I replaced a wall wart with a linear power supply

[AlainGr]

I have had the NOS1 since January 2013 and have been very happy with the SQ that was added to my audio chain. It replaced a Bryston BDA-1 dac that I enjoyed very much before.

While I was using this previous dac, I tried a few power supplies for the Audiophilleo 2 USB to SPDIF converter, that was directly plugged at the dac input. This converter is normally powered by the USB 5 volts that comes from the PC. To improve the SQ, I got  an Aqvox USB power supply, that replaced the "noisy" power with a clean 1mv ripple power.

Then I heard about the iFi iUSB power supply, with its 0.1uV ripple - better than the Aqvox. Bought it.

I "thought" that each change was doing something good. I am using  "thought" in quotes because there were slight changes, but I could not really turn them into images. Anyway, if I did not try, I would never have known...

When I removed the AP2 + Bryston in favour of the NOS1, I had concerns about putting anything in between the pc and the dac. It seemed that Peter had thoroughly thought about every aspect between pc and dac. XXHighEnd is so musical and adjustable that putting anything else would have been "heretic".

The Aqvox was not needed anymore, so was the iUSB.

And for the first time, I could use the Sotm PCIe USB card that had been idling in my PC, because there were latency and/or buffer problems (Blue Screens) with the Audiophilleo before.

With the Sotm, I can cut the power off (5V) (there is a switch right beside the USB connection on the Sotm) after XXHE has established a connection with the NOS1.

But after a while I decided to put the iUSB "back to work". I removed the SSD that was in my music pc and put it in an external enclosure, like the music hdd (2TB, 7200rpm). The two of them are connected to the pc through Esata. The SSD holds W8/64 pro and XXHE. So now the iUSB powers the SSD.

Did I hear an improvement ? Not really.

5 days ago, I received another linear power supply, this time for my music hdd. While the price of that linear power supply is twice the value of the drive + the enclosure, I again took a chance.

Let me put it this way: I am not very sensitive to slight variations in SQ, way less than many of the posters in this forum. I tried very few USB cables, had paid quite an amount of money for interconnects and AC cables... But NO, I did not really feel that there was an improvement in SQ. I can't specify if a USB port is better than another, not even if I compare them with the Sotm USB card... I apply recommended tweaks, try a few others, appreciate what I can learn from fellow posters. As for my hearing abilities, well all I can say is that I am happy to still be able to hear something, as I have abused my ears many eons ago. 

With this in mind, I was not expecting to hear a damn difference.
I have read a lot of descriptions about one's perceptions but to image what I perceive now is hard for me. The nearest approximation that I can give would be something like "Tthere is AIR between the notes"... It may not be it, but it is surprisingly good. The LF are tight, the notes are clearer... It is as if a "sizzling" was removed.

In each step of my journey, I have heard some improvement through time, but at a certain point all has been subtle and my expectation bias is not high. Most of the time, I would say that except for XXHE and the NOS1, many expenses were done for not much.

If you have the chance to find a linear power supply that will fit with your budget, try it. The power supply I bought is a TeddyPardo brand, model "Teddy12/4"). There is also the "Teddy12/2" power supply. I chose the 4 amps version because it could accommodate more powerful devices (48 watts max), but the 2 amps (24 watts) will certainly fit perfectly for a hard drive. As long as the DC voltage output is the same as the wall wart  it replaces, and the amperage is equal or higher, the power supply should be perfect for the job.

There are other brands and I am quite sure that they would achieve what I am perceiving now.

This TeddyPardo power supply is expensive if you compare its price vs the price of a hard drive + its enclosure. But if I can hear what I hear now, you will most likely hear something too.

I don't like to write superlatives, because I don't feel it is necessary. I can only try to contribute with fellow audiophiles what I am perceiving now.

A linear power supply for the PC ? I hope there will one powerful enough someday

Alain

[juanpmar]

Great post Alain, and thanks so much for sharing your findings, I´m sure you can hear more of you say and you say many useful things. Reading your post I remembered that I had in some place the Aqvox that I used some time ago, I´ll look for it to see if I can use it again. About Teddy Pardo I took a look to his power supplies and they look very interesting. Please let us know your experience with the Teddy Pardo when you have some more time with it.

Kind regards,
Juan

[AlainGr]

Thanks Juan for the kind words

The power supply is still making its magic on the rendering. Now I know that what has changed is the level in noise. It is a lot "blacker" than before and because of this, I can "feel the matter" much more.

Regards,

Alain

[boleary]

[quoteA linear power supply for the PC ? I hope there will one powerful enough someday][/quote

http://www.coreaudiotechnology.com/products/music-servers/atx-linear-power-supply/

Be sure to check out the disclaimer at the bottom of the page.

[PeterSt]

... and the extra box. And the 50LBS.

Wow ...

[boleary]

Do you think the 200 watt "S1" is enough power if fan-less and one HD?

[PeterSt]

Yes. But it sure is in the danger field. So, I was thinking about the XXHigHEnd PC. Here the processor consumes 130W max. Video card a bit, a disk a bit, Mobo, things you don't think about and ... close to it.
So with this 130W processor I would not do it. Then rather 500 buks extra. Nothing tells about 50LBS extra, so ...



But which all doesn't tell me really that this helps SQ. I just don't know. It can for sure, but don't blame me because of some enthusiastic Wow when it does not.

Peter

[boleary]

I thought the extra 500 would yield an extra 10-15 lbs.  Did you see they build a music server for $65,000?

[acg]

Hi guys,

First post here.  Downloaded XXHighEnd a couple of days ago, thought I might also try to contribute.

