Overall Isolation - network, USB, and power

[austinpop]

Austinpop, reading this, I am lost. What is the current state of your USB chain?

 

Sometime I don't remember myself!

 

But my current chain is as shown in the signature. For USB specifically, it is:

• Auralic Aries Mini <- Curious 0.2m USB -> Intona Industrial <- Vbus2 -> <- Curious 0.2m USB -> W4S Recovery <- Cardas Clear USB -> Ayre Codex
  

[MrUnderhill]

Thank you to all for this thread, as well as Andrew Everard and articles on Audiostream.

 

I bought a pair of TP-Link MC110CS, a pair of SC cables (as short as I could find) and a 9V PSU. I already have one 9V IFI that I was using on the switch I have in my audio setup, I swapped this out with the original TP-Link PSU; but, as the switch is on the noisy side of the media bridge I was hoping any noise would have a minimal influence.

 

 

Synology NAS (CAT5e) > Cisco Switch (CAT5e) > Audio Switch (SMPS) > MC110CS (CAT5e - SMPS) > SC-SC > MC110CS (AQ Cinnemon - IFI LPSU) > mR (Van del Hul AES - Sbooster LPSU) > SBooster 5v blocker > Audio Breeze DU-U8 Talema (Van del Hul AES) > Mutec MC-3+USB > Bel Canto DAC3.5vb

 

 

Adding The Bridge

 

Mary Black - Columbus - Live at Olympia: the double bass was always there but as a large presence, it has now coalesced into a musical instrument. In fact I am able to follow the bass line with increased ease.

Neil Diamond - Melody Road - Melody Road (Quboz): again the bass is easier to follow and has a increased dynamic as the player plucks the strings. Edge reduced without impacting clarity.

 

In general I find that digital edge has disappeared, I do still have some edge but it is recording dependent and not added by my system. My digital music is simply excellent.

 

I have been using a Bel Canto 3.5vb for a few years, but I was somewhat frustrated with it, it wasn't delivering what I was expecting in a couple of areas; now it is. I was thinking of trying other DACs, no longer.

 

A few months ago I bought a pair of Focal 1008be IIs, which are powered by my EAR Yoshino amps. I have enjoyed them, but now realise that on the digital side I was only hearing a truncated part of their abilities. The bridge, apart from allowing detail to shine through, has revealed all sorts of spacial cues, captivating imaging.

Conclusion

 

Well, this is a tremendous gain for me. It is allowing my digital chain to shine. Thank you again.

 

M

[lmitche]

 

Sometime I don't remember myself!

 

But my current chain is as shown in the signature. For USB specifically, it is:

• Auralic Aries Mini <- Curious 0.2m USB -> Intona Industrial <- Vbus2 -> <- Curious 0.2m USB -> W4S Recovery <- Cardas Clear USB -> Ayre Codex
  

Will the Lightspeed 2g with external power work between the PC and Intona? Otherwise the same setup should work in front of the DAC without a VBUS2.

[Superdad]

 

But my current chain is as shown in the signature. For USB specifically, it is:

• Auralic Aries Mini <- Curious 0.2m USB -> Intona Industrial <- Vbus2 -> <- Curious 0.2m USB -> W4S Recovery <- Cardas Clear USB -> Ayre Codex
  

 

Well the next big jump your system is going to take is when you receive the UltraCap LPS-1 you have on order. Use it for your W4S Recovery.

 

Right now you have the Intona providing galvanic isolation from the PC, and the W4S Recovery improving the signal integrity, but any PS powering the Recovery--besides a battery or the LPS-1--is going to somewhat defeat the Intona's GI by reintroducing a leakage current loop.

 

This is likely why in the past I have found the Intona/REGEN combo a mixed bag: The galvanic isolation of the Intona sounds good in my system (hear it in the bass especially), but its signal integrity and impedance match are poor compare to the REGEN. But before I had the LPS-1 I was just powering the REGEN mostly from the Mean Well (my JS-2 was occupied with my Mac mini and DAC; now I have second JS-2 and also a bus-powered DAC as an option), so that somewhat defeated the Intona's isolation.

