The EtherREGEN thread for various network, cable, power experiences and experiments

[auricgoldfinger]

52 minutes ago, Nenon said:

Emile has tested a bunch of different transceivers with his Taiko Extreme server and reported that he liked the StarTech SFP1000zxst transceivers the most. They are expensive at around $90 a piece, but every now and then they pop up new on eBay for $15-$20 if you are patient. I bought two of them cheap but have not tried them. They work with single mode fiber, so make sure you have compatible cable if you decide to try them.

However, the weird thing is that they are extended range transceiver designed for 80 km distance. Connecting them with a much shorter fiber cable (a few meters rather than 80 km) would have some consequences. For example, the transmit light would be very strong, and it could potentially burn the receiving transceiver module. A workaround to this would be to use light attenuators, but as far as I know Emile, and a few of his customers, don't do that and don't have any issues.

I have access to many different transceivers, including Cisco, Arista, and some third party ones but no desire or time to perform a transceiver shootout. If anyone is really interested and committed to do such thing I might be able to help with some test samples.

 

So maybe industrial transceivers have similar benefits to industrial RAM???

[auricgoldfinger]

2 hours ago, ray-dude said:

I bypassed the stock switch by my AV setup since this improved bass detail and dynamics considerably

 

I would like to know why having other switches in the chain matters if the ER A-side ports are all galvanically isolated from each other and all the noise is gone once the moat has been crossed.  Isn't the moat supposed to make such things irrelevant @Superdad

[auricgoldfinger]

1 hour ago, cat6man said:

 

i'd go with "much is still not understood, grasshopper"

 

I find the idea that the performance of the ER itself can be affected by overall network configuration to be a bit disconcerting, given it is supposed to be independent of such influences as long as it is the last device before the DAC.

[auricgoldfinger]

3 hours ago, sandervdp said:

So my question, in this setup (farad to etherregen) what would be the better option out of these three; the higher gauge Canare 4s6 JSSG360, the Gotham GAC4/1 UltraPro JSSG360 or Farad's own option using lots of thin, tin plated copper wire?

 

Audio Sensibility Signature Silver > Farad Level 2 > the other two options you mention.  Big delta between AS Signature Silver and Farad Level 2.

 

https://audiosensibility.com/blog/sotm-audio/#!/Signature-Silver-DC-Cable/p/102469015

[auricgoldfinger]

1 hour ago, stevebythebay said:

For those of you running with external clocking, have you found that clock cables make a difference?  If so, which ones actually do enhance the performance and SQ of the clock?

 

There is discussion of different cables in the REF 10 thread: 

I use the Habst Digital BNC, which I prefer to the SOtM dCBL-BNC and the generic cables from the engineering supply houses.  In the Mutec thread, people have done comparisons of a few generic cables.

 

EDIT:  At the moment, I am using a Blaxius^2 to re-clock my ER.

 

[auricgoldfinger]

17 hours ago, kennyb123 said:

I just installed a pair of 10 db attenuators on the server side and a pair of 5 db attenuators on the other side. Treble is much improved though still a little hot on recordings that have that tendency anyway.  These are the attenuators:  https://smile.amazon.com/dp/B07R5D72JL/ref=cm_sw_em_r_mt_dp_U_MM5hEbVX60JNC

 

What is the difference between the LC/UPC attenuators you bought and the LC/APC attenuators below?

 

https://www.amazon.com/dp/B07B1KH6DX/?coliid=I4B8CAK6ZS4FP&colid=1O2O3N2HH23ZK&psc=1&ref_=lv_ov_lig_dp_it

[auricgoldfinger]

19 minutes ago, kennyb123 said:

 

I just had to Google APC vs UPC.  Interesting:  https://www.belden.com/blog/data-centers/upc-or-apc

 

This helps:  "UPC connectors are blue while APC connectors are green."

 

The fiber optic cable that I purchased has blue connectors.  I would think that UPC would be the correct choice of attenuator for these.

 

Interesting.  Thanks!

[auricgoldfinger]

2 hours ago, charlesphoto said:

The big question is: how do different db attenuators sound, and would a better quality attenuator (say a Belkin) sound any better? Also, both ends vs just the receiving one? 

 

What about trying less attenuation?

