Ethernet cable-from router to streamer.

[krelldog]

Does an upgraded ethernet cable make an audible difference ?

 

If so...please recommend an ethernet cable between my Aurender 100 C and my router.

 

Thanks in advance.

 

krelldog

[PeterSt]

This may look like an advertisement. But trust me, it really is not.

 

1.

In the "Phasure" environment we use a two PC setup; one plays the audio (The Audio PC) and the other provides the music files (the Music Server PC). The music files are NOT streamed. Instead they are loaded into memory of the Audio PC (could be 4-5 albums at a time) and they are played from there. Meanwhile, the LAN connection between the two PC's is almost fully shut down. Only that which provides the RDC connection so the (headless) Audio PC can be controlled from the Music Server PC (via WiFi tablet etc. preceding the latter in the chain of now three devices) - remains. It is the leanest possible LAN connection during playback.

 

2.

At some stage people started to ask for making LAN cables in the same fashion as Lush^2 cables we provided at the time. Thus, configurable for the shielding. Usually after two or three people ask for something, we at Phasure start making it. After that, sort of hell breaks loose because "all" people suddenly want it.

But not me.

I hardly believe in all these matters, but, when dozens of people start to genuinely (I don't even try to sell it) report such and so improvements, I may dive in too. Btw, I am always last because I don't pay. Other people keep on having priority for quite ever.

 

3.

The ET^2 Ethernet cable emerged from this and a full year later I finally connected one between the Audio PC and the Music Server PC. Remember, that connection actually doing nothing - described under #1.

It so so much improved the sound ! ... I can't reason how this works, but it does.

It must have been a year ago that I put in that ET^3. But not long after - thus a year ago - people started to ask for an extra shield and the ET^3 was born. People buy it and buy it. I am still on the ET^2. As usual I am behind. Even while I know it will work for the better ...

 

End-of-not-advertisement (ahum).

 

Please keep in mind that when the Ethernet cable will be used for streaming, like in your situation, the differences will only be larger. Say as large as a Lush^2 or Lush^3 will imply (very huge differences per config). With my ET^2 in that belittled situation, it became immediately apparent that the sound became much more smooth because of it. Think more silk. This is not how a "direct" protocol cable like USB will exhibit it. That changes stages and depth and holographic's and other natures. I don't see this happening with my #1 described connection. But with yours it well could be. You'd need to try it. And send back if it doesn't work out ...

 

Peter

[Miska]

8 hours ago, krelldog said:

Does an upgraded ethernet cable make an audible difference ?

 

Just make sure it doesn't have a shield that would connect grounds of the two devices. Since this would spoil the galvanic isolation Ethernet otherwise has through transformer coupling. So for example standard U/UTP CAT6 is fine. So the cable should have plastic connector bodies.

 

[AnotherSpin]

Meicord ethernet cable from switch to streamer works well for me in comparison with noname and Supra used before.

[audiobomber]

4 hours ago, Miska said:

 

Just make sure it doesn't have a shield that would connect grounds of the two devices. Since this would spoil the galvanic isolation Ethernet otherwise has through transformer coupling. So for example standard U/UTP CAT6 is fine. So the cable should have plastic connector bodies.

 

Do all CAT7 and CAT8 cables break galvanic isolation? I believe all are shielded.

 

Is the shielding a problem if both devices are plugged into the same conditioner, i.e. same ground potential?

[Miska]

9 minutes ago, audiobomber said:

Do all CAT7 and CAT8 cables break galvanic isolation? I believe all are shielded.

 

Not all CAT7, but CAT8 yes. Luckily those are not needed for audio, since you can reach 10 Gbps on CAT6A already. And most audio devices are just 1 Gbps anyway.

 

11 minutes ago, audiobomber said:

Is the shielding a problem if both devices are plugged into the same conditioner, i.e. same ground potential?

 

Generally yes, many audio devices have floating PSU that doesn't have 3-pin mains connection.

 

[audiobomber]

27 minutes ago, Miska said:

 

Not all CAT7, but CAT8 yes. Luckily those are not needed for audio, since you can reach 10 Gbps on CAT6A already. And most audio devices are just 1 Gbps anyway.

 

 

Generally yes, many audio devices have floating PSU that doesn't have 3-pin mains connection.

 

Thanks Miska. Are all cords with metal connectors a problem, or could the shield still have broken conductivity? I assume connecting an ohmmeter between the connectors would be a way to tell?

