Optical Network Configurations

[SQFIRST]

On 4/3/2024 at 2:08 PM, jabbr said:

This isn't just clock jitter rather the "eye pattern" which doesn't pass upstream noise by specification.

 

On 4/3/2024 at 2:08 PM, jabbr said:

My last segment is 10g

 

@jabbr thank you for your thoughts on that. It makes sense. Combined with the post from @barrows I should clarify how I am using 10G and my experience of the subsequent benefits.

My last audio segment to the Lumin X1 is 1G optical as that provides the best network isolation as the Lumin has an SFP port. My audio does NOT use USB so network connection is the final digital connection. There is no further digital optimization possible such as with a DDC or USB audio device. All further enhancements are only at the DAC level and otherwise analog in my setup. You can understand the emphasis on network quality in my case and also note the unique nature of streamer DAC setup.

The benefits I derive from the 10G protocol is from using a Mikrotik CRS 305-1G-4S for just the optical 10G connections, to centralize switching at 10G speed, thereby gaining the benefits of 'fill in this space with your choice of technical understanding'. Router, Servers and audio segment all connect at 10G. There is no use of 1G in this switch and no dropping of rate using a dual rate SFP+. Everything autonegotiates to 10G for the core switching. On the opposite ends of each connection I have 10G switches that connect components with 1G copper.

If I were to remove the 10G connections, thereby connecting all devices at 1G instead, the sound becomes less enjoyable immediately. I am including uPnp as well as Roonserver and NAS media. @jabbr has the unique limitation of not using 1G connections in a manner where such tests can be conducted otherwise this would be more apparent to them.

@barrows I do appreciate you sharing the benefits of a higher quality 1G device, as well as your thoughts on possible iitter impact,  and while I have not used the OMD I have researched enough to understand its benefits. I use the EtherRegen with an external clock going from it's B-side copper to A-side optical to the Lumin. 

 

 

[SQFIRST]

Thanks for trying to help with suggestions everyone but my post was less about experiencing issues and more about how I was able to address them with the use of 10G. This is just one of the many solutions to explore and what we were discussing recently. The reason that I elaborated on my connections was to respond to the inquiry from @barrows about how I am using 10G as well as to appreciate @jabbr's reliable and consistent assistance with technical answers on the subject.

Each setup with have it's unique limitations but the basics have to be covered first, which includes power domain separation and isolation. My Lumin X1 is power isolated and separated from everything so optical is the best connection as any use of copper will break the power related implementation.

 

@dbastin congrats on finally getting the B-side benefits and it's good to hear it is working for you.

 

[SQFIRST]

There is networking and then there is the audio version of it :)

I just moved away from Corning SMF-28 to the bend-insensitive type, the opposite of @audiom3's experience. Just pointing out the variety of experience and not anything to contradict. I have some long fiber runs so that factors into it. For short 1M run if you manage to kink it up then it's on you! Jokes aside I did read up on the Corning site and there is such a thing as 'fiber fatigue'. The SMF-28's thinner core may well be playing into the perceived sound quality differences but also leading to the possible physical degradation or bend issues. That is the experience which I had reported just few posts back and @jabbr pointed out the Clearcurve.

 

Just another reminder that what is 'better' is a very complex answer and best not to figure that out without getting details and trying out for oneself.

 

 

 

 

 

[SQFIRST]

31 minutes ago, audiom3 said:

 Corning SMF-28 is bend-insensitive

Somehow I managed to get them wore out! Maybe it was just time for a new set in my case. Regardless, they do break down or can still cause bend issues as I experienced.

30 minutes ago, hltf said:

TP-Link MC220CM FMC units.  These TP-Links will be replaced with the two 10GTeks

What are you trying to achieve? Just asking so that folks responding to your questions can help you better. I do not see this change as being a big one and I say that because you mention using the Uptone JS-1 already, which is a high quality PS. Something like the OMD or ER, at least on the receiving end could make a bigger difference. It all depends on what you are connecting and trying to achieve.

[SQFIRST]

6 minutes ago, hltf said:

I thought I understood from the discussion earlier on in this thread that having the Finisair FTLF1318P3BTLs would give me substantially better optical isolation than the TP-Links that I am currently using. 

 

1 hour ago, hltf said:

For the moment I am using this same Uptone on my two connected TP-Link MC220CM FMC units

 

@hltf wish you success with your journey! 

