Converting ethernet to fiber

[MarkS]

Does anyone use these two devices to convert eithernet to fiber, then back to ethernet?  If so, what do you think?

 

https://www.amazon.com/gp/product/B0034CMZIG/ref=oh_aui_detailpage_o07_s00?ie=UTF8&psc=1

 

https://www.amazon.com/gp/product/B0000513ID/ref=oh_aui_detailpage_o07_s00?ie=UTF8&psc=1

 

I’m told these devices kill noise. Looking for additional comments/experience. 

 

[davide256]

I've used the GigE version with microRendu. Where they help is to break electrical connectivity between your network switch and attached player/renderer. This helps improves low level resolution in your audio playback. Someone else would have to explain why, all I cared about was the result. You still have to worry about the wallwart noise for the converter/wired Ethernet in front of the server/renderer, I used 5V from  an HDPlex  LPS

[crenca]

I wonder how much interaction (i.e. noise) the photons in fiber have with cosmic rays?

[phosphorein]

40 minutes ago, crenca said:

I wonder how much interaction (i.e. noise) the photons in fiber have with cosmic rays?

 

No different than electrons in Copper. The major source of noise I expect would be in the transmitter and receiver stages.

 

[d_elm]

5 hours ago, MarkS said:

Does anyone use these two devices to convert eithernet to fiber, then back to ethernet?  If so, what do you think?

 

https://www.amazon.com/gp/product/B0034CMZIG/ref=oh_aui_detailpage_o07_s00?ie=UTF8&psc=1

 

https://www.amazon.com/gp/product/B0000513ID/ref=oh_aui_detailpage_o07_s00?ie=UTF8&psc=1

 

I’m told these devices kill noise. Looking for additional comments/experience. 

 

I have used MC100CM and MC200CM, chose MC100CM for lower noise of the two, but now use a 'special' switch instead of fibre.  If you use FMCs the downstream FMC should be powered with an  Uptone LPS-1 or equivalent quality LPS. Fibre will drop the upstream noise but add the noise from the downstream FMC and jitter from the process.

 

I now use a Netgear FS105 v2 modified by SOtM for linear regulators, caps, and external clock from sMS-200ultra.  As an experiment I placed the MC100CMs upstream from the switch and the sound was less clear. A result of the fibre jitter.

 

Uptone should soon have a switch to consider.

[jabbr]

4 hours ago, crenca said:

I wonder how much interaction (i.e. noise) the photons in fiber have with cosmic rays?

Is that a trick question? Photons are not affected 

[jabbr]

3 hours ago, phosphorein said:

 

No different than electrons in Copper. The major source of noise I expect would be in the transmitter and receiver stages.

 

Photons in glass of course are much different than electrons in copper ... transmitter & receiver stages are known and their low noise profiles are well known — if you are really concerned consider the bandwidth capability of single mode fiber 

[jabbr]

1 hour ago, d_elm said:

A result of the fibre jitter.

All this is quite demonstrable with an eye pattern — take a look at an eye pattern from a good fiber and your concept of “fiber jitter” — can’t say that crap from bet doesn’t exist but show the eye pattern 

[zoom25]

I suspect at home with copper ethernet and naturally with fibre, external noise will not be a problem. I'd pay more attention to length and possible reflection issues.

[phosphorein]

10 hours ago, jabbr said:

Photons in glass of course are much different than electrons in copper ... transmitter & receiver stages are known and their low noise profiles are well known — if you are really concerned consider the bandwidth capability of single mode fiber 

Not different in terms of their interactions with cosmic rays. Of course in other properties they are different, for example photons can interact with ambient light external to the fiber.

[davide256]

Glass is not an electrical conductor, but its crystal structure allows it to bend the path of photons. Cosmic rays are incredibly energetic but because of that their wave front interaction is very small, pass through most substances unaffected beyond  random collision. The chances are far greater of an interaction inside your electronics because of greater size/mass density.

 

The key to remember here is that commercial routers and switches are made for internet data packet transmission, connecting computers for ordinary data processing. Things like leakage  current and a few milliseconds of jitter don't matter here, are expected and allowed for. Audio grade solutions would likely be better served by the type of gear used for stock market trading networks where errors and milliseconds equal lost revenue on trading.

[plissken]

On copper vs fiber: You are chasing a non-existent rabbit.

 

I've seen no measurement on the output of a DAC that shows empirically that you can see any voltage flux when that DAC/Streamer is sent, for example, a 1Khz signal.

 

I've seen no bias controlled evaluation that supports going fiber also.

 

I would say go Wireless.

[davide256]

1 minute ago, plissken said:

On copper vs fiber: You are chasing a non-existent rabbit.

 

I've seen no measurement on the output of a DAC that shows empirically that you can see any voltage flux when that DAC/Streamer is sent, for example, a 1Khz signal.

 

I've seen no bias controlled evaluation that supports going fiber also.

 

I would say go Wireless.

Its another valid way to achieve electrical isolation from your switch/ router... fine as long as  the signal path doesn't have issues. But products like the microRendu don't come  with built in wifi.

[plissken]

5 minutes ago, davide256 said:

Its another valid way to achieve electrical isolation from your switch/ router... fine as long as  the signal path doesn't have issues. But products like the microRendu don't come  with built in wifi.

 

Ethernet PHY's are all air gap'd on both ends by transformer coupled magnetics. Anyone can hit google.com/images and search. You will see where the tops of the packages are removed and see that it's all electrically isolated.

 

You can also get a WiFi bridge...

