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

[One and a half]

22 minutes ago, rossb said:

I have been using the etherRegen without an issue for months. However, today it started to produce an odd result. When the eR was connected to my network, the internet connection would drop out. If I disconnected the eR to the network, the internet connection would reappear. I tried this multiple times to confirm that I was not imagining this. Presumably there is some kind of conflict between the eR and my modem/router. Any thoughts as to how to resolve this?

It's summer, the ER is too hot, even under cooler conditions. Find a USB fan, blow it over the case and see if the ER keeps going?

[One and a half]

7 minutes ago, rossb said:

Thanks, but that’s not quite the problem. The eR is working fine. But it causes my modem/router to lose the internet connection when it is connected to the network.

 

You are monitoring the WAN connection at the router to draw this conclusion? 

[One and a half]

2 minutes ago, rossb said:

Yes, the internet connection drops out from all PCs and other devices on the network.

The router has logs, any clue in looking at those? Some of them are gobbly gook, some are actua;ly written in English. Perhaps extract and post as an attachment.

[One and a half]

@JohnSwenson, reading the post about the EtherRegen’s temperature of a nominal 50C as a ‘sweet spot’ for CMOS stability, is the ambient reference 25C? 
Thing is if the ambient is hot like 35C will the temperature increase beyond the 50C?

My question is to ask whether there is a need to force cool, arrange free air around the case or perhaps there’s a Temperature regulator looking after this.

[One and a half]

3 hours ago, JohnSwenson said:

The 50C is where two competing internal factors reach an optimum. Higher temperatures make CMOS switch slower, which increases jitter in the receiver, BUT causes less noise in the internal power/ground network in the chip. The effects of these two tend to "crossover" at about 50C. This is not a sharp narrow peak, more like broad hill. Thus getting in the 40-60 range is fine.

 

So do whatever it takes to get around 50C. In my lab which is usually running around 22C that happens with the EtherREGEN sitting flat on the bench. This is without lots of other high bandwidth stuff on the A side. If you are running several high bandwidth devices plugged into the A side, the interior heat generated will be a little higher. If your ambient temperature is higher (you live in the tropics without AC etc) or the ER is inside a closed cabinet etc, you may need to either put it on its side OR use a heat sink of some sort.

 

So put it in the normal environment, if the case is between 40C - 50C, don't worry, leave it alone. If it is hotter than 60C then do something to cool it off. If it is cooler than 40C (you live in Siberia and like to keep the windows open), you should insulate it somehow, I don't think this will happen to too many people.

 

Does this make sense?

 

John S.

That’s a workable response and can manage the two of extremes and a rise of 25-30C on the safe side. It’s trivial for me to control the temperature , the temperature ranges 5C - 40C at the extremes and the EtherRegen would only be active when the audio system is on and usually not above 32C. 
Didn’t plan on a cupboard, there’s room behind the Lumin which is open air, a few weeks of measuring and the optimum air flow should keep that temperature stable.

 

Thanks for the great response!
 

 

[One and a half]

With an impending delivery occurring for the EtherRegen, thought to put feeble brain to paper and see how to connect the device to the rest of the system.

 

With all of the cabling shielded including AC power lines, may as well continue with the signal and 9V power lines to the EtherRegen and the ifi Micro USB 3. 

 

To avoid rising voltages caused by shield currents, the 9V lines are short. Shields are connected at the plug end as well bonded to the metal frame of the junction power box. The large 6mm (10AWG) bonding the Lumin, ER (A-Side) and the junction box maintain the same potential  to the earthed 0V at the Acopian power supply.

 

The thought of a separate LPSU for the Micro USB 3 and Ether Regen did receive some thought. Not only the cost, but potential leakage paths across the AC line and bonding the two 0V togther would have created a worse problem. The noise from the A side of the ER is worse than the Micro USB in any case, a pragmatic solution with just the one PSU. In any case the noise form the A side would be far less less than the standard ifi iPower SMPS for the Micro USB3. 

 

The STP (Shielded twisted pair) of the Lumin bonds the B side to the same ground as the Lumin, antenna effects are minimal due to the short cable and being shielded anyway.

 

The Acopian supply (9V, 6.5A)  has two sense terminals, like the JS2,  that compensates for long leads voltage drops, so the desired voltage won't drop (too far).

