A novel way to massively improve the SQ of computer audio streaming

[mozes]

1 hour ago, Gavin1977 said:

Shame Mundorf stopped making these MPower supplies back in 2013... 30A of continuous low noise power.  Could build some seriously high powered servers with these.

 

I have no idea of their specs (ripple, noise floor, transient response etc....).  You will see that it uses a Mundorf HeadCap (I believe) - I was wondering if these could be added to ATX power rails to smooth the current...

 

 

 

 

 

Check out these guys. Made in the US and have LPSs up to 57 amps continuous.

https://www.astroncorp.com/linear-desktop

[Gavin1977]

19 minutes ago, mozes said:

Check out these guys. Made in the US and have LPSs up to 57 amps continuous.

https://www.astroncorp.com/linear-desktop

Anyone used one though?

[mozes]

23 minutes ago, Gavin1977 said:

Anyone used one though?

I use the 4A version to power my router, haven't compared it to other LPSs, but it has been running strong for around 3 yrs. Made a positive impact over the SMPS.

[zettelsm]

On 4/29/2020 at 7:22 AM, drjimwillie said:

Over on the “what’s best” “ Extreme” thread they’re talking about The latest Roon release making a detrimental impact on sound quality.

TA was able to fix this issue with some changes of the system. Do any of the extreme owners know if it was only changes to the Roon operating system? And if they saw what was done, can they share those changes here, please?

 

The changes were not to Roon. The changes were to the Extreme server's operating system. Dr. Bok logged in via TeamViewer and performed the changes, but he was so fast I didn't catch all of what he did. My impression is they were changes to allocation and priority of the operating system.

 

Steve Z

[ASRMichael]

Sorry didn't know where to post. Anyone else have an issue with email notifications from threads? I've had none since Tuesday?

[austinpop]

8 hours ago, ASRMichael said:

Sorry didn't know where to post. Anyone else have an issue with email notifications from threads? I've had none since Tuesday?

Site issues go on this thread below. There does appear to be an issue with some email providers...

 

 

[bodiebill]

On 4/22/2020 at 12:29 PM, bodiebill said:

So I think we agree a really good and powerful LPS for a heavy lifting server is quite a challenge. But what about devices that draw only very modest current (say up to 2A)? I see that some member's use f.i. multiple Paul Hynes LPS'es for this. Very expensive. For my JCAT net card FEMTO and my ISO Regen I use cheap Chinese LPSU's (Studer900, € 40) followed by either a LDOVR or MPAudio box that converts the output to a lower voltage using LT3045's. Or alternatively Chinese battery powered LPSU with 4 LT3045's, good for 5V and 2A.

As a rule, the larger the voltage difference, the lower the ripple/noise, so the better the SQ. Also putting two LT3045 units in series further improves the result (i.e. 9V => 7V => 5V).

To be precise:

Studer900 6V => MPAudio DLS-HPULN => 5V => JCAT net card FEMTO or

Chinese battery powered LPSU => 5V => JCAT net card FEMTO

Studer900 9V => LDOVR DXP-1A5DSC => 7.1V => ISO Regen

See pics.

The results are so good (and I compared quite a few options) that I have not been tempted yet to try expensive LPSU's for this.

Anyone else here tried something similar?

 

This week I received a Keces P8 with dual rails 9V/12V and 18V/19V (both 4A) plus USB 5V (1A). This enabled me to try quite a few permutations, applying the SR4 and the LT3045 boards where they had the most positive effect. For now, for my HQPe server - NAA endpoint setup, after only one day of breaking in the Keces, I have decided on:

• XT Power battery 5V for the server's JCAT net card FEMTO
• Paul Hynes SR4 19V for the MikroTik 305 SFP/RJ45 switch
• Keces P8 19V for the endpoint PC with HDPLex 800W DC-ATX converter
• Keces P8 9V => LDOVR DXP-1A5DSC (LT3045) => 7.1V for the ISO Regen
• iFi Power 9V into MPAudio DLS-HPULN (LT3045) => 7V into Chinese PS (LT3045) => 5V for the Elfidelity USB card

The latter (Elfidelity card) sounds clearly better than direct from the Keces P8 5V USB, probably because of the double (serial) LT3045 noise reduction.