[Miska]

3 minutes ago, audiobomber said:

Thanks Miska. Are all cords with metal connectors a problem, or could the shield still have broken conductivity?

 

Standard compliant STP cables do have the shield ends connected to the connector body. Standard UTP cables have plastic connector bodies.

 

3 minutes ago, audiobomber said:

I assume connecting an ohmmeter between the connectors would be a way to tell?

 

Yes, that would tell.

 

[audiobomber]

I just tested my three CAT7 cables. Two (Audioquest Pearl and Tera Grand) have isolated shields, the CableCreation flat cable has a connected shield. 

 

I was not able to hear ethernet cable differences in my previous setup (sMS-200 and Audiolab 8200CD). To my surprise, I was able to hear differences with the current system (exaSound Playpoint and e32). I tried various generic CAT5e, 7 and 8 cables and liked the Tera Grand best. I will continue to use one of the CAT8 cables, but I have a LAN Isolator on order. I will do another round of cable testing with and without the isolator when it arrives in May. 

[audiobomber]

15 hours ago, krelldog said:

Does an upgraded ethernet cable make an audible difference ?

 

If so...please recommend an ethernet cable between my Aurender 100 C and my router.

 

Thanks in advance.

 

krelldog

I suggest you connect an ethernet switch between your router and the Aurender. That made a more significant difference than ethernet cables in my system.

 

Pretty much everything I've read suggests not to use the router as a switch. There are lots of good inexpensive options available from TP-Link, Netgear, Linksys, etc. For a further improvement, you can use a linear PSU on the switch. Or go whole hog and buy an audiophile switch.

[intothedragon]

16 hours ago, Miska said:

 

Just make sure it doesn't have a shield that would connect grounds of the two devices. Since this would spoil the galvanic isolation Ethernet otherwise has through transformer coupling. So for example standard U/UTP CAT6 is fine. So the cable should have plastic connector bodies.

 

Hello Mr. Miska,  I'm getting electrical interference from the ethernet cable connected to my Node2i streamer. Every time I turn on any lights or the motor of the fridge turns on, the Node2i makes a click and cuts the sound while connected to the ethernet cable, no issues on the Node2i using wifi. I just realized that all my ethernet cables are shielded with metallic terminations. Would it be the cause of this electrical interference? Changing all my cables for UTP with plastic terminations would solve the problem? Also, would a Network Surge protector like this one Delock Network Surge Protector 6 kV RJ45 do any good in this situation? thank you

[PeterSt]

38 minutes ago, intothedragon said:

I'm getting electrical interference from the ethernet cable connected to my Node2i streamer. Every time I turn on any lights or the motor of the fridge turns on, the Node2i makes a click and cuts the sound while connected to the ethernet cable,

 

Although it may change things, your problem most probably will be a missing ground somewhere. For example, the ground in a balanced (XLR) cable (loose/broken wire).

 

Groeten uit Grollo Epe.

[davide256]

On 3/28/2021 at 8:49 PM, krelldog said:

Does an upgraded ethernet cable make an audible difference ?

 

If so...please recommend an ethernet cable between my Aurender 100 C and my router.

 

Thanks in advance.

 

krelldog

Don't worry to much about Ethernet cables. Data never leaves the digital data domain on an Ethernet cable, your bank would be screwed if that data could be messed

up/no longer correct.  Worry more about what voltage noise your switch/router is injecting over the Ethernet port into the Aurender.

[intothedragon]

19 minutes ago, PeterSt said:

 

Although it may change things, your problem most probably will be a missing ground somewhere. For example, the ground in a balanced (XLR) cable (loose/broken wire).

 

Groeten uit Grollo Epe.

I hope this is not the case, also I double checked all the cables and connections, from routers, modem, all power cables are brand new, I just got a new HiFi system, I'm using Brennenstuhl Secure-Tec Power strips on all my devices, different ones not stacking to many plugs on each strip. Also, this problem with the clicks and sound cuts only happens when connection the streamer to the Amp on the Digital Coaxial output, when I connect using the RCA or Optical nothing happens. I'm pretty sure the electrical noise is coming from the shielded CAT7 FTP cables. 

[John Dyson]

42 minutes ago, davide256 said:

Don't worry to much about Ethernet cables. Data never leaves the digital data domain on an Ethernet cable, your bank would be screwed if that data could be messed

up/no longer correct.  Worry more about what voltage noise your switch/router is injecting over the Ethernet port into the Aurender.

