Galvanic isolation of your DAC

Glad we agree, and no, you didn’t.

I’m interested in the concept of a ‘quality switch’. Since these devices are built for optimum data throughput and not for minimal electrical noise at the level that might spoil an audiophile music renderer I strongly suspect that a ‘quality switch’ will be good at the former but poor at the latter: even the good ones will almost certainly use a SMPS—perhaps the most hated of electrical devices amongst audiophiles!

So my next thought is… who’s using which linear PSUs with which optical isolators, and how does their sound quality compare with a GISO-like device. A comparison with no galvanic isolator will be of interest to the users themselves but of course of little use to the rest of us because of our agreed assumption that there are significant differences in the amount of electrical noise being injected into music renderers from all of those different network setups out there.

… and I’m off to read that Meridianunplugged forum topic on the same subject.

emailists said

AMG - are you saying the Ethernet goes direct from the Mac to the bridge ? My Mac pro has two ethernet ports but I connect the Mac to a router with wifi built in and that handles the modem and connection to the BII. I would think the Mac is just as noisey as the router and should be isolated as well.


I would have thought the same thing, that there would be significant benefit to using optical fiber between the Mac and the DS Sr. Nonetheless, my results jive with lots of others who tried optical fiber runs while using the bridged ethernet direct connection from their server to their renderer on computeraudiophile.com. No significant difference.

amgradmd said

… my results jive with lots of others who tried optical fiber runs while using the bridged ethernet direct connection from their server to their renderer on computeraudiophile.com. No significant difference.


Elsewhere in this thread a couple of us have agreed that what brings about a change in sound quality (or alternatively fails to do so) when done to one persons network may work differently in somebody else’s network. That’s not to discount your assertion that it didn’t work for you.

But it is a call for others to describe their experiences with optical isolation [of the last network leg, whether from a switch (stand alone or built in to your broadband router) or directly from your server].

Regards,

Peter

I’m planning on adding FMC to my network connection this weekend. I’ve bought an arsenal of parts from Amazon and prepared myself for network battle this weekend. Of course, just as I was confident in my plan, I got to thinking - would putting an Eeero mesh wifi unit in my audio room and hooking up an Ethernet cable to it’s output accomplish the same thing?

If so - Might there be any potential benefit in using some sort of upgraded power supply on the Eero?

I’m only bringing this up because I already have these units in my house, my WIFI is blazing fast and I’m wondering if this is a far less labor intentensive/parts intensive solution.

“Subscribed!”

We pass the Ethernet cable Five Times Through the Ferrite Corewink

For reference

I’ve finally added a fibre-optic link as close to my DirectStream Junior as possible.

A pair of new Comnet CNFE1005M2 fibre media converters (FMC) were on eBay in the UK at a ridiculously low price so I grabbed them. The normal price is high as these are professional modular units: so I’m not recommending them from the price point of view. I bought a short length of fibre-optic link from Amazon UK (‘TPFNet High-Quality 0,50m Fiber Patch Cable’ - each one comes with an individual test certificate).

At the back of my AV furniture is a network switch¹ (Netgear). The prior arrangement had been a network patch cable—screened cat.6—between that switch and my DSjnr. In place of that there is now a patch cable to a FMC, the fibre-optic link, the other FMC and a very short patch cable to the DSjnr. The power supplies for the two FMCs are still the original switched mode power supplies (SMPS) as supplied but I shall dig out a couple of suitable old-fashioned, transformer-based power supplies from my box of such redundant things and use them for the FMCs.

There was no setup other than connecting the various bits together.

It works.

I’m not one for fancy words to describe sound quality and I’m well aware of expectation bias so I’ll just say that my impression is that something has got better and so I’m sticking with the new arrangement.

In due course I’ll get a high quality, linear power supply just for the FMC that connects electrically to the DSjnr.

By the way the only other input to the DSjnr is via TOSlink (optical).

Enjoy the music!

Peter

¹ I’d already replaced the switch’s original, switched mode power supply with an old-fashioned, transformer power supply to reduce noise fed back into my mains and in the hope of reducing noise fed into the DSjnr through its network socket.

Peter said

. . . my impression is that something has got better . . .


This is all that matters.

