The isolation thread

[R1200CL]

@agladstone

 

A small update. 

You can use spare rail JS-2 to power LPS-1.2 😀

 

[James Stephens]

6 hours ago, JohnSwenson said:

Just a word of clarification: the shield tied/not tied Ethernet cable ONLY affects true ground loops. High impedance leakage current will go through ANY Ethernet cable (except one that is not plugged in!)

 

Quick refresher:

 

Ground loop: caused by voltage difference (50/60Hz) between safety grounds on different outlets.

 

Leakage loop: current flowing from mains AC (NOT safety ground), through power supply to DC output. Can travel to other boxes and either go through leakage in another power supply or get shunted through a safety ground connection.

 

I've written extensively about this in other posts.

 

John S.

Hi John,

 

Thanks very much for correcting me on this.

 

Best,

James

[agladstone]

12 hours ago, R1200CL said:

@agladstone

 

A small update. 

You can use spare rail JS-2 to power LPS-1.2 😀

 

This is perfect now!! Thanks sooo much! 
I just got the EtherRegen and will get it in place later today (I’m going to change up my stereo rack to try and fit all this stuff in it! All these smaller pieces of gear and cables and power supplies take up a lot of rack real estate!). 
I’m also planning on adding a wire to go from AC Wall Safety Ground to the negative of the DC output of the DC4EVER - seems like that is what is suggested for network switches power supplies (JSGT). Should I do the same also for the Power supply for the Netgear router/ switch? I am powering that with an HDPlex LPSU? 

[agladstone]

13 hours ago, JohnSwenson said:

Just a word of clarification: the shield tied/not tied Ethernet cable ONLY affects true ground loops. High impedance leakage current will go through ANY Ethernet cable (except one that is not plugged in!)

 

Quick refresher:

 

Ground loop: caused by voltage difference (50/60Hz) between safety grounds on different outlets.

 

Leakage loop: current flowing from mains AC (NOT safety ground), through power supply to DC output. Can travel to other boxes and either go through leakage in another power supply or get shunted through a safety ground connection.

 

I've written extensively about this in other posts.

 

John S.

@JohnSwenson Connecting AC Wall Safety Ground to the negative/ Shell of the DC Power Supply (JGST) for network switch is what helps this correct? 

[R1200CL]

25 minutes ago, agladstone said:

I am powering that with an HDPlex LPSU? 

Which version ?

 

Quote

Connecting AC Wall Safety Ground to the negative/ Shell of the DC Power Supply (JGST) for network switch is what helps this correct? 

 

Only if you HDPlex is lacking safety ground connection, which I doubt. That’s my understanding. 

 

Remenber the JSSG started with the discovery of a Meanwell SMPS that didn’t have ground connection. There where two was to solve this. Open the SMPS (which I did) and add a strap from “third pin” to DC minus out, or do an external wire from DC minus to your socket outlet safety ground. 
 

Also the HDPlex isn’t a SMPS 😀

[R1200CL]

@PYP

 

Here is your system first sketch. I made it before I saw your last post link beneath. 
 I think you’re only using the clock on last EtherRegen. 

https://audiophilestyle.com/forums/topic/57519-uptone-audio-etherregen-listening-impressions/?do=findComment&comment=1117574

 

So basically you’re probably don’t braking any isolation. 
But you had some questions?

 

 

Maybe turning the first EtherRegen in reverse B>A, and then powering it with JS-2 ?

 

 

 

[PYP]

46 minutes ago, R1200CL said:

@PYP

 

Here is your system first sketch. I made it before I saw your last post link beneath. 
 I think you’re only using the clock on last EtherRegen. 

https://audiophilestyle.com/forums/topic/57519-uptone-audio-etherregen-listening-impressions/?do=findComment&comment=1117574

 

So basically you’re probably don’t braking any isolation. 
But you had some questions?

 

 

Maybe turning the first EtherRegen in reverse B>A, and then powering it with JS-2 ?

 

 

 

Thank you!  Yes, the clock (75 Ohm) is powered by the JS-2 and the second eR is powered by the LPS 1.2 (first eR uses the smps provided with the eR), so isolation is intact and there is one JS-2 output unused at this point.   The connection between the eR's is copper ethernet.  The only fiber is from oM to first eR.  

 

Because of how the various pieces and cables are placed and how short the clock cable is (0.5 m), turning the second eR around to B > A, and powering with the JS-2, would be a lot of work.  Given how the system sounds, I'm not inspired to do that.   Good "problem" to have, I suppose.  

[R1200CL]

3 minutes ago, PYP said:

Because of how the various pieces and cables are placed and how short the clock cable is (0.5 m), turning the second eR around to B > A, and powering with the JS-2, would be a lot of work.  Given how the system sounds, I'm not inspired to do that.   Good "problem" to have, I suppose.  

