Rerouted Ethernet Cable

I too have heard the "contradictions".   Up to a few weeks back & a cable sceptic, I was far from convinced;  I read & noted the reports but kept an open mind.  Then a few weeks ago because of RJ45 plug failure, I changed out my all Cat7's for Cat6 MeiCord & a small but noticeable change in sound.  NB:  I did not say quality or better.   Last weekend I carried out a reinstall tidy up & made a positive effort to route the ethernet at much as possible away from AC power, phone/ISP & DC cables.  Again a small but noticeable change.

I know the data is transmitted & verified as complete & the player is playing from the buffer, not the cable. But something is doing something.    Simon-in-Suffolk suggests the data is carried by Manchester encoded voltage & as its analogue it can be changed. Meicord suggest the importance of the twisted pair integrity & RJ45 plugs with sub-optimal twisted pair geometry. They claim such errors cause signal transfer issues such as near end crosstalk, return (reflection) loss, skew delay & reduced bandwidth that potentially can cause data transfer anomalies.

Interesting ......... still learning

call it old fashioned or whatever but there is a bit in me that almost regrets selling my CDS555 when i decided to stream instead. its hairshirt enough to worry about dressing interconnects, powerlines and burndies -- but dressing ethernet cables???? wow! funny old hobby...

enjoy

ken

Hi again Ken,  in my experience the changes I mention above are subtle,  the listener would need to be "tuned in" to the system to hear the changes.    Lets say its pretty good & then gets better.   As to going back to CD,  no way !!!  I'm pretty well finished with my ethernet changes now,  the last item is a fibre optic broadband arriving in two weeks, I but don't expect any change in SQ,  maybe something on iRadio & Tidal (maybe) 

just tongue in cheek Mike -- it would be hard to fit a CDS555 into my system now without re-arranging my Fraim racks.

but other than needing to have a CD shelf in my office (which took up a lot of space) -- CD replay was incredibly 'simple'...

enjoy

ken

Dan.S posted:

Kevin Richardson posted:

Dan.S posted:

By definition, Ethernet data is an error corrected therefore errorless stream. You can do whatever you want with your Ethernet cable, lift it above ground, suspended in mid air and using a Vodka for that matter, but the differences in SQ some people claim they hear between this and a cheap €2 cable tangled around all the power lines in your house are a result of other factors. Same goes for toslinks or any other digital transmission cables where checksums are applied to all packets.

Of course a €500 2m Ethernet cable must trigger some sound improvements, who would be crazy enough to spend that amount of money for a nice colored cable that does the same thing as a €2 cable.

I'm sure you believe what you say but I have first hand experience that contradicts your position.  

The really funny thing is I also have first hand experience that contradicts me, that's why I've stated these differences are a result of other factors.

What other factors?  I only moved my Ethernet cable.  Nothing else changed.  I continued to llisten to the same album I had been listening to all day but with significant SQ improvement.  The change was not subtle.  What do you think would cause this?

Was there ever a "suspended cable" craze?

Was there ever  ....... maybe not so much suspended,  but lifted   ...........   

Mike-B posted:

.....

Last weekend I carried out a reinstall tidy up & made a positive effort to route the ethernet at much as possible away from AC power, phone/ISP & DC cables.  Again a small but noticeable change.

 

do you now believe that keeping internet cables away from AC or DC power cables as possible (i.e. not just the 'minimum' distance) is beneficial to SQ?

enjoy

ken

Mike-B posted:

Was there ever  ....... maybe not so much suspended,  but lifted   ...........   

 

Maybe centenary suspension is the answer.

So here are three systems - choice is dependent on lengths and precision required

ken c posted:

do you now believe that keeping internet cables away from AC or DC power cables as possible (i.e. not just the 'minimum' distance) is beneficial to SQ?

Its not a matter of believing Ken,  there are so many variables & to repeat again I am not convinced short distances of parallel routing with very low current in power cables do cause anything thats so bad &/or audible.   The tight twists of Cat6 do a pretty good job of rejecting noise, the regulations (specific to the other thread) control separation spacing with 230v & current flows of less than 32 amps,  they are also designed for large installs over so much longer distances than a home hifi system & finally they allow for the final 15m of horizontal run (i.e. in the office/lab/shop room) to not be separated.      Your power cables, although able to carry 32 amps it's more like 1 amp & so very little energy is present.    

Re my last tidy up.  I have a very good idea what maybe (another maybe) caused the small change.   But before that .........  my longest ethernet run from switch to NDX that was running close & parallel to a power cable was for about 1.5m & a small section was in a conduit.    The power cable is screened & has a very tight twisted 3 core & is renowned for having very low radiation.  I have moved the ethernet out of the same conduit & it is now aprx 5cm away - the regulations require 2cm.  What I did change that maybe caused something was to move the multi-way power socket with 3x SMPS on;  it was laid on the floor & across/close to some ethernet, power & telecoms lines.   That is now moved off the floor & mounted vertically on the back of a cabinet that houses my NAS, switch & UPS.  This has not only moved the SMPS's away from the other cables but also the multi-way power sockets' internal parallel connection strips that would be radiating something.     But to summarise,  yes its a good idea to keep power cables away from all other cables,  but I'm not convinced its critical, especially with screened ethernet, or that it is your only problem.

Is it possible that the simple fact of copper wire hooking into the streamer could carry interference into the unit which affects SQ?

If so, then differences would likely be audible between shielded & un-shielded cable, and interference likely may be lessened by increasing physical clearance between the ethernet cable and others e.g. power cables.

DC71 posted:

Is it possible that the simple fact of copper wire hooking into the streamer could carry interference into the unit which affects SQ?