 

As 19V supplier for the endpoint PC the SR4 sounded only very slightly better than the Keces. The SR4 made a clear difference on the MikroTik switch.

 

I also tried powering the endpoint CPU/EPS seperately with the SR4's 12V, but this did not give a clear improvement compared to the Keces powering the motherboard + the CPU via the HDPlex 800W.

 

I have ordered a 6V Chinese LPS to be able to further simplify and increase the possibilities 🙂

 

Fyi, for what it is worth...

[bodiebill]

Not much discussed here, but the € 40 the Elfidelity AXF-100 PRO III Internal USB card for me was a very pleasant surprise.

https://www.audiophonics.fr/en/pc-htpc-modules/elfidelity-axf-100-pro-iii-internal-usb-power-filter-for-pc-p-9056.html

In my setup the resulting jump in sound quality was at least as substantial as that when adding the € 435 JCAT net card FEMTO. Apples and pears of course, but still... YMWV 🙂

Both externally powered with 5VDC LPS.

[bodiebill]

4 hours ago, bodiebill said:

• Keces P8 19V for the endpoint PC with HDPLex 800W DC-ATX converter
• Keces P8 9V => LDOVR DXP-1A5DSC (LT3045) => 7.1V for the ISO Regen

 

Handy that the Keces displays the current drawn.

For my endpoint PC (motherboard + CPU) it never goes above 1.1A.

Motherboard only never above 0.7A.

ISO Regen never above 0.3A.

[beautiful music]

4 hours ago, bodiebill said:

 

This week I received a Keces P8 with dual rails 9V/12V and 18V/19V (both 4A) plus USB 5V (1A). This enabled me to try quite a few permutations, applying the SR4 and the LT3045 boards where they had the most positive effect. For now, for my HQPe server - NAA endpoint setup, after only one day of breaking in the Keces, I have decided on:

• XT Power battery 5V for the server's JCAT net card FEMTO
• Paul Hynes SR4 19V for the MikroTik 305 SFP/RJ45 switch
• Keces P8 19V for the endpoint PC with HDPLex 800W DC-ATX converter
• Keces P8 9V => LDOVR DXP-1A5DSC (LT3045) => 7.1V for the ISO Regen
• iFi Power 9V into MPAudio DLS-HPULN (LT3045) => 7V into Chinese PS (LT3045) => 5V for the Elfidelity USB card

The latter (Elfidelity card) sounds clearly better than direct from the Keces P8 5V USB, probably because of the double (serial) LT3045 noise reduction.

 

As 19V supplier for the endpoint PC the SR4 sounded only very slightly better than the Keces. The SR4 made a clear difference on the MikroTik switch.

 

I also tried powering the endpoint CPU/EPS seperately with the SR4's 12V, but this did not give a clear improvement compared to the Keces powering the motherboard + the CPU via the HDPlex 800W.

 

I have ordered a 6V Chinese LPS to be able to further simplify and increase the possibilities 🙂

 

Fyi, for what it is worth...

 

Curious to know what is your endpoint PC configuration.

[bodiebill]

12 minutes ago, beautiful music said:

Curious to know what is your endpoint PC configuration.

