English Electric 8Switch

Streaming Audio
326

Cats, Nags, Roosters - this thread has suddenly got interesting.

I worked at English Electric during my summer holidays - it was the same year that The Who released “Won’t Get Fooled Again”. They were a great company and merged with GEC. If only we could turn back time.

It seems that in the context of the audiophile, FR’s Cheshire Cat may be apt as Britanica says

The Cheshire Cat is a member of the Duchess’s household. When it appears before Alice, it assures her that everyone in Wonderland is mad and that it doesn’t much matter which path she follows if she doesn’t care where she’s going. The cat’s ability to vanish slowly by sections disconcerts Alice, who at one point finds herself in the presence of only the cat’s grin.

331

Yes short list by objective measurements and make the final decision by personal audition - works for me. Beware of external influences. Once you have found something good stick with it even if there is something better just around the corner.

I can’t see anything wrong with these switches and as they are made on a small scale then they will be expensive. So let the lamb lay down with the lion.

334

So we will have …what’s for dinner ( thread ) photo… :slightly_smiling_face:

2 Likes
335

@Peder, I of course always respect your expertise, I just used the words “more savvy” to invoke the young spirit of the single-mindedness of a Viking!!!.

1 Like
336

I’m curious which Auralic streamer you have and if you’ve ever compared it Wired vs WiFi ?

337

Hi the Cisco Catalyst devices main benefit for home audio for connected audio devices on a normal home LAN is its support for IGMP snooping as default. This keeps unrequited group data away from streamers on your LAN thereby reducing unnecessary processing overhead by your streamer and reducing the noise floor in the streamer. For many people reducing the noise floor relates to better or at least different SQ.

The Catalyst devices appear also designed to an EMC standard and PHYsical media layer termination standard that offers a better noise performance and thereby for some better SQ compared to some other switches. Certainly the former is what I would expect for commercial devices more likely to work in more sensitive environments compared to the domestic environment.

Finally for UPnP and possibly Roon, (and some home automation apps) the Catalyst devices can be configured to act as an IGMP querier. This helps home networks where the router is not running IGMP, and help keeps device multicast group memberships fully maintained without relying on less reliable application querying. This helps responsiveness and speed of discovery (almost instant) on the Naim app for example. This does require the user to log into the switch to configure.

Outside of audio there are two other main benefits for the home user … the Catalyst switches offer loop protection, thereby preventing you taking your network down by accident if you have created a loop, this can also be used for resilience uses. The switches also offer full diagnostic capability and can test connected Ethernet cables or fibres. This requires the user to log into the switch.

There are many other benefits such as configurable support of trunk, VLANs, port QoS, Ether-channel, SNMP and port security … but these are less relevant perhaps for the home user.

BTW I am not really an authority on things Cisco… I just happen to engage with the company and design with some of their products in my professional world.

For me the health and configuration of my home network is important as I rely on it for many things, including work, home audio, IPTV etc. Devices like these Catalysts give me the control and visibility I need to help maintain my home network as optimum as I can make it for the purposes I need it for. Many standard consumer devices give no visibility or control here giving no no real status of the health of the underlying network. By design TCP/IP will try and work as the network gets less and less healthy until it stops all together… this is often invisible to the user with many consumer devices.

5 Likes
338

I wrote this yesterday and then the thread disappeared.

Anyway, Stephen and anyone else who’s interested, I’ve had a few days now with the EE switch. Here’s what i’ve found.

Present setup –

Roon Nucleus (internal SSD for local music storage) + Qobuz (formerly Tidal) streaming service. Cisco 2960 8TC (blue).
25m CAT5e to lounge (Devialet 220 Pro)
20m CAT5e to bedroom (Project Streambox S2 / Mytek Brooklyn DAC+)

The Nucleus and the two systems connect to the 2960. The uplink port on the 2960 connects to my TP Link Deco Mesh router for internet and wireless connectivity. The 2960, Nucleus, Router and Virgin modem live in the ‘office’ along with a Sonos Bridge for the various Sonos devices around our house.

I’d previously been using a Melco N1A as my UPnP server and had been very happy with this setup for a few years. At the start of 2019 I’d tried Roon, found it was the control point I’d been looking for all along to make the most of my library after many years of playing with computer audio. Suitably impressed, i bought a lifetime subscription along with a Roon Nucleus. To make all this work I needed an additional switch so ordered a used 2960 based upon advice on this forum. Tidal soon followed and having access to not only my library, but a vast streaming library with no loss of quality between local and streaming sources was extremely liberating.