My suggestion is turning the first one. Since fiber between them, powering both with JS-2 should be OK. What you think ? 

[PYP]

1 minute ago, R1200CL said:

My suggestion is turning the first one. Since fiber between them, powering both with JS-2 should be OK. What you think ? 

The first eR is A > B in order to accept the fiber from the oM.  At some point, I will try powering the first eR with the unused JS-2 outlet.  That is easy to do, but I will let the current configuration settle in for a few weeks.   

[agladstone]

6 hours ago, R1200CL said:

Which version ?

 

 

Only if you HDPlex is lacking safety ground connection, which I doubt. That’s my understanding. 

 

Remenber the JSSG started with the discovery of a Meanwell SMPS that didn’t have ground connection. There where two was to solve this. Open the SMPS (which I did) and add a strap from “third pin” to DC minus out, or do an external wire from DC minus to your socket outlet safety ground. 
 

Also the HDPlex isn’t a SMPS 😀

I can actually very easily do the same inside of the DC4EVER. The ground from AC wall to case screws to case directly below where the DC - banana terminal is , I can add a short wire from ground lug to DC- lug, seems like this would be easier and less mess and risk vs having a wire go from ground in a screw on wall plug to the outer DC shell (dc-) ? 

[R1200CL]

12 hours ago, agladstone said:

I can add a short wire from ground lug to DC- lug, seems like this would be easier and less mess and risk vs having a wire go from ground in a screw on wall plug to the outer DC shell (dc-) ?

Obviously. And verify by a meter maybe. 
 

Everyone can easily measure if the power is “correct” designed or not. 
Just ohm (beep) between DC minus out and third pin. (Or chassis). Without any AC applied of cause. 

This is not the case for isolated powers. Like JS-2.

Am I correct @Superdad ?

I just measured my Paul Hynes SR4T.

It’s isolated. 
 

So I would expect the Farad to be.

Can anyone measure and confirm ?
 

A simple meter cost nothing. Just as an example:

https://www.amazon.com/Etekcity-Multimeter-MSR-R500-Electronic-Multimeters/dp/B01N9QW620/ref=zg_bs_5011672011_2?_encoding=UTF8&psc=1&refRID=AHZ22QZTYVD642S6J287

 

What about HDPlex powers ?

 

Anyway, didn’t want to turn this tread into discussion about JSGT. The subject is well covered elsewhere. 
Use the search function here at AS. 
 

[R1200CL]

I’m getting convinced the absolutely best way to wire up dual EtherRegens with clocks is as shown in first sketch below. 
 

The oM/oM deluxe can be replaced with any switch. The Cisco 2960G as a good candidate. 
 

Cause of isolation the by fiber, dual output power like JS-2 should work just fine. A Paul Hynes SR4T with special Y cable from Ghent should also be excellent. And so ought the Farad to be.

 

A second clock with separate power may be added to the first EtherRegen. 
 

What you guys think ?

And what should favor a different setup with dual EtherRegens.  Or maybe a different setup should considered be equal ? See second sketch. 
 

I’ve only done last EtherRegen in A>B configuration. Those of us using opticalRendu, won’t find the two sketches useful. I’m not sure if a second EtherRegen upfront will help in that case ? Only possible configuration is A>B>B>A. An opticalModule deluxe should probably be enough upfront. 
Ops. B>A>B>A should work as well. Hm. I make those two options next. Still is dual setup worth the extra money with the opticalRendu. I myself just ordered the opticalModule Deluxe. Should be sufficient. 
If a used 50 ohm EtherRegen ever comes for sale, maybe I will test that as well. 
 

 

 

 

[PYP]

53 minutes ago, R1200CL said:

I’m getting convinced the absolutely best way to wire up dual EtherRegens with clocks is as shown in first sketch below. 
 

The oM/oM deluxe can be replaced with any switch. The Cisco 2960G as a good candidate. 
 

Cause of isolation the by fiber, dual output power like JS-2 should work just fine. A Paul Hynes SR4T with special Y cable from Ghent should also be excellent. And so ought the Farad to be.

 

A second clock with separate power may be added to the first EtherRegen. 
 

What you guys think ?

And what should favor a different setup with dual EtherRegens.  Or maybe a different setup should considered be equal ? See second sketch. 
 

I’ve only done last EtherRegen in A>B configuration. Those of us using opticalRendu, won’t find the two sketches useful. I’m not sure if a second EtherRegen upfront will help in that case ? Only possible configuration is A>B>B>A. An opticalModule deluxe should probably be enough upfront. 
Ops. B>A>B>A should work as well. Hm. I make those two options next. Still is dual setup worth the extra money with the opticalRendu. I myself just ordered the opticalModule Deluxe. Should be sufficient. 
If a used 50 ohm EtherRegen ever comes for sale, maybe I will test that as well. 
 

