Master Clock Talk

[MartinWT]

Not quite, the R usually depicts the decimal point, e.g. 8R2 means 8.2 ohms.

 

[Zaurux]

Some use a low pass filter on a sine output (e.g. OCK-1 with R26/A26) ?

 

[gimmeheadroom]

Some use a low pass filter on a sine output (e.g. OCK-1 with R26/A26) ?

 

[Zaurux]

My humour is not British enough to understand

 

[camrector]

Hey guys quick question for ya.
I asked this elsewhere and wanted to bring it here too.

Noticed something when shopping between the LHY sw10 and the awaited EtherRegen2. The sw10 has really caught my eye because it could potentially eliminate/upgrade 3-4 current boxes.
1 eliminate first FMC and linear ps.
2 eliminate/upgrade master clock
3 upgrade current Netgear switch

Cool stuff

My current switch is connected on the “dirty” side of my network. All network stuff and cpu connects to it along with the first fiber module that runs to my second fiber module and then to my R26.

My question concerns the SW10 and it’s master clock implementation as well as its placement in the network chain.

1. If the sw10 replaces my switch on the “dirty” side of my network, why would I connect its master clock to my R26 on the “clean” side? Bridging a BNC over my fiber convertors would negate my FMCs, and especially with how prone clocks are, add noise and jitter to my R26.
2. If the switch is connected to the “clean” side of my network, by becoming the last FMC, why would I connect all of my noisy “dirty” network components to its Ethernet ports? I would also then be connecting the R26 to that same switch, again negating the FMCs and most likely adding noise to the DAC.

If the answer is a second switch, why the heck does the sw10 have 17384917 Ethernet ports?!
This doesn’t make any sense.


I ask this as a question only to see what other opinions or knowledge is out there.
My mind is pretty clear in what is going on.
The EtherRegen has a similar issue but let’s start here first.

 

[mgsu]

Hey guys quick question for ya.
I asked this elsewhere and wanted to bring it here too.

Noticed something when shopping between the LHY sw10 and the awaited EtherRegen2. The sw10 has really caught my eye because it could potentially eliminate/upgrade 3-4 current boxes.
1 eliminate first FMC and linear ps.
2 eliminate/upgrade master clock
3 upgrade current Netgear switch

Cool stuff

My current switch is connected on the “dirty” side of my network. All network stuff and cpu connects to it along with the first fiber module that runs to my second fiber module and then to my R26.

My question concerns the SW10 and it’s master clock implementation as well as its placement in the network chain.

1. If the sw10 replaces my switch on the “dirty” side of my network, why would I connect its master clock to my R26 on the “clean” side? Bridging a BNC over my fiber convertors would negate my FMCs, and especially with how prone clocks are, add noise and jitter to my R26.
2. If the switch is connected to the “clean” side of my network, by becoming the last FMC, why would I connect all of my noisy “dirty” network components to its Ethernet ports? I would also then be connecting the R26 to that same switch, again negating the FMCs and most likely adding noise to the DAC.

If the answer is a second switch, why the heck does the sw10 have 17384917 Ethernet ports?!
This doesn’t make any sense.


I ask this as a question only to see what other opinions or knowledge is out there.
My mind is pretty clear in what is going on.
The EtherRegen has a similar issue but let’s start here first.

I would dont mind the noise going into SW10 clock since it has its own PSU and i would assume it is isolated from the LAN ports.

Regarding the dirty LAN side i would not bother because all LAN ports should be isolated if the SW10 is a true audiophile switch.
Plug all your RJ45 cables to the SW10 switch and hear if it degrades the music. If so, then you could buy a SFP switch for your dirty side and attach it over fiber to SW10 which handles the clean side.

 

[MartinWT]

You cannot have too many stages of isolation.

I have 5: air-gapped (1), transformers (1), fibre optic (1), moats (2).

 

[camrector]

You cannot have too many stages of isolation.

I have 5: air-gapped (1), transformers (1), fibre optic (1), moats (2).

Agree.
But the SW10 ruins all that if you use all its features.

I guess it gives people options
Either you use it as a fiber and a clock on the clean side. Or a switch and a fiber on the dirty side.

I would not recommend anyone use it as a Master Clock, Switch, and Fiber.

 

[OCC7N]

I would not wait for any other switch if I had the money for SW10.

Also I dont think people realise how good and versatile the OCK-2 is for the money. Its a pure steal, and yes it sounds better when used with ddc and dac(AGD) not 1 but 2 devices in sync.

The clock in SW10 being bigger makes sense since it has two different domain masterclock/out and switch/in.

Im only buying the etherregen because I need clock input, and cant afford the SW10, too expensive, but its not expensive for what it is. I dont think Uptone has thought of masterclock inside, but only input.

 

[camrector]

I would dont mind the noise going into SW10 clock since it has its own PSU and i would assume it is isolated from the LAN ports.

This defeats the purpose of an external master clock.
Clocks are soo sensitive to noise and jitter. That’s one of the main reasons why we use externals. To get them away from potential noise from other components.