If the improvement I think I can hear is caused merely by my expectation bias yesterday evening then if other people take my view as a statement of fact there’s a risk that those other people may do something similar in their own systems and perhaps waste their time and money. I acknowledge, though, that the this is a hobby for most and so time spend adding, removing and exchanging equipment, bits and pieces may be enjoyed whatever the outcome.

For me what matters is that my impression that sound quality has improved persists when I next swap back and forth between the pair of linked FMCs and a single network patch cable, Either way I’ll report and so hope to add to the body of evidence one way or another.

Enjoy the music.

Peter

Everyone here is well aware of expectation bias, as well as system dependency, personal preferences, etc.

You need not worry anyone will take your subjective reports as absolute truth. You are just yet another random guy on the Internet. :slight_smile:

Given the very enthusiastic endorsements made by some of a tweak just made, I’m not convinced that they all are [aware]!

Just another random guy on the Internet? Perhaps. But we aren’t all equal, in knowledge, experience, intelligence, hearing or, indeed, expectation bias. It’s dangerous to try to judge oneself so I try not to. Sometimes I observe others with less reticence.

Enjoy the music.

Peter

Nothing personal; we are all random guys on the Internet.

It is a hobby of small distinctions. Thus, people often appear excessively enthusiastic over what is, in reality, a minute change. Have you witnessed a drag racer enthuse over hundredths of a second? They make us appear calm and rational.

If you are uncomfortable making judgments this is fine. Others love to dive in. All are welcome.

So, anyone here honestly unaware of the oft-cited concept of expectation bias? Raise your hands . . .

Bueller?

He took the day off.

Elk said

It is a hobby of small distinctions. Thus, people often appear excessively enthusiastic over what is, in reality, a minute change. Have you witnessed a drag racer enthuse over hundredths of a second?


The minute changes that so many think are the best thing since sliced bread are what drive me crazy about this hobby. Maybe I’m too rational … As for drag racers, similar, but not the same. A hundredth of a second could be the difference between winning and losing, whereas in our hobby it isn’t a race (repeat, it’s not a race). I think you do understand that …

Agreed on all counts, it is easy for many to become irrationally obsessed.

Yes, as with all similes, the analogy fails at various points. Then again, there are plenty of audiophiles that view building a system as a competition of sorts.

pmotz said
Elk said It is a hobby of small distinctions. Thus, people often appear excessively enthusiastic over what is, in reality, a minute change. Have you witnessed a drag racer enthuse over hundredths of a second?

The minute changes that so many think are the best thing since sliced bread are what drive me crazy about this hobby. Maybe I’m too rational … As for drag racers, similar, but not the same. A hundredth of a second could be the difference between winning and losing, whereas in our hobby it isn’t a race (repeat, it’s not a race). I think you do understand that …


Well said.

Peter

I was reading through this thread, and it’s been quiet for a while, but it gave me an idea which I thought I’d share.

Somebody mentioned the idea of using Wifi for isolation, and that made some sense to me.

I had an old Sonos Bridge laying around. They are deprecated and can be had for next to nothing, but still work great. If you’re running a Sonos network the Bridge acts as a 2-port wireless-to-wired bridge in addition to extending your Sonos network, and it’s super easy to use.

I also had an extra 9V linear power supply to power the Bridge (it wants 5V but that’s fine), and one of these DC purifiers to run that through: https://ifi-audio.com/portfolio-view/accessory-dcipurifier/

The two Ethernet ports are hooked up to my Intel NUC (Roon Core) which has its own LPS, and to my DAC.

So now I can pull my ethernet switch and cables and power supply out of my audio rack, and my NUC and DAC are isolated from everything but each other.

I can’t say what kind of difference it made to SQ if any. Everything was already on LPS’s with CAT8 STP cables. Maybe I’ll say something later about that, for now it’s just an experiment I thought I’d share. I hope it’s helpful.

–Matthew

Welcome, rechsm!

do not use shielded patch cables and you will be fine. http://www.crazy-audio.com/2013/10/tracking-down-noise-sources-on-a-raspberry-pi/

the best network cables for audio are f/ftp types. cat 7 (and up) cables (s/ftp) are meant for heavy traffic over long distances in data centers etc where both ends of the cable are connected to grounded devices.