Nothing right or wrong about it but just asked to better understand what you expect to achieve. The TP link and the 10gtek are close in quality as to the electronics and quality. I see now that you are planning to move from an SC type cable connection to the LC type. This will result in changes related to the SFP module, yes but the overall quality of your optical setup will likely remain quite similar. This is where you are seeing folks mention the Optical module deluxe/EtherRegen/LHY switch etc. Hope this helps.

 

[SQFIRST]

@hltf The 1318s are a praised for a reason and you may well find enough improvement to be happy. Just the fact that you are using a good power supply makes me think that this change will not be enough though, just a feeling :)

[SQFIRST]

Something often overlooked when setting up an optical network isolation is the power domain separation to support it. Sharing power supplies or circuit with components on both sides of the optical connection defeats the isolation to some degree, sometimes critically. An effective optical isolation setup takes into consideration that all components on the 'sending' end are on a separate power domain. This way the 'receiving' end of the optical connection can be within a power isolated domain and the only side where the best power supplies are targeted. This requires careful setup and some extra effort in separating components but it really is worth the time. 

 

Also something I picked up from this helpful thread and on the network isolation thread so sharing back as a reminder.

[SQFIRST]

3 hours ago, hltf said:

Do you feel it could/should be better if I were to use an IFI SMPS for the first 10GTek

@hltf as you mentioned that you hear better quality when using the JS1 for both FMCs then there is no reason to change that. What I have observed is that the optical isolation setup lines up with the required power isolation/separation. Just look at what is connecting to each side (beyond just the FMCs) and you will get a good idea of how to rearrange. That is, if you are curios to get to optimal. Nothing wrong with sitting back and just enjoying the music as is.

[SQFIRST]

That's exciting to see the tech trickle down for home use.

(Home entertainment manufacturers really needs to move up from 1G for the jitter blocking benefit)

[SQFIRST]

1 hour ago, jabbr said:

I have no idea of a 1G SFP device will handle a 25G AOC cable

 

Very unlikely that 1G cage will accept a 25G module. The form and dimensions of quad speed modules appear to be slightly bigger and different than sfp/+.

 

 

[SQFIRST]

Correction - QSFP modules will not work in 1G cage. Looks like SPF25 may physically fit into 1G. Interesting premise @Crom, do keep us updated with your journey.

[SQFIRST]

I was thinking of the size of the QSFP modules which are bigger and spoke too early on the SFP28 which IS same size as SFP. Thanks for the follow up @jabbr.

Here is a chart that I found helpful:

 

[SQFIRST]

17 minutes ago, jabbr said:

This *is* a good time to make the point that the 100Gbe devices which require vanishingly low jitter in order to work, DO NOT use special power supplies nor clock crystals, rather the clocking electronics have become rather sophisticated

This is good to know. What changes did you experience going to 100GBE devices with regards to power domain separation/control? I imagine your audio is still protected behind isolated power and clean power at the endpoint?

[SQFIRST]

6 hours ago, jabbr said:

in this hobby when you hear something more pleasing to your ear you don't need to question it

Love this statement. Easy to get caught up in the pursuit and the details and overlook the simple enjoyment of it all.

 

2 minutes ago, audiom3 said:

I am wondering how much better 100G is vs 10G and if upgrading a switch could do exactly what the OCXO in the LHY is doing

There is a kind of enjoyment and satisfaction from putting something together and seeing it all work. Similar to the tactile feel of a vinyl rig. Technical comparisons may not cut it or even make sense sometimes but nothing wrong in enjoying it. 

[SQFIRST]

Adding some thoughts from my experience:

 

On 5/3/2024 at 10:55 AM, dbastin said:

Mikrotik CRS 305 will give you SFP+ ports to connect up and downstream in the direct route to the hifi

+1 on the 305.

SWOS 'sounds' better than ROUTEROS. There are reports of the ROUTEROS in 305 having throttled throughput. It is audible if you A/B the switch.

 

On 5/3/2024 at 5:56 PM, jabbr said:

I have used RJ45-SFP+ modules which work well for networking but I HAVE NOT used these with audio endpoints

If not used directly in the audio path, they do work quite ok but I did find that they degrade sound if used on the audio path.

 

On 5/4/2024 at 10:14 AM, audiobomber said:

I have a friend with an ER/MinTeddy and SW-10. He says his system sounds best with the SW-10 after the router, with the ER closest to the Antipodes K50 and Tambaqui, the more critical position.