I agree with you -- the big risk is un intentional ANALOG leakage into the system.  Ethernet cables can participate in that problem, given a poor component design, but shielded cables can mitigate some troubles.

 

There is NO WAY to transmit any differences from the digital domain into the analog domain, unless there are bona-fide errors (or in the case of ethernett, too many collisions causing time delays.)   Any risk of random noise comes from the analog domain.   Using high quality, shielded, standard ( BOUTIQUE, elite, rebranded not needed) cables and well designed HW is good enough.   Go to industrial suppliers, get their best quality stuff -- well tested, and use what the pros use.

[PeterSt]

3 hours ago, intothedragon said:

I'm pretty sure the electrical noise is coming from the shielded CAT7 FTP cables. 

 

I know that intuitively is prone to be the case, but there is no reason to, electrically. Anyway, can't you easily test it with an "unshielded" cable (plastic connectors) ? Also, is it actually so that both ends (both devices) have metal housings for the chassis parts (where the connector goes in) ?

 

Did you try to put either or both devices into mains earth ? (or just play with it into the other way around as that you currently have it)

 

Do you have another (lousy) coax cable laying around, just to try ?

 

 

[Miska]

7 hours ago, intothedragon said:

Would it be the cause of this electrical interference? Changing all my cables for UTP with plastic terminations would solve the problem?

 

Possibly yes... Especially given that Node 2i is floating PSU device. But it is important not to establish a ground path also from it's input side, even if it gets one from the output

 

7 hours ago, intothedragon said:

Also, would a Network Surge protector like this one Delock Network Surge Protector 6 kV RJ45 do any good in this situation?

 

Spike could also leak through the PSU, but if things work with WiFi, then it is less likely to be the reason.

 

7 hours ago, intothedragon said:

Also, would a Network Surge protector like this one Delock Network Surge Protector 6 kV RJ45 do any good in this situation?

 

Those require pretty high voltage spikes to do anything. But it seems to have plastic connectors, so it likely at least breaks the ground connection through network cable shield and could fix the problem that way...

 

[R1200CL]

In near future, hopefully we standardize on singel mode fiber on any type of digital transfer. External clock may be embedded or daisy chained as well, in order to achieve the last 2% improvement. 
 

Today the $99 Mikrotik is cheaper than a good cable. Optical cable is $5. Just saying.

 

@krelldog

As others has indicated, a better cable won’t help. Or be worth the money. I do buy JSSG360 cable myself. ($78). Just so I sleep better. Better maybe test out, or consider fiber. That will at least remove any electrical connections that possible can cause anything.

 

Adding the EtherRegen between your noisy? router and the Aurender should have a positive effect, and you will get your money back if not. Be warned, you won’t return it 😃

 

 

[intothedragon]

6 hours ago, PeterSt said:

 

I know that intuitively is prone to be the case, but there is no reason to, electrically. Anyway, can't you easily test it with an "unshielded" cable (plastic connectors) ? Also, is it actually so that both ends (both devices) have metal housings for the chassis parts (where the connector goes in) ?

 

Did you try to put either or both devices into mains earth ? (or just play with it into the other way around as that you currently have it)

 

Do you have another (lousy) coax cable laying around, just to try ?

 

 

Yes the Node21 has a ethernet metal house. I'm not finding a way to ground the Node2i but the power cable provided was that flat double insulated, maybe if I change for the EU standard round one I'll get a better grounding?  

Yes I already tried many different coax cables from cheap ones to branded ones and nothing changed. What is curious is with RCA cables that are metal terminations too the issue goes away but since I don't want to use the inbuilt DAC from the Node2i I only have the option of optical where I won't get the higher bitrates from MQA or Coax that is the main issue. 

[intothedragon]

4 hours ago, Miska said:

But it is important not to establish a ground path also from it's input side, even if it gets one from the output

how I would do that, please?

[intothedragon]

6 hours ago, PeterSt said:

 

I know that intuitively is prone to be the case, but there is no reason to, electrically. Anyway, can't you easily test it with an "unshielded" cable (plastic connectors) ? Also, is it actually so that both ends (both devices) have metal housings for the chassis parts (where the connector goes in) ?