 

100% breaking the path for leakage currents (especially if someone has a regular computer plugged in) is likely why microRendu users are reporting such tremendous results with the LPS-1. The transformer of the microRendu's Ethernet input provides galvanic isolation, but regular power supplies for the mR rejoin the unit--and whatever DAC the mR feeds--back into the rest of the system's "leakage loop" paths.

 

John and I were talking over the weekend, about this, and about AC isolation transformers. We feel the issue (of leakage loops) is so important--but often hard to visualize, partly because it not the same as simple cable ground "loops"--that he is going to quickly paper sketch some diagrams (which either I or my son will digitally redraw) so that people can get a better handle on this stuff. I think a lot of folks here are getting it, but there is of course a much wider audience for this kind of education information (and yes, it obviously favors our solution too ).

 

Happy Monday all!

 

--Alex C.

[YashN]

John and I were talking over the weekend, about this, and about AC isolation transformers. We feel the issue (of leakage loops) is so important--but often hard to visualize, partly because it not the same as simple cable ground "loops"--that he is going to quickly paper sketch some diagrams (which either I or my son will digitally redraw) so that people can get a better handle on this stuff. I think a lot of folks here are getting it, but there is of course a much wider audience for this kind of education information (and yes, it obviously favors our solution too ).

 

Looking forward to it as I'm still trying to visualise it and combine it with the knowledge from the AC Filtering, Grounding thread.

 

I have a couple of tabs open to read John's writings on it from the LPS-1 and Tech Corner pages.

 

Did I say 'read'?

 

More like 'study'!

[One and a half]

Well the next big jump your system is going to take is when you receive the UltraCap LPS-1 you have on order. Use it for your W4S Recovery.

 

I'm not sure which version austinpop has of the Recovery, the newer models use a 9V supply. The best the LPS-1 can do is 7V, 22% down on nominal a fair ask. Two LPS-1 are needed, say 5V + 3.3v, that's 7% under, should work OK.

[austinpop]

Well the next big jump your system is going to take is when you receive the UltraCap LPS-1 you have on order. Use it for your W4S Recovery.

 

Right now you have the Intona providing galvanic isolation from the PC, and the W4S Recovery improving the signal integrity, but any PS powering the Recovery--besides a battery or the LPS-1--is going to somewhat defeat the Intona's GI by reintroducing a leakage current loop.

 

This is likely why in the past I have found the Intona/REGEN combo a mixed bag: The galvanic isolation of the Intona sounds good in my system (hear it in the bass especially), but its signal integrity and impedance match are poor compare to the REGEN. But before I had the LPS-1 I was just powering the REGEN mostly from the Mean Well (my JS-2 was occupied with my Mac mini and DAC; now I have second JS-2 and also a bus-powered DAC as an option), so that somewhat defeated the Intona's isolation.

 

100% breaking the path for leakage currents (especially if someone has a regular computer plugged in) is likely why microRendu users are reporting such tremendous results with the LPS-1. The transformer of the microRendu's Ethernet input provides galvanic isolation, but regular power supplies for the mR rejoin the unit--and whatever DAC the mR feeds--back into the rest of the system's "leakage loop" paths.

 

John and I were talking over the weekend, about this, and about AC isolation transformers. We feel the issue (of leakage loops) is so important--but often hard to visualize, partly because it not the same as simple cable ground "loops"--that he is going to quickly paper sketch some diagrams (which either I or my son will digitally redraw) so that people can get a better handle on this stuff. I think a lot of folks here are getting it, but there is of course a much wider audience for this kind of education information (and yes, it obviously favors our solution too ).

 

Happy Monday all!

--Alex C.

 

Alex, exactly my thinking! I think the LPS-1 is the key piece to complete the isolation boundary.

 

Of course, I still plan to compare the Baaske isolator to the current FMC-based approach, and some have cautioned about using better battery supplies for the FMCs. But these are implementation choices.

