EtherREGEN: Installation, Usage, Difficulty, Questions thread

[Superdad]

2 hours ago, Confused said:

 An easier approach would be to simply switch between the two by swapping the INT / EXT switch whilst the EtherRegen is powered up and running.  Is it OK to do this?  I am a little worried the EtherRegen might not like this for some reason and I might junk something, but maybe I am being over cautious here? 

 

Sorry, but as @Dutch pointed out from the User Guide, changing the INT/EXT switch while the EtherREGEN is powered on will not work. It won’t harm anything, but it won’t alternate the unit’s usage of internal or external clock reference. Of course no clock signal runs    through that little slide switch—its position is simply looked at once when the unit is powered up, to tell the ‘B’-side microprocessor which code-set to load to the clock synthesizer such that it programs to run from either the internal 25MHz Crystek CCHD-575 or from an external 10MHz clock. After power-up programming the microprocessor shuts off and the position of the INT/EXT switch is ignored.

[bailyhill]

I would also suggest, with what has been posted about the thermal settling of crystal oscillators, that the only way to make a proper comparison is to power up things and let them set for a while, like maybe a day and then listen.  After getting a good measure of SQ, then change to the external clock, let it settle for a day and listen again.  AB testing will not be very helpful in this case for sure.  

[RickyV]

3 hours ago, Confused said:

I am currently using my EtherRegen with a 10MHz reference clock signal from my Mutec REF10.  This works fine and sounds good, so no issues.

 

I would like to try a quick comparison between using the REF10 and using the EtherRegen's internal clock.  So far, when I have swapped between the internal and external clock, I have taking the cautious approach of powering down the EtherRegen, switching between INT / EXT clock, then powering up again.  An easier approach would be to simply switch between the two by swapping the INT / EXT switch whilst the EtherRegen is powered up and running.  Is it OK to do this?  I am a little worried the EtherRegen might not like this for some reason and I might junk something, but maybe I am being over cautious here? 

 

And maybe disconnect the clocking cable from the ER, ground loops.

 

If the ER is still warm it doesn’t matter if you disconnect and apply power again imo you don’t have to wait a day. If the ER is cooled down then you’ll have to wait a day or so before everything is stable again.

[Arpiben]

4 hours ago, Superdad said:

 

Sorry, but as @Dutch pointed out from the User Guide, changing the INT/EXT switch while the EtherREGEN is powered on will not work. It won’t harm anything, but it won’t alternate the unit’s usage of internal or external clock reference. Of course no clock signal runs    through that little slide switch—its position is simply looked at once when the unit is powered up, to tell the ‘B’-side microprocessor which code-set to load to the clock synthesizer such that it programs to run from either the internal 25MHz Crystek CCHD-575 or from an external 10MHz clock. After power-up programming the microprocessor shuts off and the position of the INT/EXT switch is ignored.

 

In External clock reference mode, what is happening in case of no 10MHz signal input ( Reference Clock device failure/Off, BNC removed)? Thanks.

[Superdad]

28 minutes ago, Arpiben said:

In External clock reference mode, what is happening in case of no 10MHz signal input ( Reference Clock device failure/Off, BNC removed)? Thanks.

 

No clock = no data flow. 

[Confused]

Thanks to all for the input to my question post #607, these things are rarely as simple as you might think, and the various responses have helped clarify my thinking here.

 

Considering everything, I think that the best way to try an A/B type test using the REF10 would be to set up the EtherRegen running on its own clock, it can then be left running for hours or days for clock and everything else to stabilise 100%.  Then switch to the REF10 providing the clock signal, this only requires turning the EtherRegen off for a minute or so to flip the INT / EXT switch and connect the REF10's clock cable.  The REF10 itself is feeding other kit, and I can't think of the last time I actually turned it off, so no time is needed for the REF10's clock to stabilise, the whole set up should be stable very quickly.

 

It should be an interesting test.  No disrespect to the EtherRegen, but the REF10 has the lower phase noise clock, after all, it is a Euro 3500 product and just a clock, but the EtherRegen does have a very good clock, with the advantage of being located exactly where it is needed on the board, the REF10's performance is necessarily restrained by having an external clock cable.  I suspect there will not be much in it, this is a case of a very low phase noise clock versus another very low phase noise clock in an upstream device.  This is not the thread for listening reports, so once I have tried this and reached some kind of conclusion I will post on the normal listening reports thread.