If so, then differences would likely be audible between shielded & un-shielded cable, and interference likely may be lessened by increasing physical clearance between the ethernet cable and others e.g. power cables.

Yes.  By a number of us, RFI is thought to be the major mechanism for SQ changes due to different Ethernet cables (hence we decorate our cables with ferrite chokes).

Simon-in-Suffolk also postulates the network stack of the media server as having an effect (and although I can't show an effect, I agree with his hypothesis), but this particular factor is unaffected by cables.

hi Mike -- thanks for taking the time to explain. i didnt mean "believe" in the religious sense -- i just meant that you seem to have demonstrated some negative effect of proximity of power (whether AC or DC) cables to net cables? i agree 100% (or should i say 'believe') that 'there are many variables'.

by the way, i dont really have a SQ problem now -- the only issue for me is how to 'domesticate' the net equipment setup -- which my daughter describes as a 'mess' followed by 'you have finally lost it' :-) Rude child!!

enjoy

ken

Ken - i'd buy yourself a small network cabinet with a couple of shelves in it and then put all the stuff in there - just put the wireless router up in the clear on one of the shelves. This will allow you to tidy it all up, use shorter patch cables and keep the rest of the family happy. Run a couple of conduits up the wall to keep the cables tidy too. 

not a bad idea James -- thanks.  i will definitely consider this. any idea where i can get one?

so, keeping switch as physically separated as possible from the wireless router was probably yet another superstition of mine. :-)

enjoy

ken

I fully support James's suggestion. I think I've posted (at some stage) a pic of my small 12U rack cabinet. It makes all the connections so much easier.
This way one can also keep power supply cables (230VAC and 12 / 24VDC) away from the signal cables.

thanks Adam, i will look out for your post. sorry to hijack this thread :-(

enjoy

ken

Hi Ken - yes Adam has a great example which i'm sure he could post a picture of again (i've seen it in another thread). Just search 'Network cabinet' and you'll find plenty of choice online.  Other than that any small shelved cabinet (ikea ?) would do - just ensure ventilation for the innards.

My suggestion for putting the Wireless router away from the other gear was to get it clear of the metal cabinet and other wiring - more for wireless signal quality and range than anything else. 

Here we go...

From top to bottom:

• Patch panel (this is where my LAN terminates)
• CISCO switch (unmanaged)
• ISP router - used only to 'feed' internet signal to the Apple Time Capsule
• Power supply for the NAS
• NAS - QNAP 251 with 2 x 6TB WD drives
• UPS / mains filter) - will support all the equipment for up to 15-20 minutes - enough to shut down the servers
• Apple TimeCapsule - sitting on top of the rack
• UnitServe SSD is connected directly to the switch, away from the cabinet (for heat reasons).

One more thing: Chord C-Stream patch cords (the ones with the bluish plugs) are used for the 'critical' connections - for the streamers.

Power supply to the whole rack is on a separate mains spur - not my decision - that how my electricians (who understand the principles of good sounding music systems as well) decided to do it. Bless 'em.

Adam Zielinski posted:

One more thing: Chord C-Stream patch cords (the ones with the bluish plugs) are used for the 'critical' connections - for the streamers.

Hi Adam,  the C-Streams are Cat7 & have the screen carried through the network via the RJ/E-45 plugs & the screened switch & patch panel ports:  my question  ....... have you or your electricians designed the network to have the interconnected screens grounded to a single (one point) ground.  This is needed to avoid the possibility of PD across the network screen as would be the case with multiple grounds.   

Mike-B posted:

Adam Zielinski posted:

One more thing: Chord C-Stream patch cords (the ones with the bluish plugs) are used for the 'critical' connections - for the streamers.

Hi Adam,  the C-Streams are Cat7 & have the screen carried through the network via the RJ/E-45 plugs & the screened switch & patch panel ports:  my question  ....... have you or your electricians designed the network to have the interconnected screens grounded to a single (one point) ground.  This is needed to avoid the possibility of PD across the network screen as would be the case with multiple grounds.   

Good question Mike - I need to check that. I've certainly not specified it that way. But knowing my electricians they have probably thought of that without actually telling me. The whole rack is connected to one, dedicated grounded mains spur. The rack is grounded to it as well (apparently by law in Poland).

However.... if I connect a streamer at the end of the network.. that is grounded itself. We have two ground connections.  The LAN run between is CAT 6.
The whole electrical network in the flat meets in one point with a common ground for all mains spurs.

I think I'm mixing 'ground' and 'earth' here... - starting to sound like an electrical philistine

Czese, OK earth & ground is a bit ....  err confusing .... blame the English language .        Naim network players ethernet port is connected to safety earth (UK) ground (USA),  but so might be your NAS & you have two grounds.  Then you have the added complication of the rack being grounded & thats on a separate power supply if I understand your description.   Not sure about the switch & patch board,  too many variables with these.

My NDX port is earthed,  but so is the NAS, its powered & earthed via its SMPS - the 12vDC switching & rectification stuff all goes on between L&N & then the 0vDC is connected to earth - that the ground/earth path for the screened NAS ethernet port;  so that gives (gave) me 2 grounds.   I fixed it with a UTP Male/Female adaptor in the NAS ethernet port.     Now I've moved on to all Cat6 & no screen ground issues.        

Mike - just had a brain wave (albeit a small one) - could that be the problem Ken C is experiencing on his network? (different thread, last week). The one you've just described above...

I cannot hear any sound quality problems on mine - the same tracks played on NDX and CDX2 via nDAC sound pretty much identical.