 

Architecturally, I think it's finally clicked in place for me.

 

In fact, here is the latest picture of my envisioned topology. Comments welcome!

 

 

I know I and many others would love to see the diagrams that John and you come up with. Thanks so much for all you do for this community. I can't wait for my LPS-1!

[austinpop]

I'm not sure which version austinpop has of the Recovery, the newer models use a 9V supply. The best the LPS-1 can do is 7V, 22% down on nominal a fair ask. Two LPS-1 are needed, say 5V + 3.3v, that's 7% under, should work OK.

@One and a half

 

I have the latest Rev B, which can operate on 6-9V/1A. In fact, I have had email exchanges with Tony Holt at W4S, who's confirmed 7V/1A is an excellent operating point to run at.

[austinpop]

Conclusion

 

Well, this is a tremendous gain for me. It is allowing my digital chain to shine. Thank you again.

 

M

 

That is awesome news. I'm so glad you're hearing the benefits. I must say there is no end to this tweaking, so at some point I will stop, and just enjoy the music!!

[austinpop]

Will the Lightspeed 2g with external power work between the PC and Intona? Otherwise the same setup should work in front of the DAC without a VBUS2.

 

You're right, I should try these configs. Maybe tonight...

[scan80269]

Well the next big jump your system is going to take is when you receive the UltraCap LPS-1 you have on order. Use it for your W4S Recovery.

 

Right now you have the Intona providing galvanic isolation from the PC, and the W4S Recovery improving the signal integrity, but any PS powering the Recovery--besides a battery or the LPS-1--is going to somewhat defeat the Intona's GI by reintroducing a leakage current loop.

 

This is likely why in the past I have found the Intona/REGEN combo a mixed bag: The galvanic isolation of the Intona sounds good in my system (hear it in the bass especially), but its signal integrity and impedance match are poor compare to the REGEN. But before I had the LPS-1 I was just powering the REGEN mostly from the Mean Well (my JS-2 was occupied with my Mac mini and DAC; now I have second JS-2 and also a bus-powered DAC as an option), so that somewhat defeated the Intona's isolation.

 

100% breaking the path for leakage currents (especially if someone has a regular computer plugged in) is likely why microRendu users are reporting such tremendous results with the LPS-1. The transformer of the microRendu's Ethernet input provides galvanic isolation, but regular power supplies for the mR rejoin the unit--and whatever DAC the mR feeds--back into the rest of the system's "leakage loop" paths.

 

John and I were talking over the weekend, about this, and about AC isolation transformers. We feel the issue (of leakage loops) is so important--but often hard to visualize, partly because it not the same as simple cable ground "loops"--that he is going to quickly paper sketch some diagrams (which either I or my son will digitally redraw) so that people can get a better handle on this stuff. I think a lot of folks here are getting it, but there is of course a much wider audience for this kind of education information (and yes, it obviously favors our solution too ).

 

Happy Monday all!

 

--Alex C.

 

A few years back my colleague and I decided we would give each piece of our audio gear a dedicated 100VA balanced transformer (except for power amp). He ordered 10 custom toroid transformers (from toroid.com IIRC) and installed them into PC power supply cases we salvaged from failed PSUs. There was an immediate sonic improvement. More recently, he is using a Furman power conditioner with PFC for the "low power" audio gear (renderer, DAC, digital crossover, etc.) and a Furman balanced transformer for his pair of Benchmark AHB2 power amps (for bi-amp setup). You can say we've been firm believers of "leakage loop" paths degrading the sound.

 

Thanks for confirming the RUR rev B can be powered by LPS-1 at 7V. I'm going to try it.

[One and a half]

@One and a half

 

I have the latest Rev B, which can operate on 6-9V/1A. In fact, I have had email exchanges with Tony Holt at W4S, who's confirmed 7V/1A is an excellent operating point to run at.

 

That's excellent news, saving another lps-1 is also healthier wallet.