Wanting to cut down the number of boxes now occupying the rack, I experimented by comparing the player port output on the Melco to using one of the ports on the 2960. I found a slight preference when using the 2960 direct which meant I could move the Melco on (I was now only using it for its player port, music storage having moved to the Nucleus). I still had a Netgear Switch (a GS105 with the older linear supply) feeding the lounge from the office where my router / modem is also situated.

A few months on, I added a bedroom headphone system initially based on my Sony 1Z Walkman but wanting access to Roon meant adding a USB endpoint – I chose a Project Streambox. Initially I used the wireless connection but drop outs playing hi-res material meant adding a hardwired connection. I ran a new cable in (standard CAT5e) terminating at the Netgear switch in the office. When I tried out the new connection, I found it sounded worse than the wireless connection I’d initially used the Streambox with. Having been very careful about the new cable routing (avoiding mains cables or crossing at 90’ where unavoidable) and having checked and rechecked the terminations were correct, this was obviously rather disappointing after the effort to put the cable in. As a final sanity check, I borrowed the 2960 from the lounge and slotted it in place of the office Netgear. What a difference – the Streambox really worked well with the 2960. As a bonus I found my lounge system took a step up too with the 2960 sat at the end of 25m of CAT5e rather than the 1.5m of CAT6 (a standard patch lead from Farnell) so the 2960 made a full time move to the office.

This new setup worked very well and to be honest, apart from some Roon updates, I’d not touched anything until a few weeks ago when a couple of changes (non Hi-Fi) in the lounge required a few more Ethernet ports to be available. Routing more cables in would have been a pain (I did the original run which is still in good condition 10 years ago when I first put in a Linn Akurate DS, my first network player) so I thought I’d try putting in a switch instead. My original Netgear GS105 (with linear PS) made things worse and I didn’t really want to put in another 2960 with a short cable to the Devialet . As I mentioned earlier in the thread, I have no illusions that the EE is just a tricked up ‘off the shelf’ switch in a nice box but with a generous return policy, a tiny footprint and a more lounge friendly appearance, I purchased one and have had it in my system for the last 4 days…

I’d like to say it’s taken my system to a whole new level but it’s more subtle than that – The start of Stimela (Hugh Masekela) where the drums grow in intensity, I now find the individual drums easier to hear as the drumming reaches its climax. The Unthanks harmonising on Magpie, their voices are clearer and easier to differentiate. These are just a few examples where I think it’s adding something good. The main thing for me is the EE switch doesn’t degrade the sound. Putting the Netgear in the same place introduces a harshness to music playback that initially sounds like more detail but very quickly becomes tiring. I’m not that good at describing musical differences, so for me the main sign that something is working well is that I’ll find excuses to spend more time listening, it’s as simple as that.

Again I’m under no illusion that these differences (or lack of effect) are down to some unique property of an ‘Audio Grade’ switch but are down to subtle electrical interactions further down the chain as digital signals convert to analogue and I just find these interactions more agreeable in the final musical result. I suspect that with a bit of time and effort I could find a combination of switch and cables that does the same but I don’t really have the time or urge to fiddle around trying to do this. For this reason the EE switch (even at £450) will be staying as will my existing 2960. When I get more time I’ll try the EE switch in front of my Streambox which seems to be more sensitive to network switch changes just to see if the effect is more pronounced but I’m in no rush to do this.

So there you go – my opinion, my system, my ears. If you’re looking for a new switch then it’s certainly worth a go and I’m sure mileage will vary.

I’ll be interested in what others find.

19 Likes
339

Thanks James, nice write up👍
I know with the Melco switch my dealer found a linear power supply by Plixir, really made a difference. I am sure MCRU will specify a PS for the EE switch at some point which might be worth trying.

340

Thanks Simon for such a generous and comprehensive reply. I use a Cisco myself and wondered why it gave me an improvement in sound quality, most notably in presence. The management facilities of the switch are beyond my ability and understanding: for me it is plug and play. As far as I am concerned, you are a valued authority on Cisco.

Someone elsewhere wrote: "Has anyone worked out how a fancier switch can make any improvement in sound quality yet, especially when data is packetised and trasmitted using TCP (rather than a streamed serial interface like SPDIF), requiring extensive reclocking and multi-layer decoupled FIFO buffering at both ends, and all Cat5 Ethernet being differential, balanced, echo cancelled, and galvanically isolated?

I would have thought if there was a difference it would be because of inconsistency of processing time of packets in the switch and the receiving end being tightly integrated in the software sense (i.e. a tiny operating system with application statically linked to it) such that playback clock was derived from audio sample packet data arriving: a complicated managed switch with many features would sound worse due to the inconstancy in packet processing time compared to a wire-speed dumb switch which will have constant latency."