 

GentooPlayer with NAA    
case    Streacom FC5 EVO
mobo    Supermicro X11SAE-M
cpu    Intel Core i3-6100T (35W TDP)
memory    2x Apacer D31.23185S.001 4GB ECC DDR4 2666MHz (wide temperature)
NIC    stock

USB    Elfidelity AXF-100 PRO III
PSU internal    HDPlex 800W DC-ATX converter
PSU external    Keces P8 19V/4A
SSD/HD    n/a (SATA deactivated in BIOS)

[Exocer]

On 5/1/2020 at 5:05 PM, bodiebill said:

Not much discussed here, but the € 40 the Elfidelity AXF-100 PRO III Internal USB card for me was a very pleasant surprise.

https://www.audiophonics.fr/en/pc-htpc-modules/elfidelity-axf-100-pro-iii-internal-usb-power-filter-for-pc-p-9056.html

In my setup the resulting jump in sound quality was at least as substantial as that when adding the € 435 JCAT net card FEMTO. Apples and pears of course, but still... YMWV 🙂

Both externally powered with 5VDC LPS.

First post after LOTS of reading of this thread. This thread contains a wealth of information.

 

This is my single machine setup, completly inspired by this thread.

 

Hardware:

Asus ROG Strix-I Z390

i7 8700T @ 3.1ghz

Optane SSD58gb (I plan to move away from USB at some point and use this for storage).

Apacer DDR4 2400 8GB

Elfidelity AXF-100 PRO III Interal USB Card

Gotham 18awg OFC DC Cables 

HDPlex 400w DC-ATX Converter (800w has been out of stock forever)

HDPlex AC-DC Converter 400w (Found a deal super cheap. This is temporary until I decide whether to go with the future HD Plex 500W LPSU Larry mentioned, Keces P8 19v (1 to start with, perhaps another directly to the 12V EPS connector in the future or another 19V if I can get my hands on an HDPlex 800W DC-ATX converter).

Etherregen

 

Software:

GentooPlayer Ramrooted - -PF-ST Kernel

UPMPCLI\MPD configured to stream from Tidal\Controlled by Bubble uPNP

 

I don't upsample as I feed the Yggdrasil A2 DAC via motherboard USB (getting to why momentarily).

 

It is a step up from my previous upnp renderer, The Pi2AES Pro Shield, which sent AES output to the Yggdrasil. The server presents a more effortless and transparent sound, with improved bass control and dynamics. The Pi2AES sounds great for the money spent though, even with stock SMPS.

 

Back to why I quoted you @bodiebill, do you use the Elfidelity card in the GentooPlayer endpoint? It wasn't plug and play with GentooPlayer (in my setup) and would appreciate any guidance on this topic. Note: I am not externally powering it yet but will soon have a 5v source for it.

 

Thanks all! What a terrific thread.

 

Cheers,

-Rob

 

 

[bodiebill]

10 hours ago, Exocer said:

Back to why I quoted you @bodiebill, do you use the Elfidelity card in the GentooPlayer endpoint? It wasn't plug and play with GentooPlayer (in my setup) and would appreciate any guidance on this topic. Note: I am not externally powering it yet but will soon have a 5v source for it.

 

Thanks Rob for sharing your build. I do expect that the Elfidelity usb card will make a positive difference once you get it to work.

 

Indeed I use the card in the GentooPlayer endpoint. It does not really make contact with the PCIe slot (plastic connection), so it cannot get its power from there. When I first installed it I connected it only to the internal usb 3.0 connection as I was going to exclusively use an external PS. However, that did not work. I also had to connect the internal power cable (in my case to the HDPlex 800W DC-ATX) even though I am not using it in the end. After that the USB port and my connected DAC were recognized

Not sure how you set it up, so don't know whether this is of any help.

 

One quirk with the Elfidelity card is that when I power down the PC, it starts switching like crazy between internal (by then absent) and external PS. So I make a habit of turning off the external PS before PC shutdown.

[bodiebill]

deleted
 

[Exocer]

3 minutes ago, bodiebill said:

 

Thanks Rob for sharing your build. I do expect that the Elfidelity usb card will make a positive difference once you get it to work.

 

The card does not really make contact with the PCIe slot (plastic connection), so it cannot get its power from there. When I first installed it I connected it only to the internal usb 3.0 connection as I was going to exclusively use an external PS. However, that did not work. I also had to connect the internal power cable (in my case to the HDPlex 800W DC-ATX) even though I am not using it in the end. After that the USB port and my connected DAC were recognised

Not sure how you set it up, so don't know whether this is of any help.