[savjam]

Just installed.  Installation was hassle free and couldn't have been simpler taking just a few minutes.

• Current Set-up: Router to Mac Mini using a generic ethernet cable; Mac Mini (in bridge mode) to DAC ethernet input using an SOTM dCBL-CAT7 ethernet cable.
• EtherRegen Set-up: Router to EtherRegen A port and Mac Mini to Ethernet A port using generic ethernet cables; EtherRegen B port to DAC using the SOTM ethernet cable. 

Will let play for awhile before making any conclusions.

[wizardofoz]

@savjam I don’t understand your setup as macmini has only one Ethernet port? What DAC are you using? What is your streaming software...and how are you getting the 2nd Ethernet port - via usb or thunderbolt Ethernet adapter perhaps?

 

edit ah I see Alex suggested you not use the thunderbolt adapter back in September, and from other posts it’s Roon...etc

 

a case of TL;DR I guess.

[savjam]

54 minutes ago, wizardofoz said:

@savjam I don’t understand your setup as macmini has only one Ethernet port? What DAC are you using? What is your streaming software...and how are you getting the 2nd Ethernet port - via usb or thunderbolt Ethernet adapter perhaps?

 

edit ah I see Alex suggested you not use the thunderbolt adapter back in September, and from other posts it’s Roon...etc

 

a case of TL;DR I guess.

Hi Wizardofoz.   To confirm,

Yes, I was using the Mac mini in bridge mode to provide ethernet access to the DAC via a Thunderbolt to Ethernet adapter since the Mini only has one Ethernet port.  

Yes, Roon is my streaming software.  
 

My setup with the EtherRegen is as follows (I am repeating here since I had a small typo in my previous post):  Router to EtherRegen A port and Mac Mini to EtherRegen A port using generic ethernet cables; EtherRegen B port to DAC using the SOTM dCBL-CAT 7 ethernet cable. 

[Pokey77]

Not sure if this is the right thread to post this question.

 

I've read the listening impressions thread from beginning to end and I have the following question.

 

Is the EtherRegen appropriate for use with an Aurender N10? I've only seen one person comment they use an Aurender. See my system config below.

 

ATT Fiber optic from the telephone pole to inside the house and to a BGW210 ATT-supplied router/modem. Monoprice Cat 5e 3' cable to an inwall run of Cat 5e run to behind my system and then 3' Cat 5e to N10.

[Superdad]

4 minutes ago, Pokey77 said:

Is the EtherRegen appropriate for use with an Aurender N10? I've only seen one person comment they use an Aurender. See my system config below.

 

I see this is your very first post here on Audiophile Style. WELCOME!

 

Yes, the EtherREGEN is ideal for use feeding music servers such as your Aurender N10. In your simple system you will just run your current Ethernet cable into one of the 'A'-side ports of our switch, and then another Ethernet cable from the lone 'B'-side port into your Aurender (assuming your N10 is the component directly feeding your DAC).

You should hear a difference right away, and, per all the user reports, the SQ will change a bit further over the following 100~200 hours of use.

 

Our products are sold with a 30-day money-back guarantee, so there is little risk to trying an EtherREGEN in your system (Though I must warn you: Out of 500 units delivered so far, only 2 people have asked to return. ).

Thanks and regards,

--Alex C.

[stevebythebay]

43 minutes ago, Pokey77 said:

Not sure if this is the right thread to post this question.

 

I've read the listening impressions thread from beginning to end and I have the following question.

 

Is the EtherRegen appropriate for use with an Aurender N10? I've only seen one person comment they use an Aurender. See my system config below.

 

ATT Fiber optic from the telephone pole to inside the house and to a BGW210 ATT-supplied router/modem. Monoprice Cat 5e 3' cable to an inwall run of Cat 5e run to behind my system and then 3' Cat 5e to N10.

Are you using any switch today, or are you simply connecting the N10 the long run from the router/modem?  I'm assuming the N10 is currently wired to your DAC via AES/EBU, S/PDIF, or Coax..  Is your music inside the N10 or are you getting it from a NAS or via the USB port on the N10 with an external hard drive?  Or is the network simply allowing you to stream from the internet for services like Qobuz?  Connecting the B side of the eR to your N10 Ethernet port will provide benefits in cleaning up any A side attached network devices like your router/modem and NAS.  Beyond that, unless I'm missing something, you're one of few who are not streaming into an Ethernet enabled DAC environment.  However, using AES/EBU with the N10 into your DAC is qualitatively better than any other option, I think.