 

Normally the audio system is powered down with the Music server always on. IP addresses are assigned by the router to the MAC address of the one of the JCAT Net Femto NICs. The other NIC is DHCP for the Lumin, now ER.. so the IP address should be transparent from the JCAT through to the Lumin, apart from a speed change, 1GB/s to 100MB/s.

 

The ER & Acopian will power up with the rest of the sources, the Lumin takes some time to establish a network connection, so the ER should be functional by then.

 

Networked audio EtherRegen _ USB3.pdf

[One and a half]

14 hours ago, R1200CL said:

I’m quite sure you’re braking the moat by this setup. Mostly by using shared power and probably the creative way to ground. (In addition Vbus power is a challenge). 

In your case I can’t see any reason to use the ground screw on the EtherRegen. Read the user manual.

 

From recent discussions here, the last few pages, the ground thumbscrew is on the A side only. So any voltage differences on the A side Ethernet shells and the Lumin's case won't be, that's the whole idea of that 6mm wire.

 

14 hours ago, R1200CL said:

You could also test with fiber in. Adding fiber interface to your PC. Or even use the opticalModule.

 

 

Without starting a discussion about the iFi Micro USB 3.0, I take the chance to suggest to remove it. At least as a test.

USB cables is an issue as well. And good Vbus power. I guess a USPCB won’t fit in your system or ?

 

This goes against everything we learned about how to do ethernet wiring. Allowing possible loops to travel along the connected shield of the ethernet cable. 

 

You may have a problem here. Data sheet Lumin says only 1 GB. @wklie can you comment on that ?

 

Further you may benefit a lot to reverse the EtherRegen and use fiber into Lumin. 

Will be interesting to hear how things works out.

 

BTW: You using both 230 VAC as well as 110 VAC ?

An other area of concern or ?

 

 

 

[One and a half]

EtherRegen listening evaluations day 5 (90 hours burn in)

 

Here the listening tests is a CD Rip of "Thom Rotella Band" Self titled CD from 1987 from DMP DR10 compared to the CD.

 

Blue corner : CD setup: Accuphase DP-720 SACD/CD Player XLR Analog out -> Pre-amp

Red Corner : Computer Audio: Music Server JCAT NET Femto -> 35m Cat5e -> ER Side A -> ER Side B -> Accuphase DC-950 DAC -> XLR Out -> Pre-amp (ER PSU = none other than ifi iPower 9V 1.5A)

 

The ifi power is used overnight to keep burning in the caps. If the ER stays, I will work out a system to keep the ifi on standby to the ER when the main audio system is off. If the main audio system is on, the ifi is closed own (from the AC side) and the linear supply kicks in. That way no leakage currents can flow anywhere from that rotten SMPS. The ifi is AC powered from a standard TN system 120V network and is separate to audio AC power (by a long shot).

 

The DC-950 has a superior DAC formation than the DP-720, so the two aren't quite equal. When the DP-720 proprietary HS-Link is used to the DC-950, it's in another class again. For the exercise, both sources can be switched A/B on the run.

 

ER: The sound stage has definitely changed form the initial 3 hours run. The image now at 90 hours is roughly like this:

1. The centre image is a bubble about 1m across, speakers are 3.5 m apart.

2. Left and right speaker have their own content about 0% off stage and 20% in stage (I mean here 20% from the speaker to the centre)

3. There's a gap between centre and mid way of that bubble to the physical speaker (about the same left and right).

 

Accuphase:

All the gaps are filled with the sound stage spread over the two speakers with a little spilling over the sides. The height here is a tad higher, not much maybe 1/2 foot. Since the soundstage is across the speakers, instruments are readily defined in space.

 

I'll swap out the ifi supply with the linear 9V, 6.5A and see what happens.

 

 

[One and a half]

EtherRegen listening evaluations day 5 (90 hours burn in) (update)

 

Substituted the ifi IPower with the 9V 6.5A Acopian PSU. Also changed to "Karla Bonoff Greatest Hits" SHM-CD and CD Rip of the same same situation red and blue corner.

 

No real differences, pretty much the image is as the same as post 1120799. Uptone Audio's guidance is that the power supply doesn't matter, which I would tend to lean to, with the one caveat is that if it is an SMPS that is it sourced from a totally different AC supply than the remainder of the audio system.  