[auricgoldfinger]

DIY types may be interested in this cable:  https://viablue.de/com/cables_ep7_ethernet.php

 

They can also be purchased as finished cables.

[auricgoldfinger]

21 minutes ago, RickyV said:

Have you compared this one to the supra cat 8? 

 

No, I haven't.

[auricgoldfinger]

1 hour ago, soares said:

They look fantastic! And the prices very competitive. Have you tried them?

 

 

I use these cables for my network backbone to provide general hygiene and maintain signal integrity.  They have never been my "audio" Ethernet cable, which I consider to be the last and most critical in the digital chain, even though the materials and construction certainly qualify them as candidates for that role.

[auricgoldfinger]

 

1 hour ago, kennyb123 said:

I think John was speaking of more of active isolation devices and not passive filters.   The Shunyata includes common mode filters - not isolators.  The SOtM also includes a filter I believe.

 

I believe the SOtM CAT7 "filter" is actually a magnet.

 

[auricgoldfinger]

21 minutes ago, Mike Rubin said:

Just using the stock Sonore Trendnet SFP in the Rendu and a similar spec'ed one (juniper-branded Finisar) for the ER.  Is it likely I would do much better with others?

 

These are the latest and greatest Finisar transceivers: https://optical.communications.ii-vi.com/optical-transceivers/ftlf1318p3btl

 

They connect my two switches to one another.  I was previously using the Planet Technology.  I did not like the earlier generation Finisar in my system.

 

 

[auricgoldfinger]

Nothing special.  This is what I use:  https://www.amazon.com/gp/product/B001B1FHJ6/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1

 

The length could possibly make a difference.  I think it's system dependent.

[auricgoldfinger]

1 minute ago, nonesup said:

Well this is another thing to experience.  All the cables he had were the typical yellow ones purchased from Amazon.  Finally I was able to buy two Corning Clear Curve cables of 5m and 10m in the USA (it cost me more to transport than the cables) and this weekend I will test them, I don't know if they will make any difference.

 

I'll be interested to know your impressions!

[auricgoldfinger]

47 minutes ago, Mike Rubin said:

Thanks, guys.  Quick question: The pictured Finisar is the FTLF1318P2BTL, which is what Amazon sellers also depict in listings for the FTLF1318P3BTL.  Is that the same part or do I really want the FTLF1318P3BTL?

 

This is what I bought:  https://www.digikey.com/product-detail/en/FTLF1318P3BTL/775-1082-ND/3507746?

Maybe it would be best to avoid Amazon sellers due to the uncertainty.

 

I don't use attenuators.

[auricgoldfinger]

A word of warning...these transceivers sounded very harsh when I first installed them.  Listening to them was at times unpleasant.  After playing music 24/7 for about a week, they began to sound better.  They needed close to a month of use before finally reaching their full potential.

[auricgoldfinger]

There's also this one:  https://www.fiberoptics4sale.com/collections/corning/products/clearcurvezbl

 

[auricgoldfinger]

1 hour ago, Mike Rubin said:

Alas, even though I am just across the Bay from FOSCO's facility, shipping is more than the cost of a short link. Otherwise, I might well have gone the Corning route, too. 

 

Thanks to @auricgoldfinger, here is what I am using for singlemode, which was $14.39, delivered, for .7m. 

 

https://smile.amazon.com/dp/B008EN3H0A/ref=cm_sw_r_cp_apa_i_trXhFb11HH700

 

I have no idea who makes the innards, but suspect the product description would have said that it is Corning if it is.  It works fine. 

 

I still wonder about attenuation with these short runs of singlemode, though.  I don't have a clear understanding of when it's necessary (I guess it matters more with certain kinds of SFP's) and what bad stuff happens if you fail to use it when you should (SQ issues?  Burn up your SFP?  Burn up your optical Rendu?).

 

I do recall hearing from Sonore that, even for multimode,they recommended using OM1 cables instead of the OM3 that everyone was using a few years ago with Trendnet FMC's because of overheating risk.  I therefore suspect that I should be paying more attention to temperature than I am. 

 

If you're concerned about attenuation, you might consider exchanging the .7m cable for a 5m or 10m cable.  Leaving the cables wound will provide some attenuation.  The tighter the coil (more bends), the greater the attenuation.  If your system is transparent enough, you may be able to hear a difference between the different lengths.