[Puma Cat]

Consistent product robustness problems with the the SFP Cage and Optical Transceiver on the EtherREGEN:

 

Since I received my ER back in October, I've experienced consistent problems with the SFP cage and optical transceiver on EtherREGEN. I am using the Sonore SystemOptique-certified 850 nM optical transceiver (purchased from Sonore) in the ER's SFP cage. 

 

My experience is that even if barely move, or in some cases, even touch the ER or the Tripp-Lite OM-1 specification optical cable that is plugged into the OT, the OT almost always disconnects in the SFP cage, and then I usually spend a considerable period of time trying to get the digital audio steam from my server to connect.

 

Today, after placing my digital streaming front end a nice maple board, I just spent >90 minutes trying to get ER to connect up. To say it was exasperating was putting it mildly, I was just about to give up and spend the afternoon listening to LPs when it "connected" again and I could stream my files to my SOtM SMS-200 UltraNeo. 🙀

 

Back when was I was working as Design for Six Sigma (DFSS) Master Black Belt, teaching product development principles to scientists and engineers for one of the leading biotech companies in the world, there was a key quality attribute that I spent a lot of time drilling home: the concept of a ROBUST product or system-level ROBUSTNESS.

 

A "ROBUST" product is defined as:  "Function performs on-target with minimal variation in the presence of NOISE factors."

 

The noise factor here is the optical transceiver module, and presently, the SFP cage/OT interface of ER is not robust to the optical transceiver utilized.

 

ER should work in a statistically robust manner with any industry-standard specification 850nM, 1000-Base SM MMX fiber optical transceiver, and in my experience, with the number of times that I (and for reference, a good friend that has also an ER) have had the problems described above, this is not the case. 

 

I know folks will chime in, here..."Oh, don't use that optical transceiver, use this OT, instead."

 

Look, gang, I get it, and I'll have to do that, I guess, in order to have a consistent and less exasperating customer experience. 

 

But, so that I am maximally clear here: that is not the point of this post.

 

The truth of the world is this: There are NO, and I mean, NO perfect manufacturing processes or perfect products, from any manufacturer in any industry. None. They do not exist.

 

The Japense have a very specific term for their philosophy about designing and manufacturing products: KAIZEN (Change for the Better, ref here: https://en.wikipedia.org/wiki/Kaizen)

 

The key point about kaizen is to work with humility and open-ness with a spirit of "continuous improvment" so that overall product quality increases continually over time. When product quality improves, the customer's ability to use the product and their experience improves, and society as a whole improves, too. 

 

Understanding this key principle with humility rather than indignation is the basis for one of the highest quality manufacturing systems the the world, The Toyota Production System (reference: The Toyota Way by James Likert).

 

This post was written in the spirit of kaizen with respect to helping Uptone Audio developing what is truly an excellent product into an higher-quality one, specifically with respect to product robustness. One that will realize its full and considerable potential all the while providing 1) a function performs on-target with minimal variation in the presence of noise factors and 2) a satisfactory customer experience.

 

Right now, its not doing that. 

 

Improving ER's "robustness quality attributes" will not only improve the customer experience, it will also lower service, support, repair, and warranty costs, a lower COPQ (Cost of Poor Quality), and result in greater profitability for Uptone Audio. And that's a good thing; I sincerely want Uptone Audio to be maximally profitable and successful as they bring a great deal of innovation and value to the audio industry. And, they're really nice guys, too. 

 

Alex, happy to work with you and Uptone Audio as a professional DFSS resource regarding the SFP/OT robustness quality attribute in any way I can. Feel free to reach out. 

 

all my best, 

Stephen

[Superdad]

1 hour ago, stevebythebay said:

Beyond that, unless I'm missing something, you're one of few who are not streaming into an Ethernet enabled DAC environment.

 

Hi Steve:

I do not understand your above sentence.  There are lots of people using the EtherREGEN with computer/server/renderer endpoints (such as Aurender, Auralic, Innous, Sonore, SOtM, Taiko, etc.)--both with and without local music storage. And whether the connections from those endpoints to the DAC is AES, S/PDIF, or USB is not relevant to the EtherREGEN doing its job. 

