ISO Regen performance Improvement Cheap!

[sligolad]

Been checking here for several weeks now and I am really surprised no one has commented on the single most bang for buck upgrade you can add to your system when using the ISO Regen.

In fact I think it would be a perfect add on that should be offered with all ISO Regens at point of sale.

 

If anyone has followed Rob over at US Audio Mart (seen you there often Alex) you will be aware of the discussions on the serious uplift in sound quality you get with just adding about $10 of cable between the power supply and the IR.

Its basically a quick and easy cable snap in and then just sit back and be blown away by the greater separation, low frequency dynamics and overall amazing depth to the music never heard before, i listen to Red Book files...its all there!

 

For those looking for an explanation on how adding cheap cables into the power in on the IR could make such a difference I have to say this has me dumbfounded and I was a complete skeptic, the low cost and easy install was the only reason i tried it.

 

Its basically adding a CAT7 cable with passive POE ends to carry the DC between power supply and the IR, when you connect up the 2 pieces just stick the power supply in the female jack on one end and stick the male jack on the other end into the IR and sit back and be amazed.

 

Some links below:

POE Ends

CAT7

[genjamon]

Wouldn't star-quad DC wire do the same thing?  I have the Canare star-quad wire recommended by John Swenson in past threads and found it to improve upon high end neotech silver/gold solid core wire with oyaide connectors in my system.  Are you claiming that this tweak improves on the star-quad?

[rah50]

I think I get it, but how about a picture of this to be sure?

[RamDawg]

I definitely want to try as well.  Your last description confused me as to the setup. I have an LPS 1 powering my IR. IR has female end, LPS 1 has female end.  The Poe ends on your link- one female one male.  Should I buy 2 sets to get 2 males- 1 to connect to the IR and 1 to connect to the LPS 1 with the cats in between?  Please clarify kind sir!  Thank you. 

[sligolad]

OK posted some pics to help.

 

The set shown uses a 2 metre CAT7 cable which is the flat cable in the link.

I also have a round Ibra CAT7 but the IR was having difficulty seeing USB signal so maybe some may struggle with the round cable though some have had success with other CAT7 cables.

A POE set comes with male and female ends so if you have a female to female power connector already t5hen you will need to order 2 POE sets.

I ordered a 5 set as they are dirt cheap and already use 3 POE sets in my system, the other 2 on the DC feeding both Startech Lex & Rex.

 

Lots of good discussions here for those interested in the last 10 pages or so:

POE from the discoverer

 

 

[JohnSwenson]

As far as I can tell there are two possible things going on with this configuration:

 

1) sort of star quad

 

2) transformers in the path

 

#1:  these PoE adapters inject DC into the the center tap of Ethernet transceivers, the result is an almost star quad arrangement of the wires inside the Ethernet cable. The result of this is much lower inductance than a usual DC cable. See the DIY DC cable thread for details on using star quad cables for DC power.

 

The description specifically states this works for CAT7, which has shielding, has someone tried this with a same length CAT6 and found it only works for CAT7?  I'm not sure what these adapters DO with the shield connections, can someone do a continuity check of the setup above and see if the cable shield connect to the DC negative on the DC connectors?

 

#2 the DC power is going through the Ethernet transformers, this may offer some high frequency filtering, but it may also increase impedance, it is hard to tell without actually measuring.

 

I would like if a couple people that are already using a star quad cable DC cable could compare that with this configuration.

 

Thanks,

 

John S.

[sligolad]

Thanks for the thoughts John.

 

Could be this particular implementation of Star Quad and wire gauge alright so would also be interested to hear feedback from someone with a good home made star quad compared to the plug and play option presented here.

 

The passive POE setup has no transformers involved so the 2 pairs in the network cable go direct to the positive and negative on the DC ends.

Works with all network cables i tried from CAT5 up but benefits improve as you go higher, CAT7 was best but someone suggests after tests that CAT8 (hard to find a cheap 1 metre) is even better so I have one on order.

 

There is no continuity on any of the cables i have between the shielding and the DC negative on the POE ends.

[JohnSwenson]

3 hours ago, sligolad said:

Thanks for the thoughts John.

 

Could be this particular implementation of Star Quad and wire gauge alright so would also be interested to hear feedback from someone with a good home made star quad compared to the plug and play option presented here.

 

The passive POE setup has no transformers involved so the 2 pairs in the network cable go direct to the positive and negative on the DC ends.

Works with all network cables i tried from CAT5 up but benefits improve as you go higher, CAT7 was best but someone suggests after tests that CAT8 (hard to find a cheap 1 metre) is even better so I have one on order.

 

There is no continuity on any of the cables i have between the shielding and the DC negative on the POE ends.