 

On the signal side, there's a CAT6 STP cable between the Lumin U1 and the ER. 

 

Erratum:

Red Corner : Computer Audio: Music Server JCAT NET Femto -> 35m Cat5e -> ER Side A -> ER Side B -> Accuphase DC-950 DAC -> XLR Out -> Pre-amp (ER PSU = none other than ifi iPower 9V 1.5A) is not right.

 

Should be:

Red Corner : Computer Audio: Music Server JCAT NET Femto -> 35m Cat5e -> ER Side A -> ER Side B -> Lumin U1> Ifi Micro USB3 -> Accuphase DC-950 DAC -> XLR Out -> Pre-amp (ER PSU = none other than ifi iPower 9V 1.5A)

 

 

[One and a half]

On 3/25/2021 at 3:44 AM, MasterWarzombie said:

My Oyaide DB 510 is burning in (300 hours); it's silver.
Impeccable quality of finish, Oyaide connector (the same as those supplied with the JS2).
I like the association with my external clock / ER.

I have this cable as well, awesome quality and response!

[One and a half]

5 hours ago, Johnnydev said:

is this combination with your clock and ER better than your AfterDark. Project ClayX Black River Carbon Fiber Woven 75ohm BNC Clock Cable?

I don't have a clock, but used for spdif connection, cheers.

[One and a half]

3 hours ago, JohnSwenson said:

One thing to try, have you listened to a UTP cable instead of the the STP cable?

 

John S.

As soon as the power comes back on will try this. It's now out close to 90 minutes, there was a loud growl from a transformer a few blocks away, and a single reclose that failed, so could be out for some time yet.

 

From the initial 3 hour assessment, there wasn't much of a change with UTP/STP, things might change 😉

 

The Lumin ethernet shell is earthed so that will mean the B side shell is also earthed with the STP. There's no bonding cable just yet between the Lumin and EtherRegen since the 0V is earthed back at the 9V LPSU end.

 

While the power is out can make up the bonding cable and take some low ohms readings.

[One and a half]

EtherRegen listening evaluations day 6 (112 hours burn in)

 

Once the power was restored, resetting lots of clocks, Internet, all good to go.

 

Listened to Kara bonoff again, time for a change to The Doobie Brothers "Livin on the Fault Line" SACD rip vs SACD DP-720 player.

 

With the STP from ER to Lumin U1, could have been additional time, the bubble in the middle was wider than before. OK, shutdown the Lumin and swapped STP Cat 6 with a generic Cat5e 1m cable. This is closer to the SACD, missing those details around voices, image height is still a little down, but 'could live with it'.

 

Enough of subjective data, now for the OBJECTIVE DATA 🙄

 

With a Fluke 289 (I like when Fluke names their models after famous engine sizes), I measured from the +9V of the PSU to the shell of the XLR at the DAC. Reading was 9.063V.

 

Removed the ER out of circuit totally and disconnected its 9V power. So now the Cat5e from the music server went direct to the Lumin U1 as before. With music playing, the voltage at the same points was 9.003V. Across the power supply was 9.003V.

 

So with the ER is circuit it adds voltage of 600mV, this is with the UTP cable.

 

If the ifi MicroUSB3 added noise it would also appear from the 9V to the XLR shell, but it's not there. 

 

The SUBJECTIVE data now is that the soundstage is restored across the two speakers, no bubble any more. There's a VERY fine line between resolution of the Lumin system and the SACD player. Like before the ER, I could listen to either and be 'OK'.  There's not much point now to keep going with the burn in on the ER for a month to maybe reduce the voltage further, when a direct connection from the JCAT is the optimum and put the lounge back together.

 

 

[One and a half]

I need new glasses, the differential is 60mv, not 600 mV, duh!

[One and a half]

8 hours ago, R1200CL said:

I’m thinking using same PS for EtherRegen and the Micro USB3 Isn’t optimal. Maybe worth testing out separate PS ?

The problem of adding the voltages are not due to the power supply, since the adding of Volts is via the signal leads. So even if the ER was supplied by a separate power supply, the higher voltage on the XLR shell would still be there.

 

I should have measured the AC component on the DC rail, in retrospect to find out where that signature came from. In any case removing the ER collapsed the added voltage. 