 

Did you try to put either or both devices into mains earth ? (or just play with it into the other way around as that you currently have it)

 

Do you have another (lousy) coax cable laying around, just to try ?

 

 

another thing, yes I already tried using the last part of the ethernet, the one connected to the Node2i with a UTP plastic termination cable and the problem stays, I'm not sure if the other FTP cables is the reason. 

[PeterSt]

39 minutes ago, intothedragon said:

I'm not finding a way to ground the Node2i but the power cable provided was that flat double insulated, maybe if I change for the EU standard round one I'll get a better grounding? 

 

No, that is not related. But the fact that the device does not provide earth (using the clips in the wall socket) encourages for problems (OK, not problems, but you'd have less solution possibilities). Also see upcoming post.

[PeterSt]

39 minutes ago, intothedragon said:

5 hours ago, Miska said:

But it is important not to establish a ground path also from it's input side, even if it gets one from the output

how I would do that, please?

 

What Miska tells is that a "floating" PSU is a PSU which formally does not require any outside ground. It creates its own ground (think of a middle between plus and minus voltage - the middle is a "ground reference"). Now:

 

The ground reference will be a first "opportunity" for ground to not match the ground (reference) of a connected next or previous device in the chain. Thus, if one device references to mains earth and the other creates its own ground, then both almost certainly will not be equal and a current will start to flow somewhere where it was not intended. This can be over your coax shield, for example, that molesting the signal of that connection.

 

The dumb thing is that these issues are solved with explicitly MAKING the extra ground connection (over Ethernet). Now I also have a nice reason to show our cable, regarding this:

 

   

(connectors are plastic !)

 

 

 

So these extra wires (with the eye) are to provide that ground you initially wanted out. Just in case because it is up to you whether to connect them (at each side).

 

This may also give you the feel of how to provide an extra ground connection, whether from the device to the next device or to the previous device, just provide that extra "wire". Some devices have provisions for this (so the eye you see can connect to the designed position), for some you can use a screw in the chassis (see the crocodile wires). But some devices have connected nothing to the chassis, and then nothing will work really and it will be signal ground providing the chassis-chassis connection (never good).

 

It depends on so many things. And it can well be that you have a device in the chain that just won't "cooperate" regarding this all.

 

For now I stay put on the idea that somewhere is ground is missing. And IIRC I ran into situations that this is not even in your audio chain. so think of a missing wire in your low voltage led. You connect the hot wire with the switch but the ground return is missing. But your power amplifier (or anything) can provide it.

a. CLICK !

b. Poor sound.

 

It will be a matter of reducing and deducing. You already did that. Try to think of some more steps. Maybe you can plug in headphones and eliminate a lot (preamp, poweramp (, speakers)).

[plissken]

Here we go again. Maybe a sticky for this type of question:

 

1. You shouldn't used chassis tied shielded cables. IE they make a common ground connection between switch and endpoints on that switch. An end point is any multiple devices that have a ground connection over shield to the switch. Switch being the 'hub' and endpoints being 'spoke'

 

2. You can use a cable with a floated shield. That is a shield that is only there when we are running 300 cables in a bundle and want to reduce inter-cable cross-talk.

 

3. You should use UTP Cat5e or CAT6. Both are good for 10GBe applications. 147 foot for CAT5e, 180 foot for CAT6. CAT6A by standard are supposed to be shielded. Don't use them for anything else other than non-audio applications. They are good for 10GBe at 328 foot.

 

4. Use a quality BJC, Panduit, Tripplite cable. All are guaranteed to spec out and meet IEEE / TIA standards for Skew, NeXT, Interpair crosstalk, propagation delay.

 

5. There is nothing but anecdote that cables of a Category (5e or 6) make an audible difference. The issue being is that the standards that the cable has to meet determines the design. Doesn't matter if it's silver coated or CCD or ETP. They all hit air gapped magnetics on BOTH ends. At the switch and at the end point.

 

6. Empiric evidence on everthing from $30K audio analyzers to $100K Agilent scopes show us that cables that meet spec simply meet spec and do nothing magical for us.

 

[PeterSt]

24 minutes ago, plissken said:

6. Empiric evidence on everthing from $30K audio analyzers to $100K Agilent scopes show us that cables that meet spec simply meet spec and do nothing magical for us.

 

25 minutes ago, plissken said:

Here we go again.

 

Yes, it is almost USB. Nothing makes a difference whatsoever.