The only way to do a passive POE without transformers is two pairs for data and two pairs for power, thus limiting the Ethernet part to 100Mb. I know you are not using the Ethernet part, but is it rated at only up to 100Mb?

 

If there really are no transformers involved then this really is just getting something similar to star quad cable.

 

On the cable front, some cables use 24AWG and some use 23AWG wires. It MAY be that the 23AWG ones sound better. The higher CAT numbers have tighter tolerances on the wire geometry, I'm not sure this could have anything to do with a DC cable though.

 

John S.

[genjamon]

I sprung for the $10 to buy the POE adapters and have an extra 1M Supra CAT8 cable I can use to connect them.  Will let you guys know what I find compared to my 6" DIY Canare star quad between LPS-2 and Ultrarendu.  

[charlesphoto]

Tried this on a lark betweeen micro and LPS-1. Maybe a slight improvement, some smoothing, and could tell subtle differences between cables. Seemed to suck a bit of the dynamics from the music though compared to my Ghent Audio dc starquad with 45 degree Oyaide connectors. Maybe too many connections with the poE? Pretty subtle stuff - probably chasing diminishing returns, at least in my setup. I have an lt3045 coming - maybe it will work better on one leg of that, and the Gent on the other? 

[Quadman]

Big thank you to @sligolad who made me aware of the thread over at US Audiomart, I guess I spend to much time listening to music and not enough time reading forums, tho 224 page threads are a bit intimidating to peruse.  It never ceases to amaze me the paths that computer based audiophiles will travel down in the name of improving the audio stream from a PC.  Hats off to you all.

 

Well my uGreen 1m cat 7 cables and passive PoE adaptors showed up yesterday, despite reading very favorable comments from folks who use this setup to power devices I was skeptical, how could adding more connections in the power path be a good thing.  I was using shielded (JSSG mod) starquad power cables connected directly to PSU at one end and 2.1x5.5mm barrel connector at device end, now I was removing the direct connection at LPSU end installing a female 2.1mm connector and plugging a PoE adaptor into that, a ugreen cat 7 cable into the ethernet part at both ends of PoE's and the barrel of the second PoE into a Isoregen located at the dac (T+A dac 8 DSD) and a regen located at the PC.  So I had now replaced 2 shielded direct connected (at LPSU) starquad cables with this mess of cat 7 and PoE adaptors.  It took several on/off, plug/unplug cycles for the PC to see the Dac (that IR is a fussy piece) and I set a playlist to play with the amps turned off my goal to get 4-5 hours of burn in before sitting to listen.

 

Well it was now time to sit and listen and listen I did, it was a 6 hour session and for most of it I was playing music from my most played albums on Roon, the first 3 pages of albums, albums I know very well as I listen an average of 100 hours a month, not background listening, but sit and focus on music listening.  Dumbfounded is all I can say this was a pretty big change, a very positive change.  Wider, deeper soundstage with much more energy inside it.  Backgrounds just disappeared and you are left with just the performers, more solid than before even at the edges of the sound stage, their presence was just more than anytime before.  Leading edges of transients where fast and very well defined giving a nice dynamic jump to the sound from pre PoE.  I detected no edges of digital glare at all to the sound 6 hours in I still wanted to listen, but the clock told me to go to bed (it was 2:30am).  Overall the sound was more like I was at he performance than the performance playing on my stereo.  It had more of a human quality to it, more dynamic, more transparent, more speed.  Very pleased with this PoE chain and I have no idea why it does what it does.  If you read the thread over at Audiomart start around page 204 and you'll get 98% of this PoE experiment.  My next move is to direct connect the PoE adaptor to the LPSU's and add another one to power my SLC OS SSD inside my PC.  I will also add some in series LT3045 regulators to step the voltage down from my LPSU from 7V to 5V and I will power the SSD, IR and regen at this voltage, that I expect will be another jump up based on comments of those who have done the same.

[sligolad]

Great feedback Quadman and perfectly details the jump in listening pleasure I experienced.

It is all a bit baffling but no doubt about what you will hear as a result.

 

From my end I managed to get the IR to work with the 1 metre IBRA round Cat7 with double foil shielding. Can be fussy but once it settles in it works reliably from then on.

 

I also soldered a flying lead to the gold shield on one end of the CAT7 cable and ran this to an earth i have from a spike in the garden, this to bleed off any energy that may be pulled out...no idea if this is what is happening but with 3 of these POE sets in my system the music is incredible.

