi?  Have you then repeated this measurement with PLAs in place?  If so can you please publish your findings.

The actual facts of the matter here are that a number of people use PLAs for domestic networking and many of these are also using streaming technology for HiFi and can detect NO negative affects on SQ.  Just look at some of the posts on this thread but you totally disregard their findings.

To say that using PLAs negates the use of using 24bit (versus 16bit) source files is totally ridiculous and is based on what?

Yes the ideal solution would be to have the NAS next to your streamer but this has a couple of problems:

(a) NAS Noise - this just about kills it,

(b) If you are using nStream to control your streamer then this mini network will have to be connected to a WiFi hub.  So you would need a WiFi hub in your music room,

(c) If you have a WiFi hub in your living then to get network radio you will need to connect to your WAN modem, which in turn will have to connect to your Broadband port.  Mind you for most this won't be an issue because they don't listen to iRadio.

So if you can accept a noisy NAS or perhaps use a dedicate (quiet) media server, rather than a NAS and do not use nStream then you are off and running.  I don't think it will sound any better but you will DEFINITELY have fewer network problems.

Which reminds me why have NAIM not supplied an application that runs on web browsers?  Then you would not need a WiFi hub in your living room.

I must disagree on the NAS in the listening room Big Bill

I have mine in the room & its nothing like the picture you paint,  anyhow my set-up aside - I would suggest  ............

NAS near the streamer - modern NAS units are not noisy, most with a fan are 20dB or less,  the most annoying issue over & above noise for me is the flashing LED's.  Mine is in a cabinet & that fixes the lights & reduces the very low 17dB on full fan,  not the even lower normal speed fan,  to effectively silent.

The WiFi Hub need not be in the same room, it can stay were it is.  

The NAS can (best) be connected to the streamer via a network switch - fanless & silent - but also with flashing LED's that need to be in a cabinet - a network switch is so much better that connecting via a wifi hub anyhow, they are not suitable for streaming & performance is questionably close to inadequate. 

The link from switch to hub can be via a single ethernet or if that's a problem, a wireless link. 

Well said - I used Powerline for some time and have recommended it to colleagues who have used it with no issues. I took the trouble of cabling direct to eliminate Powerline and cannot detect any difference in sound.

You may not,  but others in your neighbourhood might 

There are European & global discussions going on about banning PLA's,  Ofcom have the lead for UK & UK Gov has issued a response saying at the moment they do not agree an outright ban, but to test & prosecute non-compliant makes/models.  Norway has banned a number of TP Link's. 

I use an faness Intel NUC as a NAS - zero noise.

You can run an Android emulator on your PC for the app BUT I found it just about kills PC performance running it.

Yeah if you have a quiet media server using Windows you could use fooBar2000 to control the streamer I guess - but not ideal.  Not sure if anything in Linux would have worked.  Would have been fairly easy to implement a browser solution.

I take your point Mike but you are connecting to your WiFi hub which is outside of the living room, so you might as well move your NAS outside the living room too.  Ethernet (when cabled not WiFi) does not see walls and cable lengths for 100% operation are huge.  Basically what you describe is exactly the same as mine, except my NAS sits close to my WIFi hub.

ps my QNAP is noisy and gets more noisy the older it gets, which is typical of machines with fans.  solwisesteve is on the money using a fanless PC.  You could build one yourself.  Use a powerful processor and then underclock it.  Put it all in a case designed for fanless operation.  As I said earlier control is now the problem.

There is a company called QuietPC who have a lot of information available for those wanting to find information on building fanless PCs.

There are also other resources that I can't recall off the top of my head.

P.S. my Synology is effectively noise free (general noise form the external environment is louder).

As I said PLAs work and commercially available PLAs all should be legal. I have even worked with some of the design teams that developed the technology early on..and yes the RF side effects were understood then as well.

But in use on the mains wiring they do produce RF radiation , to the extent these 'side effects' have been discussed and debated many many times by the EMC professionals, standards bodies and the PLA lobby.

So perhaps the key point is wheather this RF radiation becomes interference is a matter for the end user as long as any interference does not affect anyone else... And that is sort of where the compliance debate wasgoing  when I was involved with it.

So my point earlier was it would appear to me counter intuitive to go to extensive efforts to reduce RFI on the mains, which some people do on this forum, in order to improve SQ, if you are only to re inject RF carriers onto the mains again.

However if RFI does not affect you, or specifically the RF from PLAs does not cause interference to you then fine..

The point of SMPS is that the switch frequency and its harmonics need to be filtered on the mains. The hundreds if not thousands of RF carriers injected onto the mains for PLA must not be filtered. 