 

One quirk with the Elfidelity card is that when I power down the PC, it starts switching like crazy between internal (by then absent) and external PS. So I make a habit of shutting off the external PS before PC shutdown.

Thank you! I appreciate the tip.

 

 I did the same as you, connecting only the internal USB 3.1 Gen 1 connector. That little red light on the rear panel was a bit deceiving.

 

Cheers.

[bodiebill]

7 minutes ago, Exocer said:

Thank you! I appreciate the tip.

 

 I did the same as you, connecting only the internal USB 3.1 Gen 1 connector. That little red light on the rear panel was a bit deceiving.

 

Cheers.

 

Yes, here red is always on, green means internal power and yellow shows that external powering is succesful.

 

Will you let us know how you like the card? I suggest a few days of break in.

[Exocer]

1 minute ago, bodiebill said:

 

Yes, here red is always on, green means internal power and yellow shows that external powering is succesful.

 

Will you let us know how you like the card? I suggest a few days of break in.

Excellent. I now understand how the lights work (I lost the little paper it comes with in the storm of building the HDPlex case :))

 

Looking forward to sharing my impressions. I suspect that this USB card will hang around until I can get proper clean power to the server... So several weeks/months.

[bodiebill]

3 minutes ago, Exocer said:

Excellent. I now understand how the lights work (I lost the little paper it comes with in the storm of building the HDPlex case :))

 

Looking forward to sharing my impressions. I suspect that this USB card will hang around until I can get proper clean power to the server... So several weeks/months.

 

Do you expect that clean power to the server will make the card redundant? That is not my experience. Even with the best of LPSU's the motherboard appears to add noise that ends up in the stock usb connection.

[Exocer]

1 minute ago, bodiebill said:

 

Do you expect that clean power to the server will make the card redundant? That is not my experience. Even with the best of LPSU's the motherboard appears to add noise that ends up in the stock usb connection.

 

Good question. No, I do not believe the clean power will make the card redundant. My comment was a bit misleading. What I mean is I will prioritize clean power to the server first. Then perhaps I will consider replacing the Elfidelity card with something higher end, ala Pink Faun, JCAT Femto USB etc. 

 

So, the card will not immediately come out once I power the server via clean(er) power but the planning of the replacement card will begin once everything is settled in.

[bodiebill]

1 hour ago, Exocer said:

 

Good question. No, I do not believe the clean power will make the card redundant. My comment was a bit misleading. What I mean is I will prioritize clean power to the server first. Then perhaps I will consider replacing the Elfidelity card with something higher end, ala Pink Faun, JCAT Femto USB etc. 

 

So, the card will not immediately come out once I power the server via clean(er) power but the planning of the replacement card will begin once everything is settled in.

 

I have the impression that my DAC (a Denafrips Terminator) is less sensitive to USB signal upgrade devices. Maybe because it is fine-tuning the stream with its own clock. However, cleaning the USB signal from electrical noise still seems important. That is why the Elfidelitry usb card (€ 40) appealed to me, as that is exactly what it is doing, without the need for a riser card to connect to the PCIe slot (which could add to the burden of the motherboard and lessen the benefits of the expensive cards).

And it did result in a substantial SQ improvement that does not leave much to be desired. Therefore I am fighting that urge that is part of the audiophile disease, namely to buy a "much better because much more expensive" card. Of course I would be happy to try and compare, but not by spending another € 400 when already perfectly happy.

[Exocer]

1 hour ago, bodiebill said:

 

I have the impression that my DAC (a Denafrips Terminator) is less sensitive to USB signal upgrade devices. Maybe because it is fine-tuning the stream with its own clock. However, cleaning the USB signal from electrical noise still seems important. That is why the Elfidelitry usb card (€ 40) appealed to me, as that is exactly what it is doing, without the need for a riser card to connect to the PCIe slot (which could add to the burden of the motherboard and lessen the benefits of the expensive cards).