I have similar finding with my setup. With the ER having a power domain separation internally and the clocking taking place while crossing the sides, I have not been able to get similar clarity from any other 1G device. There is something very unique about the ER's use of a stable clock signal that is easy to overlook at first as it is somewhat tricky/difficult to setup in the correct manner. It is also very setup specific and not every use case will see benefit. If proper power domain isolation is managed combined with a stable external clock, the output from 'crossing the moat' is a very transparent audio ethernet signal. I use it as a final optical 1G connection to a streamer-dac (Lumin X1).

 

On 5/4/2024 at 8:05 PM, jabbr said:

2) the topic of clocks for network switches is more complicated that simply the "OXCO" lingo that is thrown around

I realize the optical networking thread may not be the right place for discussing ethernet clocking. Since our discussions do stray into network 'sound' a lot of times, there is some need to acknowledge that something does happen in such instances., which makes an audible difference that is pleasing. There is no argument with the fact that all ethernet switch uses clocking and that optical can eliminate common mode noise. There is certainly some misinformation around this perhaps from lack of understanding (both in reports as well as those refuting) , so it may seem easy to write it off as wrong information. However, there are also some carefully reported findings with well described instances where it does merit some room to for the possibility that at least some more definition is needed.
I agree that the focus needs be less about the ocxo and more about what is causing the better sounding ethernet signal in these instances.

 

[SQFIRST]

@audiophilac I do not use the 309 and do not know of any similar impact. I do know that SWOS is lighter compared to RouterOS as it is focused purely on switching.

[SQFIRST]

3 hours ago, jabbr said:

1Gbe is frankly ancient at this point and 10Gbe is a previous generation. Modern Ethernet electronics has come a long way and an external (OXCO) clock has not been shown to improve the Ethernet signal. 

This is very true. But it is a viewpoint contained to technology alone and not a system. Please consider that audio endpoints and home entertainment devices are only 1G capable. There will be a dropdown to 1G when connecting to them, regardless of the cleanliness from higher protocol connections upstream. So the challenge is really the last connection which is still 1G. The approach and solution for this varies depending what each person has tried. 

 

20 minutes ago, audiobomber said:

Despite all the chatter in this thread about the technical superiority of 10G vs. 1G, there are no comparisons that I've read directly comparing a 10G implementation to an audiophile switch.

10G discussions in this thread have been about network connections and not specifically to audio endpoints. If one has tried 10G in their network, as myself and multiple others have reported, there IS overall gain in reducing network jitter. How this propagates to the audio endpoint is dependent on the individual setup and implementation. 10G and newer protocols are superior at reducing jitter - this is industry proven. But that does not eliminate the possibility of issues when connecting with 1G to audio. Basically, your proof would be possible if the audio endpoint is also 10G capable.

[SQFIRST]

6 hours ago, McNulty said:

If Sonore, Lumin, Hifi Rose etc. are reading this,  maybe they can consider releasing endpoints with SFP+ ports instead of SFP as a next generation of their existing products…

That seems like a sensible approach doesn't it? It may pose some engineering complications with regards to sensitive audio design. If that is not the case then it is a no brainer. Just from an adoption standpoint, there will be resistance as the norm has been 1G and lower up until now. Add pricing increase to that, the case becomes quite difficult. But certainly a ground breaking design waiting to happen.
 

2 hours ago, Superdad said:

the Analog Devices synths we are moving to for EtherREGEN Gen2 have dramatically lower phase noise

This is great to hear as the benefit from the ER1 is audible and seems attributable to the phase noise lowering aspect.

 

2 hours ago, Superdad said:

this leaves out the noise/jitter that gets embedded by optical transceiver modules (differences in which I believe is the primary cause for the variations folks here between SFP brands/models; and why Finisars generally end up as preferred).

Which brings us back to the earlier discussion about the advantage of using 10G and higher protocols to remove jitter.

The point being that a concerted effort targeting digital 'noise' removal, which could include multiple specialized solutions, maybe needed for quality output. One approach does not have to negate the other. Also, this type of effort may not be for everyone as people have differing levels of patience, idea of fun and interests. 

 

1 hour ago, audiom3 said:

I don't know why the 10G modules refuse to lock into the Lumin U2

I have the X1 and did notice the same issue and narrowed it to the transceiver locking mechanism. The Lumin did not allow for the lock to open properly which prevents the module from seating into the cage. I did get a generic 10G to work with it but did not like the performance. It is a good idea to not try to make it work and heed Lumin's recommendation.