Certainly I have observed  the EM field at my property with a band scope and the frequency and harmonic noise from various SMPS on the mains was typically minute .. Activate a pair of PLAs  and I got wideband carriers and their intermods appearing everywhere from around 3MHz to above 30 MHz and is significant.

But PLA, whether new or old, does produce wide band RF electro magnetic radiation from domestic UK mains wiring (ie unshielded). It is relatively low level and whether it affects you and causes RF interference is clearly subjective and dependent on your circumstances, but the RF radiation is there.

Simon

I agree about using fanless PCs, NASs, Servers etc if you can, I use Netgear NAS with a variable speed fan - and it is rarely audible as it is usually on tick over on streaming duties - goes quicker/louder when doing backups. The actual media server I use is a RaspberryPi 2  - which runs MinimServer and is small and totally quiet and fanless- and is the size of a large box of kitchen matches. I would say my Sky+ STB is pretty similar in noise, if not slightly louder, than my Netgear NAS in regular use.

Simon

Should have also said that diskless is ideal, I am afraid that disks seem to get noisier as they get older.  Now that SSDs are becoming very cheap the reality of a totally silent PC is now here.

........  that's probably my next NAS move sometime in the future, might be my disk space is nearing size limits, getting noisy or a failure.   My only concerns are reliability & durability as its my understanding a HDD is more robust & SSD can just fail for no reason. It may be all rumours or folk law from days gone by,  or maybe a few more years HDD will be as dead as .....  a CD  

Mike, I know what you mean.. I suspect it's folk lore.. Despite that  however  both my NASs have conventional low energy disks whilst my Laptop has a SSD... 

I don't know very much about this subject, but this appears very logical to me...

you use expensive power supplies and powerlines, but you wouldn't mind to pollute the mains?

you can't brew a good beer with dirty water, not?

Or am I seeing things to simple?

Mike, Simon,

The unreliability of SSDs mainly applies to earlier generations, the current generations / technologies have much higher MTBF.  Wear levelling has helped a lot, but the underlying technology is also more than an order of magnitude more durable than 1st gen systems.

If building a silent system to be NAS for an audio network, I would advise against using the most powerful CPU you can get that can run fanless; rather use the lowest TDP and clock speed that are clearly adequate for the job + a bit of headroom.  This will reduce the exposure to RFI.

Another problem area is the SMPS, I've never seen a review of computer PSUs that give a clear indication of the level and frequencies of RFI injected back onto the mains supply.

Whether PLAs may constitute a significant problem for SQ depends on the PLAs themselves, the audio system components in use, the state of the existing mains supply presented to them (including the arrangement of and quality of the wiring in the ring main) and possibly other factors.  I feel that the situation is just too complex to say whether they are, or are not, a problem in any given installation, without having definitive evidence from that specific installation.  I also think it would be unwise to claim that they'll always be a problem, or, conversely, that they'd never be a problem.

PLA performance seems to depend on essentially uncontrollable factors.  I wanted to get UPnP from the downstairs Synology NAS that sits close to my main router and cable modem to a headphone listening setup upstairs (SoTM sMS-100 + Schiit Audio Bifrost and Asgard 2). I tried two different PLA models, ZyXEL PLA-407 (recommended by a colleague who is happy with his) and Netgear Powerline 500. In both cases but especially with the ZyXEL, I had frequent packet losses and highly variable ping times, terrible to listen to. I wasted a lot of time on this until I just gave up on PLAs and ran 25m of CAT-6 from my 1Gb router downstairs to a 1Gb hub upstairs. Problem solved, zero losses, fast ping times with low variance, very good listening results. This is in California, 30 year old mains wiring, relatively dense neighborhood, so it could well be that something is really messed up in our mains wiring or interference from appliances, but it's just not worth it to fight this battle. The exposed network cable is not pretty, so I plan to eventually get the house wired with CAT-6 professionally inside the walls, but for now I'm not tearing my hear trying to figure out why "easy to use" PLAs can't reliably attain the advertised throughput.

It's interesting to view the differing attitudes between WiFi and PLC. I mean EVERYONE must have suffered from rubbish WiFi but people just shrug their shoulders and accept it. They don't start blaming the manufactures etc... Also WiFi throws tons of micro-waves into the air but no one moans. Ditto mobile phones... Over half the calls I receive from mobile phones are frankly rubbish! Poor quality. Drop outs. Disconnects. Plus the air is full of high power micro-waves because of them and with multiple reported investigations concerning health issues. Again people shrug and accept it.

SSDs have a troubled history and many people are worried about their reliability, after all like me many people use an SSD as their boot drive.  So we are right to be concerned.

