Do I really need an " Audio Grade Network Switch "? Part 2

[Psilonaught]

Given the previous thread was rudely terminated I've created a part 2.

So I've been experimenting with various ethernet methods for connecting my Aqua LinQ to the network. I've found that a stock Zyxel 8-Port sounded worse that just using a direct connection to my TP Link WiFi repeater.

So I can definitely hear how a switch influences sound.

I've gone ahead and looked at the specs for an Innuos phoenixnet and found this unit from Jay's Audio.

https://www.beatechnik.com/product-page/lhy-sw-8

Highly accurate OCXO with dedicated linear power supply rail
Dedicated additional power rails for other parts of the switch
Milled aluminium case.



I'll happily report my findings once I receive it. Not suggesting it is as good as a phoenixnet but user reviews on various forums suggest is a highly capable switch.

 

[Cable Monkey]

Pretty much the standard mods to this kind of switch. Standard ‘business grade’ switch should have better iron in the isolation transformers which will be at the heart of any improvements. Six dollars each. Then the widely recognised (in our hobby) clock and PSU upgrades. Finally the milled case to give it some heft. Always creates a good initial impression when you hold it for the first time. So plenty of what they have done is formulaic. The big question is does the formula work? Or will your straight connection still win out. 🍿

 

[Psilonaught]

Well as I have a NUC in the chain with direct ethernet connection I need a switch, so this seems the least bad option.

 

[Cable Monkey]

Well as I have a NUC in the chain with direct ethernet connection I need a switch, so this seems the least bad opportion..

I don’t disagree. My musings about the impact of networks on audio are on these pages and date back two or three years.

 

[TheFlash]

Given the previous thread was rudely terminated I've created a part 2.

So I've been experimenting with various ethernet methods for connecting my Aqua LinQ to the network. I've found that a stock Zyxel 8-Port sounded worse that just using a direct connection to my TP Link WiFi repeater.

So I can definitely hear how a switch influences sound.

I've gone ahead and looked at the specs for an Innuos phoenixnet and found this unit from Jay's Audio.

https://www.beatechnik.com/product-page/lhy-sw-8

Highly accurate OCXO with dedicated linear power supply rail
Dedicated additional power rails for other parts of the switch
Milled aluminium case.



I'll happily report my findings once I receive it. Not suggesting it is as good as a phoenixnet but user reviews on various forums suggest is a highly capable switch.

That milled aluminium case is nice, less pervious to RFI ingress than the PhoenixNET steel case actually. Power is good. As pointed out before, clock accuracy in the ethernet world is a red herring if suggested it has any effect on sound quality. If, like Innuos, you have a ready supply of them because you use them in the digital side of the house then why not use same. But to suggest that the change of clock to OCXO in a switch is an “upgrade” is misleading. Completely different in the streamer/DAC side of course!

 

[TheFlash]

So I've been experimenting with various ethernet methods for connecting my Aqua LinQ to the network. I've found that a stock Zyxel 8-Port sounded worse that just using a direct connection to my TP Link WiFi repeater.

Sorry if this has been covered in previous thread but I’m intrigued as I think this is the first time I’ve heard a negative report:
- length and type of cable from repeater to switch?
- length and type of cable from switch to LinQ?

Thanks

 

[tuga]

Sorry if this has been covered in previous thread but I’m intrigued as I think this is the first time I’ve heard a negative report:
- length and type of cable from repeater to switch?
- length and type of cable from switch to LinQ?

Thanks

Personal preference may be playing a part here, even expectation bias.
That's often the case when opinoins diverge.

 

[TheFlash]

Personal preference may be playing a part here, even expectation bias.
That's often the case when opinoins diverge.

The first thing I would like to do when someone has what sounds like an outlier of an experience is to understand whether they installed thing X the same way others do/did. It could be personal preference, it could be that something else is going on, it could be that it's objectively worse for some reason (though I'm not sure how we'd know!).