And it did result in a substantial SQ improvement that does not leave much to be desired. Therefore I am fighting that urge that is part of the audiophile disease, namely to buy a "much better because much more expensive" card. Of course I would be happy to try and compare, but not by spending another € 400 when already perfectly happy.

@bodiebill, yeah, I am sure the Terminator does a swell job with the internal clock. I have read at least one person on the web who preferred the Yggy without re-clocking from an external USB PCI-E card. Would be interesting to see what others have to say on this topic.

 

BTW, I just got the card working and it was indeed plug and play (when powered). Will give this a few days to break in. The card sounds very good so far! Although the USB 3.1 port on the Asus Strix z390-I doesn't sound bad I will say that this card already leans in the direction I prefer. Too early to tell though.

 

Thanks

[bodiebill]

Good to hear. I would give it two days...

And much depends on the external LPS. Enjoy!

[austinpop]

On 3/28/2020 at 1:22 PM, Nenon said:

 

I tried Melco --> Buffalo BS-GS2016 (modified) --> EtherREGEN as I mentioned in my post below.

It sounded better than Melco --> EtherREGEN, which sounded better than just an EtherREGEN. 

Diminishing returns but it gets better.... 

 

@Nenon had kindly sent me another modified Buffalo BS-GS2016, as I was very curious to try it in my system and compare it with both the Melco S100 and the etherRegen. He had warned me it took  a long burn-in period to sound good, so I used it as a regular switch in my home office for a couple of weeks before listening.

 

Comparison between the Buffalo BS-GS2016 and the Melco S100

 

As @seeteeyou first alerted us, the Buffalo BS-GS2016 appears to have an identical board to the S100. The difference is that the Buffalo BS-GS2016 has an inbuilt SMPS. @Nenon's mod was to remove the SMPS, and expose a DC input. Another difference is that the Buffalo's management UI is accessible, while the S100 does not appear to expose any UI. Based on the Chinese forum that @seeteeyou linked to, I set the Buffalo to the following settings:

• top row of ports off
• Ports 2/4/6/8 at 100Mbps, 10/12/14/16 at 1Gbps
• LLDP off
• 802.3az off (aka EEE) on all ports
• ADP (Auto power down) off on all ports
• Flow Control is ON - this is off by default. 

After the extended burnin, I first attempted to answer the biggest burning question in my mind: does the Buffalo match the SQ of the Melco? To my ears, the answer was no. Let me be clearer. The Buffalo had all of the other great attributes of the S100: density, saturation, body, and depth. The family resemblance was very clear. The big difference to me was that the Buffalo had a slightly fatiguing harshness, or sharpness that the Melco did not. And this was with both devices powered by an SR-7 DR rail.

 

My optimum switch chain since I published my eR review has been:

• copper > 'B' eR 'A' > fiber > S100 > server. 

In the above experiment, I was interchanging the S100 and Buffalo as the last switch in the chain. I also tried reversing the order like so:

• copper > (100) Buffalo (1G) > fiber > 'A' eR 'B' > server,

and comparing the Buffalo and S100. Unfortunately, that did not change the outcome. With that said, the Buffalo, priced at just $200ish, is a bargain compared to the S100, and perhaps in a different system, the harshness I was hearing may not be as noticeable or bothersome.

 

At this point, I packed up both S100 and Buffalo and sent them over to Roy, as he has been keen to try these. He posted his impressions in another forum here: https://www.whatsbestforum.com/threads/network-improvements-and-their-impact-on-sound-quality.30455/post-647211

 

Obviously, in his system, the Buffalo was much more tolerable than in mine. This could be down to many things, but ultimately it just shows that beyond a point, you really have to use your own ears to truly evaluate gear!