When I built my current tower system I wanted powerful components at a reasonable price, so this meant buying kit that is not quite at the forefront but has been around only a short while, but is of a higher spec than that in the average of the shelf box.  This is probably the way of maximising price v. performance.  I purchased an SSD that failed almost immediately and sent it back to the suppliers. I got talking to one of the support guys who gave me the frightening info that one of the major manufacturers were using SSDs that were flagged as failures by other manufacturers.  At the time these guys were the cheapest around - no kidding wouldn't you just guess that.

So if you are going to build a PC with an SSD than make sure you get one from a reliable manufacturer.  At this time the support guy recommended a Crucial SSD which his company did not sell but Amazon did.  The PC was built in Oct 2012 and the SSD has functioned perfectly

So be warned...

One of the issues with the PSUs used with PCs is that are built with cost in mind, much too much IMHO.  The criteria for a PSU for a £5,000 pre or power amp is vastly different from that for a £30 computer SMPS.  As I see it there ain't no way around that, because nobody makes high quality SMPS for computers and who would believe anyone who say that they do.

Being a tube freak at heart, then as far as I am concerned anything other than a choke regulated PSU is dubious, that alone SMPS.  But as I have got older I can accept a solid state rectifier.

But if you could buy a computer that:

(1) is fanless,

(2) is diskless

(3) can output 24bit digital

(4) has a quality psu

You wouldn't need to buy a NAIM (or any other) streamer, nor would you need a NAS.  Some manufacturers have produced such boxes (that do some of 1 - 4) but the last time I did a survey none of them could play 24bit files and in fact they were complete boxes - ie had a DAC built in and so were analog out.  Analog out from basically what is a computer is always gonna be compromised!  In fact thinking about it I don't think any of these boxes did any of 1 - 4!

Is it pointless reducing RF on the mains?  But to say that PLAs would re-injected after this cleanup is not correct?  That is not how it would work.  If you applied RF reduction to the whole of your mains then PLAs would not work on your setup. DOH!

If you want to minimise the affects of RF on your equipment then clean it up just before it goes into your kit.  Come on this not rocket science.  A simple, inexpensive mains filter can be put together using some mains rated caps and a enamelled copper coil around a ferrite rod.  You can also add a surge protector.

I use one of these on all my kit apart from my record deck which is powered by a lead/acid battery.

I have checked the performance of these with a scope and believe me they work.

Remember too that any decent bit of kit will also employ measures to reject RF.  Today with all the RF rubbish polluting the atmosphere from phones, WiFi hubs, fridges, microwave ovens, computers, cat flaps etc etc.,  it is the stuff that your kit picks up from the atmosphere rather than mains wiring that is more of a worry.  This is much more difficult to deal with.

So it looks as though we have established some things...

1a  Some people's mains supply has little RFI coming from outside

1b  Some people's mains supply has a lot of RFI coming from outside

2    Point to Point PLA performance is affected by the mains wiring and the amount of noise on the incoming mains (these problems can sometimes be sufficiently severe to cause partial or complete failure of the PLA communication).

3a   PLAs do inject RFI on to the Mains

3b   SPMPS do inject RFI onto the mains

          (some are relatively benign some are very noisy [in my experience, this variance includes PC power supplies])

4a   Mains borne RFI leads to radiated RFI (to some degree or other)

4b   The effect of the penetration of radiated RFI on the audio equipment can be reduced (but not eliminated) by simple means (such as ferrite chokes)

4c   Some people using or intending to use PLAs are not taking the RFI reduction precautions from point 4b.

5a   Simple (inexpensive) mains LC filters remove a lot of the mains borne RFI

But then there are also some other relevant points taken from other posts on this forum...

5b   Many people on this forum find that simple (i.e. inexpensive) mains LC filters "suck the life out of the music" when used with Naim equipment.

5c   Both mains borne RFI and radiated RFI can lead to a reduction of sound quality in audio equipment if they are allowed to penetrate the system.

5d   I have conducted experiments on alternative mains filters (qualitative empirical results published in other threads on this forum) that indicate: Within my system, the mains capacitors used in LC filters are the main source of the SQ degradation leading to the "suck the life out of the music" effect caused by these devices.

My conclusion (others may reach different conclusions for themselves):

These points suggest that PLAs can often be used successfully without major impact on sound quality:

Either because the mains is already so polluted that no further degradation of Sound Quality is heard, or, if appropriate and sufficient precautions against penetration of RFI are taken (protection from all forms of RFI is needed).

There is also no guarantee that PLAs will function successfully in any specific given environment.

However these conclusions only cover sound quality and communication function, it doesn't cover the (contentious) possible health issues, nor does it cover any potential effect on other equipment or any potential effect on your neighbours.