 

[Maverick]

Personal preference may be playing a part here, even expectation bias.
That's often the case when opinoins diverge

Correct the length of cable in a domestic digital environment is utterly trivial given the distances the various standards define. Unless you own a large country estate that is, I don’t.

 

[Cambs12]

Correct the length of cable in a domestic digital environment is utterly trivial given the distances the various standards define. Unless you own a large country estate that is, I don’t.

You dont?! 😲

 

[Maverick]

You dont?! 😲

Sadly there will me no King Paul anytime soon. Although wasn’t there a country where the DoE (RIP) was considered a god? Maybe position vacant?

 

[Cambs12]

Sadly there will me no King Paul anytime soon. Although wasn’t there a country where the DoE (RIP) was considered a god? Maybe position vacant?

You'll have to be chancellor first.

 

[TheFlash]

Correct the length of cable in a domestic digital environment is utterly trivial given the distances the various standards define. Unless you own a large country estate that is, I don’t.

This is completely incorrect, Paul.

If the switch is to have any impact on sound quality at all it must be installed just before the streamer rather than just after the router. This is a fact not an opinion. If by "digital environment" you mean the digital signal, it's nothing to do with that - no bits are claimed to be saved or corrected or other mumbo-jumbo - it's to do with reducing noise using the galvanic isolation inherent in the design of any ethernet circuit board you can think of.

The hifi enthusiast's use of a switch simply exploits a "side effect" of this design to the benefit of sound quality - for those who choose to have a wired rather than wireless system. If you install a switch as originally intended - just behind the router to give you more ports - then the only noise it will stop reaching the streamer is the noise from other devices attached to the router. That might be of benefit but nothing like installing it just before the streamer.

That's why I asked about cable lengths. It's pivotal to performance.

See you soon.

 

[robin]

Sorry Nigel, I think you could be incorrect,

 

[Psilonaught]

It's pretty obvious that a dinky little £17 switch would sound worse than NO switch. Hardly a revelation.

My interest is in if the Jay's Audio switch can either offer no detraction of sound Vs no switch or indeed improve the sound being outputted from my WiFi repeater (which will invariably be noisy) by cleaning it up.

I now need a switch as I've bought a fanless i5 NUC and good quality linear power supply and I want the NUC connected physically into the same network as tye LinQ.

 

[Maverick]

This is completely incorrect, Paul.

That's why I asked about cable lengths. It's pivotal to performance.

So I mention cable lengths but you talk at length about ”reducing noise using the galvanic isolation” (which I never mentioned but is another red herring) then assert without ANY evidence that cable lengths are ‘pivotal’, something that the standards bodies do not support, well not at the cable run lengths used in a domestic environment.

 

[Psilonaught]

This ethernet malarkey is a path to mucho money wastage I worry...

Listened to my system last night via my new (old) NUC, connected via its switching power supply and through a £10 Netgear prosafe switch with switching power supply I found in a drawer and it all sounded excellent...

Will there be much audible improvement with better NUC power supply and "hifi" switch, who knows! Items arriving in next week.

I will focus on buying new tubes next...a more reliable upgrade path me thinks.

 

[antonio66]

Will look forward to your thoughts as and when your Jays turns up. Which LPS have you purchased, and I assume you will try one of them first before inserting the other in the system. Strangely I was looking at the Jays on Saturday when a AliExpress 'take a look' email came through.

 

[Chumpchops]

What length tubes

So I mention cable lengths but you talk at length about ”reducing noise using the galvanic isolation” (which I never mentioned but is another red herring) then assert without ANY evidence that cable lengths are ‘pivotal’, something that the standards bodies do not support, well not at the cable run lengths used in a domestic environment.

90m ethernet design max run, before repeaters etc are needed, it’s 5m for local patch cables.

 

[Maverick]

What length tubes

90m ethernet design max run, before repeaters etc are needed, it’s 5m for local patch cables.

so I'm probably OK with standard cable at 0.5m