[austinpop]

That's excellent news, saving another lps-1 is also healthier wallet.

 

Tastes great, AND less filling!

[austinpop]

Thanks for confirming the RUR rev B can be powered by LPS-1 at 7V. I'm going to try it.

 

That's excellent news, saving another lps-1 is also healthier wallet.

 

Just to confirm I'm not hallucinating, here is a question to that effect:

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/wyred-4-sounds-recovery-universal-serial-bus-industry-standard-cables-connectors-and-communications-protocols-between-computers-and-electronic-devices-reclocker-findings-audio-performance-curated-thread-27140/index20.html#post586858

 

And here is the official post from W4S regarding the rev B PSU requirements:

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/wyred-4-sounds-recovery-universal-serial-bus-industry-standard-cables-connectors-and-communications-protocols-between-computers-and-electronic-devices-reclocker-findings-audio-performance-curated-thread-27140/index20.html#post587065

[scan80269]

I just tried feeding a W4S Recovery with an LPS-1 set to 7V, and it works great!

 

My bedroom audio setup is starting to look more like austinpop's, except my Aries Mini is currently set to streaming from high 5GHz wireless band, and the RUR USB output feeds a set of KEF X300A bookshelfs.

[austinpop]

I just tried feeding a W4S Recovery with an LPS-1 set to 7V, and it works great!

 

Awesome!! Thanks for confirming.

 

My bedroom audio setup is starting to look more like austinpop's, except my Aries Mini is currently set to streaming from high 5GHz wireless band, and the RUR USB output feeds a set of KEF X300A bookshelfs.

 

I hate to say this, but I really did find a significant improvement going wired, especially with the fiber isolation. I really wish I understood why this is the case, when a strong 802.11ac signal should be the cleaner option, galvanically.

 

The only way you'll know if it's worthwhile is to try it. However, it really depends how resolving your system is. And, whether getting a wired ethernet to your Mini is even feasible.

[scan80269]

Awesome!! Thanks for confirming.

 

 

 

I hate to say this, but I really did find a significant improvement going wired, especially with the fiber isolation. I really wish I understood why this is the case, when a strong 802.11ac signal should be the cleaner option, galvanically.

 

The only way you'll know if it's worthwhile is to try it. However, it really depends how resolving your system is. And, whether getting a wired ethernet to your Mini is even feasible.

 

Switching the Mini to wired Ethernet is easily done, so that's what I'll try next. I can even throw in the EN-70HD isolator.

 

However, my bedroom is not the best venue for serious listening of sonic differences of audio tweaks. My house is close to a main street so there is often quite a bit of traffic noise. And while I enjoy my KEF X300A speakers immensely, I'm not sure the built-in DAC is resolving enough to support these experiments. That is why I've been scheduling sessions with my retired colleague, who has a very high resolution system ideal for doing A/B comparisons.

 

Back on the Mini, I sometimes wonder whether using wired or wireless networking results in lower overall electrical noise on its main board. When wired LAN is active the wireless is supposed to be quiescent, and vice versa. Without knowing the quiescent & active power of wired and wireless subsystems, it is hard to predict which one will gain the upper hand in SQ. Also, things like how hard the data stream recovery logic has to work to retrieve the audio bitstream will impact the board noise level. For wireless the signal strength and airwave congestion will matter. For wired, I suspect the length and quality of the Ethernet cable can be factors.

 

I'll propose to my colleague to do another round of wired vs. wireless comparison for his Aries Femto. He has settled on wireless over dedicated high 5GHz band, but previously the wired setup had no tweaks so it's worth another shot. He told me recently that a wireless router change or even just a router configuration change (e.g. from router mode to AP mode) can affect the sound for streaming. There is clearly much more awaiting experimentation and tweaking than just the audio gear.

[austinpop]

 

I'll propose to my colleague to do another round of wired vs. wireless comparison for his Aries Femto. He has settled on wireless over dedicated high 5GHz band, but previously the wired setup had no tweaks so it's worth another shot. He told me recently that a wireless router change or even just a router configuration change (e.g. from router mode to AP mode) can affect the sound for streaming. There is clearly much more awaiting experimentation and tweaking than just the audio gear.