 

 

 

Those both look good.  The second is very similar to my setup.  My clock is fed by the JS-2.  And, for now, the first eR uses the included SMPS.  Thanks for doing this.  

[James Stephens]

3 hours ago, R1200CL said:

I’m getting convinced the absolutely best way to wire up dual EtherRegens with clocks is as shown in first sketch below. 
 

The oM/oM deluxe can be replaced with any switch. The Cisco 2960G as a good candidate. 
 

Cause of isolation the by fiber, dual output power like JS-2 should work just fine. A Paul Hynes SR4T with special Y cable from Ghent should also be excellent. And so ought the Farad to be.

 

A second clock with separate power may be added to the first EtherRegen. 
 

What you guys think ?

And what should favor a different setup with dual EtherRegens.  Or maybe a different setup should considered be equal ? See second sketch. 
 

I’ve only done last EtherRegen in A>B configuration. Those of us using opticalRendu, won’t find the two sketches useful. I’m not sure if a second EtherRegen upfront will help in that case ? Only possible configuration is A>B>B>A. An opticalModule deluxe should probably be enough upfront. 
Ops. B>A>B>A should work as well. Hm. I make those two options next. Still is dual setup worth the extra money with the opticalRendu. I myself just ordered the opticalModule Deluxe. Should be sufficient. 
If a used 50 ohm EtherRegen ever comes for sale, maybe I will test that as well. 
 

 

 

 

 

Maybe Alex can do a special deal on 4 LPS 1.2's? ..... lol just kidding.

 

I am enjoying looking at these drawings.

 

Best,

James

 

 

 

 

[James Stephens]

5 hours ago, R1200CL said:

Obviously. And verify by a meter maybe. 
 

Everyone can easily measure if the power is “correct” designed or not. 
Just ohm (beep) between DC minus out and third pin. (Or chassis). Without any AC applied of cause. 

This is not the case for isolated powers. Like JS-2.

Am I correct @Superdad ?

I just measured my Paul Hynes SR4T.

It’s isolated. 
 

So I would expect the Farad to be.

Can anyone measure and confirm ?
 

A simple meter cost nothing. Just as an example:

https://www.amazon.com/Etekcity-Multimeter-MSR-R500-Electronic-Multimeters/dp/B01N9QW620/ref=zg_bs_5011672011_2?_encoding=UTF8&psc=1&refRID=AHZ22QZTYVD642S6J287

 

What about HDPlex powers ?

 

Anyway, didn’t want to turn this tread into discussion about JSGT. The subject is well covered elsewhere. 
Use the search function here at AS. 
 

 

 

My BG7TBL is currently powered by a Teddy Pardo (TP) 12/2 and I realized today that I may have an issue with leakage currents making it to my endpoint since the TP is plugged into a power conditioner that has 2 iFi iPower X 12V DC power supplies plugged into it and John write in another post that he measured a lot of leakage currents from those.

 

I unplugged the unit and got my multimeter out and was not able to measure continuity from the DC minus to the safety ground.

 

I quickly made a ground shunt for the clock DC supply with a couple of DC jack plug adapters, green wire and an AC power plug and installed that to act as a ground shunt.

 

A little later I decided I would order an LPS 1.2 to power the clock and use the TP along with a DC Y-Cable to power the devices (Tivo and Roku) currently powered by the iFi DC power supplies so that I can get rid of those and then I won't potentially be sending those leakage currents through the HDMI to my PrePro. Of course that depends on the behavior of the TP and I don't know about the distribution of Type 1 vs Type 2 noise with it. As I understand the TP supplies are neither SMPS or LPS .....

 

Thoughts anyone?

 

Best,

James

[Mouna]

Hi RCL2000 - here I am with my system with a second ER

 

My network setup is Router> LAN  - 2 LAN cables - one from the Router and one from PC ( to manage Roon on a NAS connected to the router )  > ER powered with LPSU Farad 3 7V + emperor double crown clock with LPSU Giesemann + LPSU Farad 3 7V  > Finisair 1310 + single mode cable on both ER's > ER powered with LPS1,2 with emperor double crown clock with LPSU Giesemann ( second output ) > LAN >DNP APL STreamer>DRT>APL DAC MKII -

 

The question where to put the 2 LAN cables...... only possible in A?  or putting one in A one in B?

 

In A is the fiber....Doesn't seem to make sense....

 

Leave it there - hope I could make it clear...

thank you in advance!

 

[richard_crl032]

On 3/11/2021 at 12:42 AM, agladstone said:

I also just got a JS-2 but can’t use second rail for EtherRegen since it’s also powering my DAC. 
I also have a LPS 1.2 and a Vinnie Rossi Pure DC4Ever 12V power supply, which of these would be best for EtherRegen and which best for my ISORegen which will be on B side along with my DAC? 