[Superdad]

1 hour ago, Puma Cat said:

The truth of the world is this: There are NO, and I mean, NO perfect manufacturing processes or perfect products, from any manufacturer in any industry. None. They do not exist.

 

The Japense have a very specific term for their philosophy about designing and manufacturing products: KAIZEN (Change for the Better, ref here: https://en.wikipedia.org/wiki/Kaizen)

 

The key point about kaizen is to work with humility and open-ness with a spirit of "continuous improvment" so that overall product quality increases continually over time. When product quality improves, the customer's ability to use the product and their experience improves, and society as a whole improves, too. 

 

Understanding this key principle with humility rather than indignation is the basis for one of the highest quality manufacturing systems the the world, The Toyota Production System (reference: The Toyota Way by James Likert).

 

Hi Stephen:

We agree with your 100%.  

I am sorry that you have had such a hard time with the combination of the SFP optical transceiver from Sonore mated to the SFP cage of your EtherREGEN.

You are not the first person to have difficultly with the optical connection using that transceiver, AND to-date that model of transceiver is the only one for which we have record of difficulty.

 

The Amphenol SFP connector and transceiver used in the EtherREGEN are items chosen for their quality and not for their cost.  And the PCB positioning of the connector inside and relative to the cage is a precise, manufacturer specification which we adhere to. It might be possible for us to move the connector forward about 0.1mm, but we are not particularly inclined to do so and it might not solve the issue with the particular brand of SFP module you are using.

 

I have just now ordered for you this $14 SFP module: https://www.amazon.com/gp/product/B00U77VPX2/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

It will be delivered to your address tomorrow. I have been using these in production testing and they have worked fine.

 

Please try it and let us know if that solves your problem. If it does not, then permit us to exchange your EtherREGEN board for a new one. Life is too short to have to suffer exasperation with your audio toys. 

 

Thanks and regards,

--Alex Crespi

[stevebythebay]

Sorry.  Wasn't examining the possibilities.  I guess if the N10 is receiving data streams from one side or the other of the eR, it then moves the results via other means to the downstream DAC.  The N10 requires a wired Ethernet connection to the wifi/router, as there is no WiFi capability in the N10 (and using an iPad for instance, communicating via the wifi/router allows for managing playback).  So, this would require a connection from the A or B side into the N10's network port, likely using the B side, and allowing the A side to support multiple other Ethernet connections (NAS for instance, and of course the WiFi router).  Not sure where the "biggest bang for the buck" sound-wise would come in making connections, though I'm inclined to favor B port to Aurender.  That way the eR maximizes the quality of all upstream bit streams that get fed into the N10. 

 

The confusion I had is my own point of reference (my network and system) and failing to acknowledge the flexibility and possibilities in using the eR in so many scenarios.

[Pokey77]

Superdad + Stevebythebay,

 

I appreciate the welcome.

 

My setup is pretty straight forward as you can see. I simply stream Tidal and so am looking to clean up the signal more before it gets to the N10. I plan initially to run out of the router to the A side of the ER, then out the B side and through the wall to the N10 as a first experiment. Then secondly I would try moving the ER behind the rack in my listening room, but ideally it'd be in the other room far away from my system. In the second position, I could run my network connection for my TV out of the A side. In the end, whatever place the ER ends up will be where it sounds best. Then after that I can work with Supra 8+ network cables I've been breaking in via nightly Netflix.

 

BTW, I am running AES/EBU to my TAD dac. To me, it is ever so slightly better than USB.

[wizardofoz]

Alex, and anyone else that might like to chime in.

 

I have an ER running optical from my Unifi SW24 SFP ports, Using FS.com modules. On the B side is a microRendu connected to a USB DAC (OPPO HA1 or Sonica) and this is fine for one setup. There is also an Allo USBridge that I have connected to the A side, and potentially a Roon ROCK core or Windows 10 Core or MacMini running Roon and god forbid a NAS too.

 

I'm considering an Allo USBridge Signature for a downstairs system and was wondering if I would be better to connect this optically to A side and move the ER to a wired port. I have a FS.com FMC that I could use on the ER A side optical port with a 15M fibre run to the other area. Optionally I also have a wired connection to the other area (Cat6) that I could use too on the A side.

 

lots of options...DAC on the USBridge will probably be Denafrips Ares II or Soekris DAM 1021 R2R's. 