[SQFIRST]

1 hour ago, jabbr said:

Its entirely unclear to me that the endpoints need to handle 10G ethernet given the assumption that the switch device which is capable of 10G would presumably perform in a similar fashion when operating at 1G

 

1 hour ago, jabbr said:

Again all compliant 10G+ switches remove jitter and all fiberoptic switches remove common mode noise. You don't need to worry about this. 

 

 

Based on your statements it would seem that the Mikrotik CRS305 is not compliant.

I have been using it to reduce network jitter with 10G end to end. There were other solutions necessary for optimal sound from my 1G endpoint which is a streamer-DAC. 

 

1 hour ago, jabbr said:

Its not the 10G SFP which removes jitter though any device attached to the DAC might affect SQ.

The conversation here was more about the physical issues in seating SFP+ modules in a Lumin and not so much about 10G itself.

 

 

[SQFIRST]

12 hours ago, jabbr said:

Its entirely unclear to me that the endpoints need to handle 10G ethernet given the assumption that the switch device which is capable of 10G would presumably perform in a similar fashion when operating at 1G

 

@jabbr I really appreciate your efforts to right track information. I am not questioning ethernet standards and have no confusion with that. Just need some clarity on the theory in practice.

 

My actual usage experience with the Mikrotik CRS305, which is central in my network, is not following the assumption that you have mentioned. There is a clear difference when the switch is connecting at 10G end to end vs. when connecting at 1G. This difference is only based on my hearing the output from audio setup and not by any other measurement. To be clear, the CRS305 does not appear to be performing in the similar fashion at 1G as with 10G. In my experience so far, unless the connection is running with both of the ends operating at 10G, the benefit is not present in the audio in my chain. 
 

[SQFIRST]

27 minutes ago, jabbr said:

If you are using a fiberoptic Lumin at 1G how does the switch connect at 10G? 

 

Really its the last connection to endpoint which I feel is most important

This is exactly what it comes down to and it would be great if the CRS305 handled 1G as it does 10G but that is not the case. The option to connect 1G optical from the switch to the Lumin is good but in my case, I get better performance using the EtherRegen as the final 1G device.

 

My endpoint setup is simple as there is only network input terminating at the DAC (X1). It is a different experience from using USB to DAC.

[SQFIRST]

I suspect your mellanox or the 100g devices you have in place offers a high level of compliance where you do not see the similar issues that us mere mortals have been reporting. If you really want to test things you may end up working backwards a bit :)

[SQFIRST]

38 minutes ago, taipan254 said:

Is Impression 1 right or wrong? Is Impression 2 right or wrong? 

@taipan254  Yes :)

 

I can share what I have tried and from that limited scope:

 

For 'Impression 1' I will expect you to be getting the 10G benefits (assuming it is an end to end 10G connection) and will add that moving the 10G connection closest to endpoint should provide maximum benefit, since any device in the path can add some of it's own jitter.

 

I was never in the 'Impression 2' camp but that is the earlier comment from @jabbr stating as should happen which I have recently responded to. In practice I have not seen 10G benefits from using a singular 10G device. That is, reap benefit of 10G standards from using just one 10G device. For my experiments it has been primarily the Mikrotik CRS305 as a switch. So what I can conclude is that either the theory as @jabbr states does not hold up (jabbr is usually thorough and accurate) or this particular model is not fully compliant.
There is no new impact with this information other than it reiterates why I needed to further treat the final 1G connection after the 10G, as per my preference. 

 

I suppose it depends on what you are using and how you can test to discern the benefit.

[SQFIRST]

7 hours ago, jabbr said:

To be clear it’s not a theory, the specification mandates compliance testing whereby noise is injected into the upstream signal and the noise needs to be eliminated by the receiver. 
 

true a model might not have been rigorously tested — none of the 1G Ethernet devices need have this testing done 

 

Indeed. The actual experience of the network with an end to end 10G in operation is unmistakable.

 

Takeaway - compliance to standards is the key and just running a 1G connection off a 10G device does not always mean one is getting the higher standard benefit. 

 

 

 

 

[SQFIRST]

24 minutes ago, Jud said:

Keeping the audio system inside the home off the same circuits devoted to in-house noisemakers like refrigerators, A/C systems, LED lights, even appliances with microcontrollers, is a far more effective means of reducing noise than generating your own power.

+1

 

With regards to networking blocking critical switches/components is key.