 

 

[Nenon]

I was trying not to post this publicly, but now that word is out, and given the comment below, I decided to share:

Quote

 @The Computer Audiophile requested review samples of the Aqvox SE and the Melco S100 switches. Neither were provided. 

 

Shortly after Rajiv's review of the etherREGEN switch, he sent me the Melco S100 switch he used in his review. Rumors were that the Melco S100 is identical to the Buffalo BS-GS2016, but installed in a nicer case (and possibly with a better clock). I opened both switches, took the boards out, removed and the heatsinks and compared the two. A picture is worth a thousand words, so here you go:

 

The board on the left is from the Melco. The board on the right is from the Buffalo. Let me know if you spot any differences. Even the printed numbers are the same. I thought that Melco used better components and upgraded clock on this board, but that was not the case. 

 

The Buffalo switch has built-in off the shelf SMPS that is easy to remove. Then you can drill a hole for a DC connector, solder two wires, and you are done. The switch is ready to be powered with external power supply. I did a few other things on mine - dampened the chassis, installed proper footers, used Mundorf silver/gold wire, etc. 

 

The Melco comes with an external SMPS. Part of the "voicing" of the Melco was adding a bank of extra capacitors to smoothen the noise of the SMPSS it comes with.

Those look like carefully selected capacitors that they must have used to "voice" the switch. Good I am not writing in the objective forum or I would have been torn apart for my last sentence... voicing a switch :). Yes, by manipulating / filtering the DC before it enters the switch, they were able to give it a smooth character.

 

My thinking was the following:

1. Those carefully selected capacitors are voicing the switch to be used with the included SMPS. If we are using a really good LPS like the DR SR7 @austinpop used, they may not be needed. They can even harm the sound as they are arguably lower quality than what I am using in my LPS.

2. The Melco wiring is done with cheap/regular wire. I used the best wire I have found for DC. That may have a benefit...

 

Was that the case? When I first plugged in the modified Buffalo, it was very harsh. I could hear some of the properties of the Melco, but there was a big difference in favor of the Melco. If I did not know the boards were identical, I would not have had the patience to wait for the Buffalo. It took about a month of non-stop running of the Buffalo to come close. At that point an A/B comparison was close but they still sounded different. The Melco was always the smoother sounding switch. It had rounded edges. The Buffalo was more aggressive, more edgy, but not necessarily edgy in a bad way.  

 

Let me put it this way. If you like reduce the speed of your streamers to 100 Mbps to get a smoother sound, you may not like the Buffalo sound signature. If you have an etherREGEN and going across the moat (i.e. A to B) is your preference, you will probably not like the Buffalo switch. But if you like me prefer to use the A-side of the etherREGEN only, the Buffalo might be a really good choice. In fact, adding the modified Buffalo to the etherREGEN makes as much difference as adding the etherREGEN to my system before I had any audiophile switches. Essentially the Buffalo doubled the effect of the etherREGEN for less than the cost of the etherREGEN (which was a bargain to begin with). 

 

[RickyV]

About a month ago I ordered a Buffalo  BS-GS2016 switch (90 euro) so I could experiment a little more. Feeding my ER with optical or cable, buffalo first ER last etc. But first modify the buffalo switch.

 

After some investigating I learned that the switch has 3 switch mode regulators 3.3V, 1.5V and 1V. So I made the plan how to power the switch with linear power as much as possible. So I have ordered parts to build a dual rail studer900 power supply, one rail for my modem/router and one rail for the buffalo switch. I like the dual regulator concept so I try to implement as much as possible. That’s why I build a 3A ldovr stack for the 12V power. I am still waiting on the transformer so I temporarily used the original smps which outputs 12.2V to feed the ldovr stack. I also have the plan to disable the 3.3V sm regulator and feed the 3.3v also with a ldovr regulator when parts arrive, you can do that with the 1.5V and 1V too. Will also experiment with buffer capacitors for the switch mode regulators.

Sorry if this is a bit to much diy.