I do not have the technical knowledge to comment on this but would welcome your thoughts. It certainly seems to fly in the face of many user’s experience here.

2 Likes
342

That was a very thoughtful and helpful write-up, particularly since you provided such a full account of the context, both geographical and over time.

2 Likes
343

How about using practice and theory in harness together…

That approach works better than eitter one standing alone.

5 Likes
344

:small_blue_diamond:NO,.Not In This Case,let me explain…

If you shall understand this,.then you should be with someone who are knowledgeable,and can performs this.
This so that you can see and understand different contexts,.in the process.

Of course Xanthe,.it is very good,to also have read the written instructions available about this method.
This,.before you start to practice this practically with someone knowledgeable.

These instructions can be found on Lejonklou Hifi’s forum.
Previously there were also instructions on Linn’s now defunct forum.
If the instructions are on the new WAM (Linn) forum I do not know.
But if I’m to be honest,.when I’ve read these instructions,then I don’t understand how to do,.and I am very knowledgeable in this method.

• Fredrik Lejonklou (Lejonklou HiFi),.one of Sweden’s three most knowledgeable on this method has said.
That he has throughout his career,.only met two people who have managed to learn the method,.just by reading the instructions and practicing himself.

In my case,.I have not encountered anyone who understood to perform the method to 100% without help.

So the best is,.like I said in the beginning…

“If you shall understand this,.then you should be with someone who are knowledgeable,and can performs this.
This so that you can see and understand different contexts,.in the process”.

Stay safe out there…
/Peder🙂

345

Sounds like a stiff challenge!

1 Like
346

Although you haven’t actually stated it (and I find the above post quite difficult to understand), I assume you are referring to the ‘Tune Dem’ approach to assessing HiFi systems (N.B. you can’t asses individual components using Tune Dem as it requires an entire system to be functioning for Tune Dem to work, and there is no controls on inter-component interactions).

If you looked at a number of my posts (particularly those on the old forum) you will observe that I use the techniques of Tune Dem along with other techniques in assessing systems.

I do not rely on just one system or approach since, as with other incompletely controlled methods of assessment, doing so increases the chance of erroneous results due to the failure to use full experimental controls. The more different ways you inspect something, the more chance you have of finding important differences. I prefer to keep many tools in my arsenal rather than always relying on just one.

There is another adage that is also appropriate:
“When the only tool you have is a hammer, all problems begin to look like nails.”

4 Likes
347

But if they are effectively only nails, the hammer is all you need.
:rofl:

348

Thank you James for a very thorough report from your experiences and findings so far with the EE switch. It is very very much appreciated. I will have to read through your post many times to get a hold on things. It seems you’re quite happy with your decision in keeping the EE (new kid on the block?). I am about to order one myself with some Shawline Ethernet cables and after a trial period will give the forum my verdict too. .

Again, thank you so much for your time here to have posted your verdict. :sunglasses:

1 Like
349

I also highly respect @james_n’s point of view, so also the EE is a Chord product, so it sounds like a respectable device to be considered.

3 Likes
350

I’m looking for something to really beat my wireless connection. So far the Cisco or Ciscos have not. Hopefully a single EE switch will. :sunglasses:

351

By the way, I just wanted to say that when I used the term ‘bandwagon’ above, that was in response to being asked why I didn’t buy and try one of these switches.

But I really do appreciate the test reports and ideas of those forum members who try new stuff. It’s one of the most interesting and useful things on the forum - as well as sharing ideas for new music.

3 Likes
352

There are several aspects… there is interframe timing consistency but that tends to governed by the router or media server on your LAN subnet. This timing is many multiples of the symbol clock frequency as used on the serial lines over Ethernet segments such as from switch to streamer… so really the switch PHY clock has no real impact on this.

The transport dynamics of the TCP protocol can affect the processing on the end streamer. This is governed by the server bandwidth and latency. This comes more into play usually with internet accesses. Again the switch PHY clock has no bearing on this.

However the symbol encoding frequency on the physical Ethernet line… these are the minus, zero and positive analogue voltages on the serial line(s) are synchronised by the switch PHY clock. These modulated analogue voltages will couple to connected devices… and may well modulate then ground plane of the receiving streamer and radiate if there are aspects of the connection that are not purely balanced. This ‘carrier’ modulation acts like a micro FM transmitter, and if there is phase distortion in the original clock, then this will produce noise sidebands in the coupling … and likely cause perturbations in the Receiver (streamer) ground plane through electrical and magnetic coupling. This in turn can modulate the streamers clock and associated electronics by minute amounts… this is what some of us can notice.