Regarding the dirty LAN side i would not bother because all LAN ports should be isolated if the SW10 is a true audiophile switch.
Plug all your RJ45 cables to the SW10 switch and hear if it degrades the music. If so, then you could buy a SFP switch for your dirty side and attach it over fiber to SW10 which handles the clean side.

Don’t disagree here. Just don’t connect the Master clock to your R26.

 

[camrector]

I would not wait for any other switch if I had the money for SW10.

Also I dont think people realise how good and versatile the OCK-2 is for the money. Its a pure steal, and yes it sounds better when used with ddc and dac(AGD) not 1 but 2 devices in sync.

The clock in SW10 being bigger makes sense since it has two different domain masterclock/out and switch/in.

Im only buying the etherregen because I need clock input, and cant afford the SW10, too expensive, but its not expensive for what it is. I dont think Uptone has thought of masterclock inside, but only input.

I think you are missing my point entirely here.
The sw10 touts itself as a Master Clock and a switch with a fiber input. It has one clock input and one clock output.

What I’m saying is that you cannot use it as a Master Clock, switch, and fiber convertor, without introducing noise and jitter into your DAC.

 

[Jake2]

I would dont mind the noise going into SW10 clock since it has its own PSU and i would assume it is isolated from the LAN ports.

Regarding the dirty LAN side i would not bother because all LAN ports should be isolated if the SW10 is a true audiophile switch.
Plug all your RJ45 cables to the SW10 switch and hear if it degrades the music. If so, then you could buy a SFP switch for your dirty side and attach it over fiber to SW10 which handles the clean side.

I would put the question on port isolation of the SW-10 to Alvin/Beatechnik. I have the LHY SW-8 which although a great sounding switch feeding my R26's LAN input is sensitive to sources of noise on other inputs. Also connecting my TV via ethernet to the LHY adds a layer of haze and flattens dynamics when listening to music from my server.

Though the counterintuitive catch-22 is each audio source, notwithstanding they are each also a source of ethernet borne noise, sounds best itself when connected to the SW-8 rather than further upstream, more remote to my Netgear GS108E switch (Alpha Audio recommended) fed by an Ifi Power X . Ie. AV sources sound much better when the TV is connected to the SW-8, ditto with music where my server is directly connected to it. Recsntly adding ifi LAN iSilencers in between server and upstream Netgear switch connections to the LHY has effectively added the high quality port isolation it was missing, including improving the sound from all sources.

Ideal solution IME... two high quality switches the best one solely for DAC/streamer and the upstream one for server /AV sources. With full galvanic isolation between.

 

[OCC7N]

This is how you use the SW10.

First of all placement as close to you source/ethernet/DAC as possible.

Only one output option(1) should be used to endpoints/clients

Input: SFP

Output Option 1: Ethernet/Switch - stream dac
-the most obvious one

Output Option 2: SFP
-this for daisy chaining one more switch. Use this if you plan to buy one more

Only 1 Clock I/O should be used at a time, if not daisy chaining with one or more SW-10

OUT: Masterclock out to DAC/Streamer

IN: The clock input if you have a masterclock like OCK-2 or another

Its a versatile switch.

 

[camrector]

I would put the question on port isolation of the SW-10 to Alvin/Beatechnik. I have the LHY SW-8 which although a great sounding switch feeding my R26's LAN input is sensitive to sources of noise on other inputs. Also connecting my TV via ethernet to the LHY adds a layer of haze and flattens dynamics when listening to music from my server.

Though the counterintuitive catch-22 is each audio source, notwithstanding they are each also a source of ethernet borne noise, sounds best itself when connected to the SW-8 rather than further upstream, more remote to my Netgear GS108E switch (Alpha Audio recommended) fed by an Ifi Power X . Ie. AV sources sound much better when the TV is connected to the SW-8, ditto with music where my server is directly connected to it. Recsntly adding ifi LAN iSilencers in between server and upstream Netgear switch connections to the LHY has effectively added the high quality port isolation it was missing, including improving the sound from all sources.

Ideal solution IME... two high quality switches the best one solely for DAC/streamer and the upstream one for server /AV sources. With full galvanic isolation between.

Can you explain your logic/reasoning for having a switch after your second fiber convertor instead of direct into the R26?
That would add another point of potential noise especially if using the R26 LAN.

 

[Jake2]

I think you are missing my point entirely here.
The sw10 touts itself as a Master Clock and a switch with a fiber input. It has one clock input and one clock output.

What I’m saying is that you cannot use it as a Master Clock, switch, and fiber convertor, without introducing noise and jitter into your DAC.

Hi Cam I agree it is theoretically not optimal, and totally get your reservations. But as with many things it very much depends on implementation. It may be that despite the single box there is superb - or at least sufficient - isolation between the clock and switch/FMC sections* (which may also prove to each have been implemented to a high standard) that they each work extremely well, delivering the typical LHY high bang for buck performance.

It is not cheap though, so with next to no reviews of the SW10 yet one would be taking quite a chance, and with the EtherRegen v2 being conceptually purer and likely technically more advanced I'm personally leaning towards it. I remain SW-10 curious though...

*The clock is on a separate board but not in its own separate chassis shielded section I see.