Hi Agladstone/R12000CL,

 

Just realised that I may be breaking the moat in my new setup with 2nd ERs with dual outputs lps that are not isolated and chance on this topic here. I am having the lps powering my wireless mesh node and the AD double crown clock down the equipment chain between my Antipodes cx server and ex player but it certainly sounds great.

 

Why DAC and ER both being on B sides will break the moat ?

 

Moreover, there must be a server/player in between and does that offer high impedence to prevent this "moat breakdown" ? Else with 2X ERs that I have, there is always some connectivity between A and B in AB-?-AB or are we needing to go AB-?-BA or BA-?-AB ?

 

Sorry for these dumb questions if indeed the case but hope to use the 2x ERs effectively while also wanting to minimise more boxes of lps and $.

 

Btw, I am using stock smps of ERs to avoid this confusing topic due to my lack of knowledge but needing 12v for the wireless mesh and AD double crown clock.

 

Cheers.

 

Richard

[richard_crl032]

On 3/11/2021 at 6:06 AM, R1200CL said:

This is something very quickly drawn and totally unfinished. I need @agladstone input as I lack information of his system. He may not even have an endpoint. 
I’m quite certain the A side gets the power. I shall look again at Alex picture he posted yesterday. 
 

Edit. Yes power is on A side. 

 

 

Just to keep it updated that power is indeed on B side as advised by Alex to avoid more confusion to a confused me. Cheers.

 

 

[richard_crl032]

On 3/11/2021 at 9:46 AM, GryphonGuy said:

 

Don't panic. Just enjoy. As long as you are powering the eR with its own exclusive power supply, there is little chance that you will have "a broken moat" to use your words.  After the first few hours of operation, the sheer excellence of the music created by the clean ethernet signal will make you forget all these nit-picking worries that only concern a few people. I hope you don't become one of the few people.

 

Regards

GG

 

Hi GG,

 

Indeed and I am sticking to stock smps but having 2 units and added clock needing 12v power makes it not as straightforward and hoping for further clarification here.

 

Cheers.

 

Richard

[R1200CL]

@richard_crl032

Can you quickly draw something on a piece of paper ?

A>B>A>B is just fine. Both sitting between the Antipodes server and player ?

[richard_crl032]

1 hour ago, R1200CL said:

@richard_crl032

Can you quickly draw something on a piece of paper ?

A>B>A>B is just fine. Both sitting between the Antipodes server and player ?

Hi R1200CL,

 

TIA .. I hope to avoid more LPS/$ and I am keeping the Uptone 1.2 for my Bel Canto FM1 tuner that wish for the best available.

 

Please find simplified schematics which should be correct and the concern areas in red ... hope it is ok but do not hesitate to query me.

 

Cheers.

 

Richard

 

[R1200CL]

5 minutes ago, richard_crl032 said:

I hope to avoid more LPS/$ and I am keeping the Uptone 1.2 for my Bel Canto FM1 tuner that wish for the best available.

How about use the Uptone for the clock, and use ZeroZone for your FM tuner ?

 

Or just use the SMPS that I assume you still have for the WiFi mesh, and only use the ZeroZone for the clock. 

[richard_crl032]

Hi R1200CL,

 

Thanks for the quick response ...

 

Will opt for 2nd option and just use the many meanwell smps that I have for the wifi mesh then ... local fm stations via Bel Canto broadcast chinese music/programs that I need to keep in best possible sound quality since none in full flac via internet stations.

 

Will have a redundant 12v rail from the zerozone then but indeed confirmed that I am breaking the moat in the ER#1 ? 

 

Frankly, the system sounds fine as it is especially with the recent oladra upgrade of the Antipodes and not sure what improvement is expected while losing lps to the wifi mesh ... I will find out late nite today's audition.

 

Cheers.

[richard_crl032]

52 minutes ago, R1200CL said:

How about use the Uptone for the clock, and use ZeroZone for your FM tuner ?

 

Or just use the SMPS that I assume you still have for the WiFi mesh, and only use the ZeroZone for the clock. 

Hi R1200CL,

 

Should I be using the ifi ipower that I have or the meanwell one that came also with my Bel Canto FM1 ?

 

https://audio “science” review/forum/index.php?threads/mains-leakage-currents.5552/

 

Cheers.

 

Richard

[R1200CL]

13 minutes ago, richard_crl032 said:

Should I be using the ifi ipower that I have or the meanwell one that came also with my Bel Canto FM1 ?

Impossible to say. I guess the iFi. As long as it has required volt and ampere.

Here is a PS you can test in the future;

https://www.aliexpress.com/item/1005002207035576.html?spm=a2g0s.12269583.0.0.2d3855beLQoCWi

I ordered mine with the TALEMA option. 
 

The AD clock has almost no circuitry inside, so I think a good PS is essential, hence why I suggest the LPS-1.2.