[Quadman]

Interesting the shielding solved the ibra issue you reported in AM thread.  I may have to try that but once I finally got the pc and dac to see each other the IR was stable all evening and I had no further issues with playback.  This was one of those R U F'ing kidding me moments in SQ jump.  Digital from a PC and ways to improve, many times, just don't seem to make sense but the results speak for themselves.  Well time to crawl back down the rabbit hole.

[lmitche]

On 5/10/2018 at 3:23 PM, JohnSwenson said:

The only way to do a passive POE without transformers is two pairs for data and two pairs for power, thus limiting the Ethernet part to 100Mb. I know you are not using the Ethernet part, but is it rated at only up to 100Mb?

 

If there really are no transformers involved then this really is just getting something similar to star quad cable.

 

On the cable front, some cables use 24AWG and some use 23AWG wires. It MAY be that the 23AWG ones sound better. The higher CAT numbers have tighter tolerances on the wire geometry, I'm not sure this could have anything to do with a DC cable though.

 

John S.

John, yes these Poe adapters support 100 mbps only. That is because two of the four pairs are used for power.

 

Here is a diagram that has been confirmed with my Multimeter.

 

 

My guess is that the benefit comes from the Ibra Ethernet cable construction which is rather like the JSSG design with dual layers of foil shielding.

 

The Poe splitters don't look large enough to contain a transformer. Perhaps we should cut one apart?

 

Larry

[sligolad]

Already have cut up the POE connector and nothing in there Larry.

Just direct connection to the pins.

[Quadman]

I cut one open this afternoon and was shocked on the power lead wiring, which leads me to believe that we should just buy RJ45 female, perhaps shielded and solder our own power leads to it.  Either shielded or not then connect other ends of leads directly to LPSU's.  The existing power lead to PoE is to the right.  No transformers.  This wiring does confirm your multimeter readings tho.

 

[lmitche]

Way cool! Many thanks for the picture. It’s just like I suspected. Yes indeed I’ll cut open the Ibra cable in the am and solder some connectors at each end. I am also curious to learn how the shielding is terminated.

[AnotherSpin]

Very interesting findings and discussion. Will this give any improvement with Regen first generation (amber Regen)? Thank you.

[seeteeyou]

FYI - I posted this a month ago and maybe we could make something even simpler on our own?

 

https://szxlydz.taobao.com

https://item.taobao.com/item.htm?id=553920538613

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?page=304&tab=comments#comment-807018

 

[lmitche]

Ok so an IBRA cable was dissected. Here is a picture:

 

 

The foil around each twisted pair is encased a stranded metal sheathing. Using the multimeter, continuity is maintained throughout the foil and metal sheathing to the gold tabs on the side of the Ethernet plugs.

 

So now we know that the mechanism delivering a SQ increase here is largely the cable shielding, not the POE splitters.

[Theobetley]

Just so I understand: you jury rig a standard power supply connector to a POE (std to ethernet convertor) and on the other end do the opposite and lo and behold you  have a high performance power supply feed. Is this correct?

[lmitche]

20 minutes ago, Theobetley said:

Just so I understand: you jury rig a standard power supply connector to a POE (std to ethernet convertor) and on the other end do the opposite and lo and behold you  have a high performance power supply feed. Is this correct?

What do you mean by jury rig and "lo and behold"?

[Theobetley]

I am not degrading the tweak just trying to understand. Once I understand I will try on all my power supplies. I guess jury rig does sound pejorative but I meant once you hook up the power supply feed to route through an ethernet cable (via the POE's) you get clean power at the other end.

[lmitche]

15 minutes ago, Theobetley said:

I am not degrading the tweak just trying to understand. Once I understand I will try on all my power supplies. I guess jury rig does sound pejorative but I meant once you hook up the power supply feed to route through an ethernet cable (via the POE's) you get clean power at the other end.

Based on the POE splitter breakdown pictures above, the splitter delivers zero SQ benefit and just connects the Ethernet cable to the DC plugs. For someone without DIY skills, this is a good and cheap way to obtain the benefits of shielded DC cabling.

 

Nevertheless, this also means that in theory one can make a DC cable just by adding DC plugs and connectors directly to the Ibra shielded cable with the Ethernet plugs removed in a starquad, or double star quad, type of way.

 

I'm not getting the "wow" impact to SQ here, but I am already using heavily shielded cabling already, and wouldn't expect to hear an improvement but I certainly don't hear any degradation either.

[lmitche]

We should probably be careful about current limits with the 26 awg Ibra cable and other cat 7 solutions. The POE splitter puts current on 2 26 awg conductors on each side, for an effective awg of 23. That yields a max current of 0.729 amps according to the following table.

https://www.powerstream.com/Wire_Size.htm

That may explain why my Netgear switch is flakey when powered with the Ibra cable and POE splitters. However there is more than enough current for an ISO Regen.