I suspect they do use local step down buck regulators…this new switch…does away with those which should in theory be a good thing…as they are horrid little things…which are far too handy!
its not the step down I am thinking of with standard regulators - its the polarity reversal that likely requires the current pumps. Remember Ethernet is balanced with +/- voltages around 0v, therefore if you are feeding +12 Volts DC then you need to transform the voltages.
I suspect in many low cost units these functions are built into the controller chip.
Hello I am very interested in your results. I have newly bought a used Silent Angel Bonn N8 and tried it out today. I also use a Cisco 3560 with factory settings (I think with IGMP snooping). The Silent Angel is slightly better in resolution than the Cisco alone. Better still is the combination: Router - Cisco (with Tv Box) then - Silent Angel with Streamer. On its own, the Silent Angel sounds a bit thin in the basic tone and bass.
I think I’ll keep the Cisco and use it in the garage as the first switch for the entire house wiring (including the solar system, heating and wiring throughout the house). The Silent Angel will then remain in the living room for the streamer, the TV box and the WLan access point). I might try a DC supply for the Silent Angel (SBooster or IfI IPower Elite with 5 volts). I’m curious to see how it works in the end.
Back to the LHY Switch…. I have seen the new Silent Angel Bonn NX Switch on the web who seems to have a special technical architecture too like the LHY. But the NX will cost about 3‘500 Euro, with external powersupply and clock more then 6‘000 Euros…
Ok I have just connected up the LHY…which I assume will burn in somewhat…I have given it a quick cold listen…results…very positive. Good detail and bass performance - sound stage is very nice. One thing that I immedetly noticed was piano - this sounded really good. So this was a quick 10 minute test…just to make sure it works. I will over the next week or so have some more serious listening sessions with familiar tracks… Now I currently use the Aqvox SE with a high performance Belleson LPS…this is my benchmark - so far the LHY has not been embarrassed.
As a final test I shall install the LHY on the end of the Aqvox - to see if that gives me anything…
It’s not really a special architecture at all from that info… it’s just following good design principles at reducing out of band noise (FM phase noise on the serial lines) and reducing the propensity of common mode noise coupling to attached devices.
I suspect the architecture is the same as a little consumer £10 network switch.
In essence it’s focussing on the out of band issues rather than addressing network switch related noise issues… it’s a shame, as it’s a missed opportunity. However such a device that addressed network noise would probably cost significantly more… such as you see in managed switches.
Is there any point having all these fancy clocks on these switches? as once the signal gets to the dac, the dac has a buffer and will re clock it anyway.
I agree and thats what I thought…but if you experiment the truth is power supply is critical…somehow something is effecting the dac … and its audible…very audible. Simon is right … the LHY switch is a very basic commercial switch, which has no management capability. I will continue testing - I am especially interested to see what happens when I add the LHY on the end of the Aqvox… I notice that Paul Pang has daisy chained 4 switches…in a row with the Quad 4…which seems a bit crackers…
This switch thing is confusing. It really makes a difference but which is the optimal option? Many new products and most of them are expensive and difficult to try. I moved my uptone LPS 1.2 from powering a cheap telecom router to powering a cheap switch ( Cisco SF110D) and it was clearly better option.
Now i have Router->Cisco SF110D-> eR (stock ps)->NDX-2 This is the best combination i have achieved with the networking devices i have and makes me wonder that there is a potential with better switch and ps. I have to try wifi (i have not yet) before any upgrades, maybe save me a lot of money and time.
Was using a cheap ZeroZero LPS before, as I had this from earlier usage of a Netgear GS105 (now repurposed upstream) and then a Cisco SF110D-05, however, the Farad3 brings noticeable improvement what the EtherREGEN is doing.
Considering an external Clock generator for the EtherREGEN over implementing an optical fibre based feed into the EtherREGEN.
I use these…
• Ediscreation
Fiber box II
• Ediscreation
Silent Switch OCXO
It is a Fiber solution,.which together with the Switch creates a synergy effect that must be experienced.
This to understand what improved soundquality these two units contribute with.
Also note,.that you can break the ground at every input on the Switch.
My Cisco 3750G has the ports to accept fibre…I might in the future take a run of fibre up to my hifi room and hopefully that would help reduce the nasties getting into the hi-fi side of things…
@Dunc that is a good question… and a totally understandable question given how these devices tend to marketed to audiophiles who for the most don’t really know how these things work or understand what is really going on… you can argue why should they?
The clock stability has NOTHING to do with any sample clocking… absolutely not.
What the stability is about reducing phase noise coupling in the serial lines or fibre which couples into the receiver which depending on the receiver’s design can cause an increased noise level in the receivers (ie streamer) electronics which may affect how the resultant audio is perceived.
Think of phase noise power as a tiny narrow band FM transmitter sending white noise energy into the leads of your streamer….
I guess some find some of this confusing… but at the end of the day different types of very high frequency noise can effect resultant audio band signals… and so a lack of noise or too much of a given frequency of noise can detract from the audio replay experience. Less is not necessarily better… from an enjoyment perspective. For example in many contemporary production techniques noise is deliberately carefully added to the signal to make it sound better and more ‘analogue’.
Also remember Naim themselves ‘noise shape’ with their firmware to allow firmware to sound preferable or more enjoyable…
So yes it can be confusing…and is not straight forward … hence why searching for zero noise is really a fools errand.
In your experience, based on what you have said, are there any switches that you would recommend, or at least aim to adequately address the issues you are referring to?
Well I go back to when I possibly started this Hi-Fi/audiophile switch fad on this very forum all those years ago.
A used Cisco 2960 or 3560 work well, and in my experience provide good network benefits at relatively low cost as well as reducing OTT common mode or phase noise like some other more consumer oriented devices do/did.
However for noise ‘tuning’ then the larder is full of exotic as well as simple/cheap variants.
My view is that it’s probably better to go for a robust quality commercial switch like the 2960 or 3560 and then ‘tune’ with different Ethernet cables if you have to… (remember price means nothing here)… but in my personal experience I largely decouple my transport streamer from my DAC to mitigate variances from networks/Ethernet leads etc.
I fully respect what Simon is saying but I found that power supply seems critical…I don’t know about clocks … this is from me just trying… the little LHY seems pretty darn good…I will report back later…