 

I see where you’re coming from in that having the two devices on the same supply in parallel causes the added voltage as they interact? That would be true if the power supply was ungrounded, but all the XLR, USB shells, Lumin and ER chassis points all measure the same at 9.003V, so all grounds are tied to earth (PE) solidly.

 

The B side is uncoupled from the A side. The only other way to increase voltage on the shell is via the Lumin, but without the ER the added voltage doesn’t exist, so it cannot come from there.
 

[One and a half]

Of any use, a repeat post the other day on power supplies:

 

The EtherRegen doesn't care what feeds it, really. What matters is what the power supply does to the rest of the system is where the differences are heard. This is based on listening using two PSU a linear supply AC fed from audio balanced supply, and the Uptone SMPS on an IT only 120V AC. Different devices, with very little difference with the audio.

 

Depending on the power supply, it produces switching noises which couple to earth and feedback through the AC network, and to DACs or amplifiers for that matter, where it the effects are audible. Switching may be regulators or bonded rectifier diodes or FETs, or an ICs that are capacitively coupled to the chassis of the power supply. It's how the PSU filters these artefacts will determine how it sounds.

[One and a half]

EtherRegen listening evaluations day 10 (252 hours burn in)

 

Last post saw elevated voltages on the normally grounded connectors with the ER in place at 6 Days burn in. Today, finally, the voltages are under control, there's maybe 1mV DC offset between the 9V on the PSU and any point that references to 0V, XLR sheels, USB shells, Lumin ground point. The BNC connector floats above ground, which it should for clocking in case anyone has an interest there.

 

How does it sound? Like it's supposed to, wide sound-stage, instruments spread out with no holes in the image, just a seamless landscape. So there's a correlation between elevated voltages on the grounds and sound stage, interesting!

 

Comparing SACD/CD to ripped versions of the same, it's razor close. CD is a fraction warmer presentation, that's about it. I would consider the EtherRegen burned in. 

 

Changed the wiring from the power supply to the ER and ifi Micro USB3, used a redundant length of Furutech AG power cable (14 gauge SPC Shielded) and parallel feeds with 2 core and shield instrumentation cable, guess at 16 AWG. Sensing cable was an old Monster Bass cable, 2 core and shield. 

 

There's also a 1m STP between the B side and the Lumin input, will leave this in for a while.

[One and a half]

8 hours ago, PYP said:

Is the eR powered all the time?  Just curious since it sounds best that way in my system.  

Well, this morning, the EtherRegen was powered down all night. Left the power on 30-45 minutes before listening, seems to have stabilised during this time, it's about the same time for other electronics to settle down. I would say all attributes have returned since yesterday. Will work on a changeover system when the main system is off the EtherRegen stays on from another PSU and switches to the linear supply when listening. Would be nice to have some data go through, but that's really messy.

[One and a half]

1 hour ago, agladstone said:

Any thoughts on why I need to constantly turn off and turn back on (reboot) my eR? 
My Aurender which the eR is feeding tells me there is no IP and I need to reboot the eR almost every listening session to get my Aurender to register an IP address again? 
I leave my eR and Aurender powered on at all times. 
@Superdad 

I wonder if the router is refreshing the lease, some do after 24 hours. 

Some routers allow a refresh time of 999 hours, others don't have a setting at all. If the latter, perhaps assign the Aurender MAC address a permanent IP address, in that way when the refresh occurs, the assignment works right away.

[One and a half]

1 hour ago, JohnSwenson said:

I have had this happen with regularity in my house with different equipment. Some endpoints always get the same IP address when the DHCP server renews the lease but others seem to frequently get a new IP address which can wreak havoc with network music systems. The weird thing is that after doing this a bunch of times the all of the sudden it will then settle down and get the same address on every  renewal. I never did figure this behavior out.

 

But I don't think this is the issue here since it seems the endpoint is saying it is not getting ANY address. I have also had that happen and that is almost always a fault of the DHCP server itself. On MANY routers the DHCP server has this weird tendency to get slower and slower  over time and eventually will not respond fast enough to DHCP requests. Adding switches to a path (such as the ER) can slow the path down enough that responses from the DHCP server to not make it to the endpoint soon enough and they time out.  The usual fix for this is to reboot whatever device is your DHCP server, which will then start working quickly and again, over time, start having problems.

 

John S.