 

I look forward to your findings. Please report back!

 

 

 

Sent from my iPhone using Computer Audiophile

[YashN]

I hate to say this, but I really did find a significant improvement going wired, especially with the fiber isolation. I really wish I understood why this is the case, when a strong 802.11ac signal should be the cleaner option, galvanically.

 

You have to take into account the nose profile of the digital wireless circuitry too.

[austinpop]

You're right, I should try these configs. Maybe tonight...

 

So I tried 3 configs tonight with the Lightspeed (LH) 2G split USB cable and the Tecknet 5V battery.

 

The Tecknet claims to put out "Output: Each Port - 5V/3.0A max , 3.4A in Total." Certainly it has no trouble powering the TP-Link FMC. However, as you'll see below, I am really beginning to wonder what its current capabilities truly are, and how poor its regulation must be.

 

USB configurations

 

Baseline:

• Chain: Aries Mini > Curious 0.2m USB -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
  
• Result: Works fine.
  

Experiment 1: Use split cable and Tecknet 5V to power Intona

• Chain:
  
  • Aries Mini > Vbus2 > LH 2G USB (data) -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
    
  • Tecknet 5V battery > LH 2G USB (power) > Intona
    

   

  [*]Result: Intona light blinks, but does not sync. Turning off battery removes power, so clearly the connection is correct, but Intona won't establish sync. My guess is the battery voltage is sagging under load.

Experiment 2: Same as Experiment 1, but no Vbus2 on Aries Mini output

• Chain:
  
  • Aries Mini > LH 2G USB (data) -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
    
  • Tecknet 5V battery > LH 2G USB (power) > Intona
    

   

  [*]Result: Same as experiment 1: Intona light blinks, but does not sync.

Experiment 3: Use split cable and battery to send 5V to the DAC, but no Vbus2

• Chain:
  
  • Aries Mini > Curious 0.2m USB > Intona Industrial > Curious 0.2m USB > W4S Recovery > Lightspeed 2g USB (data) > Ayre Codex
    
  • Tecknet 5V battery > Lightspeed 2g USB (power) > Ayre Codex
    

   

  [*]Result: Same as before. Codex does not connect to Aries Mini

 

I have a Teralink X1/X2 9V/5V/1A LPS that has been in transit for a month from Hong Kong - probably stuck in customs. If/When it gets here, I'll use its 5V output to try this again.

 

Not a ringing endorsement of the Tecknet, is it!

[lmitche]

You have to take into account the nose profile of the digital wireless circuitry too.

Agreed, a Roman nose profile is best!?

[YashN]

Agreed, a Roman nose profile is best!

 

[lmitche]

So I tried 3 configs tonight with the Lightspeed (LH) 2G split USB cable and the Tecknet 5V battery.

 

The Tecknet claims to put out "Output: Each Port - 5V/3.0A max , 3.4A in Total." Certainly it has no trouble powering the TP-Link FMC. However, as you'll see below, I am really beginning to wonder what its current capabilities truly are, and how poor its regulation must be.

 

USB configurations

 

Baseline:

• Chain: Aries Mini > Curious 0.2m USB -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
  
• Result: Works fine.
  

Experiment 1: Use split cable and Tecknet 5V to power Intona

• Chain:
  
  • Aries Mini > Vbus2 > LH 2G USB (data) -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
    
  • Tecknet 5V battery > LH 2G USB (power) > Intona
    

   

  [*]Result: Intona light blinks, but does not sync. Turning off battery removes power, so clearly the connection is correct, but Intona won't establish sync. My guess is the battery voltage is sagging under load.

Experiment 2: Same as Experiment 1, but no Vbus2 on Aries Mini output

• Chain:
  
  • Aries Mini > LH 2G USB (data) -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
    
  • Tecknet 5V battery > LH 2G USB (power) > Intona
    

   

  [*]Result: Same as experiment 1: Intona light blinks, but does not sync.