 

 

[octaviars]

2 hours ago, Pokey77 said:

BTW, I am running AES/EBU to my TAD dac. To me, it is ever so slightly better than USB

 

Nice to see another TAD owner here, is it a D1000mk2 you use?

 

I use a dCS Network Bridge with AES to my TAD, the eR is feed with a 1.2m long CAT7A LAN cable from the B-side. On the A-side i have my Roon server and fiber to my router.

[jcn3]

8 hours ago, Pokey77 said:

Superdad + Stevebythebay,

 

I appreciate the welcome.

 

My setup is pretty straight forward as you can see. I simply stream Tidal and so am looking to clean up the signal more before it gets to the N10. I plan initially to run out of the router to the A side of the ER, then out the B side and through the wall to the N10 as a first experiment. Then secondly I would try moving the ER behind the rack in my listening room, but ideally it'd be in the other room far away from my system. In the second position, I could run my network connection for my TV out of the A side. In the end, whatever place the ER ends up will be where it sounds best. Then after that I can work with Supra 8+ network cables I've been breaking in via nightly Netflix.

 

BTW, I am running AES/EBU to my TAD dac. To me, it is ever so slightly better than USB.

 

alex and others have repeated said that the ER needs to be as close to the endpoint (the N10) as possible, so putting it behind the N10 would be best.

[simon_pepper]

If it helps anyone, I have mapped out how the music playback path within my network is connected, separating WiFi traffic, from other device traffic.
(taken out backup NAS units, printers, networked UPS, networked AV equipment etc.)

 

[mourip]

I am considering an ER but am not certain of its best use in my setup. I am a bit late to this thread and have skimmed looking for a match for my own system but did not really see one.

 

I have a Roon server PC (internal library and Qobuz streaming via the Internet) with a NUC endpoint out to the rest of the system. My server has a JCAT Femto ethernet card in bridged mode so that one port goes directly to the NUC and the second is attached to a small Cisco switch with some other non-audio ethernet connections and of course Internet/LAN access.

 

I have tried very hard to have a "clean" side and a "dirty" side to isolate my system. Until recently I used two FMCs to isolate my server from the LAN. After doing some experimentation I found that a straight ethernet cable sounded better than the fiber in spite of my theories.

 

When the ER became available I figured that it would be perfect to isolate my server from the LAN. After reading this thread it now sounds like I should use it between my Roon server and my NUC endpoint (I even have a REF10 to use). I guess that I could buy two ERs however I am not sure my CEO will like that 🙂

 

Am I understanding its use correctly?

 

Best,

Paul

[Pokey77]

On 1/6/2020 at 7:20 AM, jcn3 said:

 

alex and others have repeated said that the ER needs to be as close to the endpoint (the N10) as possible, so putting it behind the N10 would be best.

I do understand that and that is why I posted what I did about my system setup. I'm going to try several different positions with the ER. I'd prefer for it to be in another room where I don't care about the heat is generates. However, I will put it where it ultimately sounds the best.

[Pokey77]

On 1/6/2020 at 1:44 AM, octaviars said:

 

Nice to see another TAD owner here, is it a D1000mk2 you use?

 

I use a dCS Network Bridge with AES to my TAD, the eR is feed with a 1.2m long CAT7A LAN cable from the B-side. On the A-side i have my Roon server and fiber to my router.

I have the Evolution setup - TAD C2000, M2500, E1. I have a very simple network setup now - Fiber from pole to inside house, not sure if it is fiber to the ATT router (could be copper), cheap monoprice copper from router to junction box, then a run of around 15' Cat5e through the wall to another junction box behind my stereo, then cheap copper (monoprice) to an Aurender N10. I have tried Supra CAt8+ and found it to have a better midrange but it is not as good in the top end. So I have those two 1M Supra cables being broken in on my TV doing streaming duty for a few months and will take them back to the audio side and see if that helped.

[LowMidHigh]

1 hour ago, Pokey77 said:

I do understand that and that is why I posted what I did about my system setup. I'm going to try several different positions with the ER. I'd prefer for it to be in another room where I don't care about the heat is generates. However, I will put it where it ultimately sounds the best.

 

I experimented with 0.5 meter and 20 meter cat6. If there was a difference, I couldn't pick on it. 

 

What is the working theory behind the adverse effect of reasonably long cat6 or higher  cables? Or there's none?