Just weird unexpected behaviour, send me nuts. Like for the jcat net Femto card essentially uses two nics, each one has its own MAC.

 

Wired : Router > Nic 1, Nic 2 > ER > Lumin

 

I’ve bridged the Nic 1 & Nic 2 connection in Windows and the router assigns the IP address to the ‘bridged’ Nic MAC address which is usually Nic 1.  Nic 2 is not assigned and the Lumin is setup for DHCP, so it works.
 

 Now for some reason, on a reboot, Windows assigns the other nic’s MAC and the Lumin doesn’t connect cause it’s not getting the IP address although it remains connected. Lumin then displays network error. Even on a reboot of the server won’t fix, have to find the MAC address that’s active for the bridge , rehashing the router and off it goes.

[One and a half]

11 hours ago, R1200CL said:

@One and a half

Wouldn't assign IP address in your router (DHCP) solve what you’re talking about ?

 

I guess there is some idea behind bridging ? Can you explain ?

 

Yes, this is normal practice now, even before the EtherRegen was installed. All devices are DHCP with the router assigning fixed IP addresses according to MAC Address. The one exception is the SACD ripper BD, where the address is assigned locally.

 

The reason for using bridging is here,, related to the first post: A novel way to massively improve the SQ of computer audio streaming. Bridging avoids the connection to another switch (this eliminating the crap that's picked up along the way there), and maintains the integrity of the JCAT Net Femto card.

[One and a half]

12 hours ago, roman410 said:

 

Try disable and delete bridge connection in Windows, and make new bridge connection. This was fix my issue connection between Logitech media server and mysqueezebox.com to load Deezer app.

Also try take off EtherRegen from chain, because if I remember correctly, Alex mentioned somewhere, for some reasons Ether Regen do not like bridged connections.

Thanks for the tip! So far the Lumin always connects with the ER, maybe an update fixed the Windows swapping the MAC address. If it plays up, will follow your tip.

[One and a half]

21 minutes ago, mitch751 said:

I am going to order another ER to pair with the existing one, how should I connect these two ERs, via SFP or RJ45 (B>A sides)?

Depends where the existing one is, how the network is structured and where the files are stored. A little drawing or a workflow would help heaps. 😀

[One and a half]

I really like Teflon (PTFE) since the only wire that can handle that heat when extruded is silver plated copper. SPC doesn't corrode, heck the other day I used some 12 gauge I bought in the 80's and the conductors are as shiny today as they were then. PTFE has dielectric better than air.  For hobbyist, McMaster Carr supplies cut to length per foot, for mains input cable I used this type.  

 

For DC, the same cable could be used for a star quad application. PTFE is not exactly flexible, but it's bendable and a small gauge cable can handle a lot of current. the 18AWG can handle 10A at room temperature, so more than enough for DC rails.

 

The downside to PTFE is it needs special stripping tools, or a very sharp knife will also do, but takes longer.

[One and a half]

5 hours ago, Superdad said:

 

Huh?  A vacuum has a Dk of 1.0; Air is something like 1.006; Standard PTFE is 2.1.

Duh, was thinking of a table in Wikipedia that compared the electrical resistance properties of different materials. Air has a resistance of 10^9 to 10^15 Ohms, whereas Teflon has 10^23 to 10^25 ohms resistance, so concluded the dielectric would be 'better' . Never mind, Teflon is pretty tough material and for wires just about the ideal. For the voltages experienced in hi-fi systems, Teflon wouldn't allow any breakdown or leaching, dare I say would not require burn in either, cause the capacitance wouldn't change. That can't be said though for Furutech AG power cable, the other day used it to distribute 9V to the EtherRegen and MicroUSB3, the copper braid actually stuck on the insulation and was a bit of an issue to remove the braid to peel it back. That cable was only a couple of years old since I bought it new. I've also read of Monster Cable speaker wire disintegrating just by being stored on the drum.

 

On the same table in Wikipedia, Silver is listed as the lowest resistance material, followed by copper. Wonder what a combo of SPC is? I gave up on looking after reading this article on differing copper properties and which copper would be used for the Teflon cable manufacturing.

 

 

Quote

 

 

I switched to heat stripping for Teflon long ago. Almost the only way to avoid accidentally nicking or cutting fine strands. Using this tool:

OOOOO, nice ! Is it cheaper than one of these?