Experiment 3: Use split cable and battery to send 5V to the DAC, but no Vbus2

• Chain:
  
  • Aries Mini > Curious 0.2m USB > Intona Industrial > Curious 0.2m USB > W4S Recovery > Lightspeed 2g USB (data) > Ayre Codex
    
  • Tecknet 5V battery > Lightspeed 2g USB (power) > Ayre Codex
    

   

  [*]Result: Same as before. Codex does not connect to Aries Mini

 

I have a Teralink X1/X2 9V/5V/1A LPS that has been in transit for a month from Hong Kong - probably stuck in customs. If/When it gets here, I'll use its 5V output to try this again.

 

Not a ringing endorsement of the Tecknet, is it!

Yeh, something isn't right with that battery. Do you have a DMM to measure the voltage drop when something is drawing power?

[Cornan]

So I tried 3 configs tonight with the Lightspeed (LH) 2G split USB cable and the Tecknet 5V battery.

 

The Tecknet claims to put out "Output: Each Port - 5V/3.0A max , 3.4A in Total." Certainly it has no trouble powering the TP-Link FMC. However, as you'll see below, I am really beginning to wonder what its current capabilities truly are, and how poor its regulation must be.

 

USB configurations

 

Baseline:

• Chain: Aries Mini > Curious 0.2m USB -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
  
• Result: Works fine.
  

Experiment 1: Use split cable and Tecknet 5V to power Intona

• Chain:
  
  • Aries Mini > Vbus2 > LH 2G USB (data) -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
    
  • Tecknet 5V battery > LH 2G USB (power) > Intona
    

   

  [*]Result: Intona light blinks, but does not sync. Turning off battery removes power, so clearly the connection is correct, but Intona won't establish sync. My guess is the battery voltage is sagging under load.

Experiment 2: Same as Experiment 1, but no Vbus2 on Aries Mini output

• Chain:
  
  • Aries Mini > LH 2G USB (data) -> Intona Industrial > Vbus2 > Curious 0.2m USB > W4S Recovery > Cardas Clear USB > Ayre Codex
    
  • Tecknet 5V battery > LH 2G USB (power) > Intona
    

   

  [*]Result: Same as experiment 1: Intona light blinks, but does not sync.

Experiment 3: Use split cable and battery to send 5V to the DAC, but no Vbus2

• Chain:
  
  • Aries Mini > Curious 0.2m USB > Intona Industrial > Curious 0.2m USB > W4S Recovery > Lightspeed 2g USB (data) > Ayre Codex
    
  • Tecknet 5V battery > Lightspeed 2g USB (power) > Ayre Codex
    

   

  [*]Result: Same as before. Codex does not connect to Aries Mini

 

I have a Teralink X1/X2 9V/5V/1A LPS that has been in transit for a month from Hong Kong - probably stuck in customs. If/When it gets here, I'll use its 5V output to try this again.

 

Not a ringing endorsement of the Tecknet, is it!

 

Are you 100% sure that it is not something wrong with your Lightspeed G2 cable...or that you are connecting the correct leg (black) to power?

Some DACs is more "power hungry" than others...but 5v/3A (3,4A in total) should really be enough to make contact. I would actually check that Lightspeed 2G cable. Check if you get a connection with this setup: Aries Mini>Lightspeed 2G (black=power)>Tecknet & Lightspeed 2G (data)>Ayre Codex. If you have a charge/DC switch on that Teknet Powerbank make sure it is set correctly (ie. to DC/output or similar). Often problems like this are due to cables IME.

[Cornan]

@austinpop

I think you could also check the GND connection on the Lightspeed 2G using a probe (continue check). Since the Intona blinks it feels like it is waiting for the handshake which is made through the GND.

Side note. Checking this video

makes me wonder if Intona allows to remove the power leg when the handshake have been done? Could be worth checking out anyway!