My Experience with Fiber Ethernet

Thanks Harry I appreciate your comments.
I did not post earlier as I don’t like to get caught up in threads that go off in different directions than the op meant them.

I generally believes bits are bits, but also believe everyone has the right to spend their own Money however they want, and it’s one of those strange area’s.

Fiber is an interesting conundrum to me, given how much I work with it and see the benefits it has In a noisy factory environment, where cat 5 and unshielded cat 6 fall over Very quickly.
While our home’s are not generally like that, maybe I should run some fiber between the upstairs and downstairs and try it out in the future.

2 Likes

Michael, you’re triggering a memory of my first industrial experience with fiber.

I’m an old software guy now, but back in the 80’s I also did some electrical engineering. I worked for one of the biggest names in industrial controls, in their research group. Our group was tasked with building a better way to test giant electrical equipment, like huge oil-filled transformers. They were tested outdoors, in an open-to-the-sky courtyard surrounded on three sides by a five-story building. And we had to figure out a way to keep the test computer, a DEC PDP-60, from blowing out every time one of these transformers failed. And these failures were pretty dramatic in their own right.

The plant had an energy management system that scavenged waste heat from the building (overhead lights and such) and stored it in giant tanks in the basement. But there was usually plenty of extra heat, so to dissipate it copper pipes were laid in the sidewalk around the buildings, and the extra hot water was circulated through them to dump the heat. This was in the snow belt, and the sidewalks were always clear, even in the snowiest of winters.

Unfortunately, the courtyard where the tests were done was also underlaid with the pipes, to keep the ice and snow off. And one day a 14-foot transformer failed (along a weld seam, IIRC) under test, and the current found its way to the copper pipes. It burst a tank in the basement, and all the sidewalks had to be re-laid. On the plus side, the fire suppressors worked.

Anyway, the noise problems caused by contactor arcing and other industrial electrical control artifacts were simply overwhelming the PDP-60 (gallium arsenide architecture, I think). We finally had to move the test computer to another building, and connected it with fiber. Worked OK for a few years.

2 Likes

Have you installed fiber into a networked music system? Have you listened to that system with the fiber in place?

You should redo your tests and listen BLIND. Not sure if you then will get the same results.
For me this falls into the same category as cables for some 1000$ (Voodoo) and this is called Placebo effect - patients that are given a pill think they are suddenly doing better.
…also an electrical engineer and tinkering with audio and other hardware since 35 years.

Use the solution that has the least cost.

MM fiber (multimode) is good for 900 yards and 1-400GBe. Speed and distance depends on the OM version.

SM fiber (singlemode) OS2 is good for 200GBe and distance to 10KM.

SM is also available in BiDi configuration so you only need 1 fiber strand. This is helpful when you have a 12 strand pulled. You don’t have just 6 connections available you have 12. It also makes Tx/Rx a lot simpler.

MM with its various OM1-5 variations (and 3 and 4 are the same color) is going way of the dodo. It’s due to to it limitations and that the price of SM transceivers and cabling are now at parity and the fact you can effectively get 2X the number of connections out of a single armored pull.

Have you tested the effect of fiber yourself?

I assume, based on your comments, you are not aware that SmallGreen Computer, the distributor of the UltraRendu, and the manufacturer, Sonore, has posted multiple how to videos on YouTube encouraging the use of fiber to eliminate noise before it can find its way to to their UltraRendu and even their OpticalRendu. They even sell the fiber and transceivers. So apparently they think the UltraRendu can use the help.

Wow Bill that could have been a very bad situation.
Industrial machinery and it’s components required to make things work and get rid of waste can cause terrible situations.

My company used to do chemical based printing (all water based now) and we had some crazy fire’s where the ink or waste would just self combust.

It made for some amazing CCTV viewing but was scary. No copper cables allowed anywhere near the ink mixing machine as you can imagine :roll_eyes:

Or they think they can make money selling fiber and transceivers?

2 Likes

No doubt, but both things can be true at the same time. That is, the fiber works and Small Green Computer sells it. Sonore has long been into reducing noise of all types in a system. Unfortunately, some seem unable to understand noise would be more accurately described as interference rather that something overt like hum. Installing a couple of transceivers, a length of fiber and one clean power supply is not expensive. Almost anyone can try it and listen. And, as I understand it it the fiber offers excellent surge protection as a bonus.

No, I didn’t know that. Interesting, I guess. Thanks. Personally, I’m just using Raspberry Pis and Google CCAs.

1 Like

Yes, that was a nasty one. I should have added that the transformer was from another company; we were testing it under contract.

1 Like

Have they posted any videos showing the actual noise? I had the mR v1.3 and my Pi4 4gb as Roon endpoints.

Feeding SMSL SU-9 to HP amp and my AKG 701’s. Couldn’t tell a lick of difference using the reference tracks from 2L.

The Pi4 has the upside of true galvanic isolation due to it’s WiFi support. No need to run any copper, fiber and all the associated FMC and their own switch mode power supplies.

For giggles I setup my AP for VHT160 on the 5Ghz band and with iPerf was able to hit 60mbps(75MB/s) sustained from my server.

The Pi4 was $99. Cost was less and I have more flexibility.

Wireless, properly implemented, gives you all the isolation and speed needed since playback isn’t generally sensitive to latency. You don’t have to mess around with getting cable to the spot you need.

4 Likes

WiFi can be very good relative to noise if things are properly designed, but I do not believe WiFi can achieve absolute galvanic isolation like a fiber optic cable can. Please see attached a nice, short article on the subject.

2 Likes

Oh it can, what is there to isolate it from on the network side. Standard ethernet can as well via isolators, no fibre necessary. I’ll not argue that the fibre network ideas here aren’t pretty neat though.

Everything I have read says EMI is an issue with WiFi. The subject is an interesting one. here is another article that you might find interesting:

StackPath.

Can you outline this in more detail?

Well depending on your POV WiFi IS EMI :wink:

3 Likes

Ok. I’ll bite:

What’s your definition of galvanic isolation? Wifi is 100% air gapped.

1 Like

As far as Darko is concerned: He doesn’t realize that he doesn’t know what he’s talking about. Unfortunately people will decline to employ any critical thinking and swallow hook, line, and sinker.

Hit the Siemons paper titled “The Antenna Myth” vs his statement “(Side note: the Ethernet cable itself can act as an aerial, attracting airborne electrical noise)”

Now if you have a product, that regardless of cost, has an improperly designed wired or wireless connection then that’s another conversation.

3 Likes