[jabbr]

45 minutes ago, davide256 said:

The key to remember here is that commercial routers and switches are made for internet data packet transmission, connecting computers for ordinary data processing. Things like leakage  current and a few milliseconds of jitter don't matter here, are expected and allowed for. Audio grade solutions would likely be better served by the type of gear used for stock market trading networks where errors and milliseconds equal lost revenue on trading.

Right except that jitter does matter which is why the eye-pattern is ubiquitous. 

 

Stock market trading networks look for latencies measured in the 10s nanosecond range. These are fiber NICs like the Solarflares that I have. 

 

The little tiny transformers in the RJ-45 PHYs do not prevent leakage current from passing —- high interesting capacitance — obviously no current passes through fiber 

[crenca]

12 hours ago, jabbr said:

Is that a trick question? Photons are not affected 

 

Yyyeeaaaaaa......

[jabbr]

37 minutes ago, plissken said:

You will see where the tops of the packages are removed and see that it's all electrically isolated.

 

You can also get a WiFi bridge...

“isolated” is a relative term because current passes across the transformer — likewise a capacitor doesn’t have a direct electrical connection yet current passes quite readily. Elthernet isolators were developed to reduce this eg hospital setting.

 

Wireless — yes good isolation but poor latency if that matters. I use three types of network in my house. 

[plissken]

4 minutes ago, jabbr said:

 

The little tiny transformers in the RJ-45 PHYs do not prevent leakage current from passing —- high interesting capacitance — obviously no current passes through fiber 

 

Still keen on seeing this as measured on the output of a DAC/Streamer.

 

1 minute ago, jabbr said:

“isolated” is a relative term because current passes across the transformer — likewise a capacitor doesn’t have a direct electrical connection yet current passes quite readily. Elthernet isolators were developed to reduce this eg hospital setting.

 

Correct and relatively inexpensive. I'm still keen on seeing this measured on the output of a DAC/Streamer.

1 minute ago, jabbr said:

 

Wireless — yes good isolation but poor latency if that matters. I use three types of network in my house. 

 

My Wifi is under 2ms throughout my house. It doesn't matter.

[pkane2001]

12 hours ago, jabbr said:

Is that a trick question? Photons are not affected 

 

Electrons are, and I have seen cosmic ray effects first hand in a long, deeply cooled CCD exposure. They are infrequent and extremely low energy. Fun to see them, as they often appear as a short meteor strikes in random directions in an image. Living at higher elevation increases frequency. 

 

[jabbr]

55 minutes ago, plissken said:

 

Still keen on seeing this as measured on the output of a DAC/Streamer.

 

 

Correct and relatively inexpensive. I'm still keen on seeing this measured on the output of a DAC/Streamer.

 

My Wifi is under 2ms throughout my house. It doesn't matter.

Look, fiberoptic, copper, Ethernet, Infiniband, Wireless all have their advantages, disadvantages and roles. 

 

They arent the same and what might not  matter to you in your system may matter to someone else. Just because a good measurement hasn’t yet been done hardly means a difference doesn’t exist. I’ve heard audible differences with different networks and different electronics YMMV

[plissken]

1 hour ago, jabbr said:

Look, fiberoptic, copper, Ethernet, Infiniband, Wireless all have their advantages, disadvantages and roles. 

 

They arent the same and what might not  matter to you in your system may matter to someone else. Just because a good measurement hasn’t yet been done hardly means a difference doesn’t exist. I’ve heard audible differences with different networks and different electronics YMMV

 

I agree they all have their roles in ways to deliver asynchronous data. I agree they all have their advantages and disadvantages delivering data.

 

So you are stating there are no good measurements? That's odd.

 

I'm still willing of offer 5:1 dollars on this unexplained, in-measurable, phenomenon of human hearing. I have a client/server/switch/wifi adapter all ready to go.

 

I even have fiber and copper PCIe cards available along with a Cisco switches with both RJ45 and SFP modules.

 

Let me know when you want to do this.

 

 

[jabbr]

9 minutes ago, plissken said:

So you are stating there are no good measurements? That's odd.

I said good measurements haven’t been done — at least I haven’t seen ones that correlate to subtle differences that I’ve heard with DACs — might involve , for example, phase error measurements within the DAC logic (not merely the clock)

 

What are you ready to do? These measurements?

[plissken]

50 minutes ago, jabbr said:

I said good measurements haven’t been done — at least I haven’t seen ones that correlate to subtle differences that I’ve heard with DACs — might involve , for example, phase error measurements within the DAC logic (not merely the clock)

 

What are you ready to do? These measurements?

 

I'm ready to perform some bias controlled testing.

[Ralf11]

 what about interactions with neutrinos?

 

I hope you all have shielded for that problem

[R1200CL]

On 21.4.2018 at 9:21 PM, MarkS said:

Does anyone use these two devices to convert eithernet to fiber, then back to ethernet?  If so, what do you think?

 

https://www.amazon.com/gp/product/B0034CMZIG/ref=oh_aui_detailpage_o07_s00?ie=UTF8&psc=1

 

https://www.amazon.com/gp/product/B0000513ID/ref=oh_aui_detailpage_o07_s00?ie=UTF8&psc=1

 

I’m told these devices kill noise. Looking for additional comments/experience. 

 

 

If your mission is better SQ (kill noise) you probably better purchase a Cisco SG 100D-8 (or 5) V2 from eBay and do the JSGT ?

 

Later you should consider the Uptone Audio, Audio Switch (UAAS) or EtherRegen as it’s called. The switch is under development.