Linear supplies...I'm playing around with one at the moment that feeds the usb output from the PC.  Now my PC is a quite a bit different to the ones recommended for the Phasure PC...passively cooled i7-3770T, 8GB ram, ssd for os, pico psu et cetera.  My dac is also a lot different to the NOS1.  

Two things that have improved my setup are: a customised PCIe - USB card (similar to the SOtM card that AlainGr has installed, but it has been modified with better TCXO, a little silver output transformer and some other goodies and must be fed its own power source); and the EA Shortblock Common-Mode filter.

On trial at the moment I have the Adnaco S3B with a high quality linear psu to head-to-head against the modded USB card mentioned previously.  I can tell you that the linear psu makes a significant difference to the Adnaco, but I am not yet sure if the Adnaco is for me.  I will power the modded USB card with the same linear psu to compare results.

Another project that I have on the go is to try to build an atx linear psu for the pc, but that is not going to happen overnight.  One thing that I have gleaned from others that have gone the linear psu to the pc route is that the quality of regulator module and components is paramount:  not only low ripple but ultra-fast response also.
I will do some more digging on this site and try to figure out the recommended PC build for XXHighEnd and the NOS1.  Maybe then I can help some more with the linear psu and other recommendations.

One question for PeterSt.  Do you think that a common mode filter between the PC and NOS1 will help at all?  I am not familiar with the USB input on the NOS1, but the maker of this filter has posted jitter measurements here. (see post #9)

Cheers,

Anthony

[acg]

Sorry PeterSt, but I have another question for you, and please forgive me if you have answered this elsewhere.

Are you able to offer any insight into why a linear psu feeding a pc may improve sound quality in a dac?  I have been asking this question on another forum and nobody there seems to be able to provide an answer other than "your dac is broken" or "it just does" (neither answers are particularly helpful but they are insightful).

By the way Peter, I only stumbled across your NOS1 project a few days ago but it seems very forward thinking to me.  I would love to hear one. 

[AlainGr]

Hi Anthony,

I am very interested in the conclusions you will pull from your comparison between the Adnaco S3B and your modded PCIe USB card.

You should see in one of the thread that it was tested about a year ago... The results were mixed at the time, more on the negative side... My attempts with it were not successful then (had problems with the Audiophilleo and BSODs)...

I did not like the sound then when I could play it for a few songs...

http://www.phasure.com/index.php?topic=2089.0

Alain

[acg]

Hi Alain,

I have done my critical listening today comparing the Adnaco S3B and the PPAstudio USB card both fed with a high quality linear psu of similar quality to the Pardo unit you have Alain.

I have had the Adnaco + linear psu in my system for three days now (on loan from a friend) and I must say that in general I thought that it was a step up from the PPAstudio card when fed by the internal PC smps power.  I have previously tested the SOtM card against the PPAstudio card and for me the SOtm was easily bested.

One thing that I can say about the Adnaco while it was in my system was that for all the improvements in clarity and cleanliness it brought it also had some unwanted baggage.  The upper mids and treble were 'steely' and 'thin' to me, and not as accurate as I thought they could be.  I never tested it with the standard wall-wart, but the owner of the unit said that the linear psu brought a very significant step-up in sound quality.

This morning I made up a cable to extend the linear psu feed to the PPAstudio card so that it received exactly the same feed as the Adnaco.  There is not much to say other than for me the PPAstudio card fed linear power was significantly better than the Adnaco.  The entire sound signature was fuller and more refined and the 'steeliness' had disappeared.  My brain is better able to read the spatial cues and when playing rock like The Drones, Daft Punk or The National the drums seem more solidly placed within the soundstage and just a little further behind the guitars than with the Adnaco, more 3d if you like.  I like it, I like it a lot.

I hope this information is useful Alain.  I will now do a search and see what the Adnaco thread has to say here.  I may even agree with someone!!

Cheers,

Anthony

EDIT:  Have read the Adnaco thread...I am in general agreement with the main protagonists there...phew.

[AlainGr]

Hi Anthony,

Thanks for the information - really appreciated Noise is hard to conjure, but it was about the first time I discovered its merits when I added the linear power supply to power my music hdd. It really has been a revelation. It seems I am more sensitive to certain aspects of audio, while for other things I can't appreciate a difference...

There is something I am not sure. I think that Peter mentionned it more than one time, but can't locate it. It is about elimination of the noise right from the start.

Does the noise that would have made its way through the signal prior to the Adnaco still remains even if the data is converted into light, then back as an electrical signal... This I wonder...

Would jitter explain the SQ you experienced with the Adnaco, as it is often described because of the conversion at the start and end ?

Do you have a link where I could get more information about the USB card you mention ?

By the way, welcome among us

Regards,

Alain

[AlainGr]

Anthony,

Would it be the Paul Pang USB card ? I read some info about it some time ago...

I wonder if he also makes Esata cables...

Alain

[acg]

Yes Alain, it is the Paul Pang card.  Have a look near the bottom of this page.

He makes quite a few things including sata, lan and usb cables.  I have one of his sata and usb cables on loan here at the moment and am going through the process of getting my head around them.  I can't think of a single good reason why a sata cable would improve sound.  The only thing that I can come up with is perhaps if it acts as an antenna for emi/rfi but that may be tenuous as best

Tomorrow I will make up a power cord to run my os ssd from the linear psu.  Then I will try it on the music ssd (which strictly speaking does not exist because I network the music files in from a NAS at present) but I will pull one out of another computer and plonk some music files on it to do some listening.

You are not the first to report that powering the ssd or hdd has an impact on sound so I will be very interested to do these little experiments.

I can tell you that on two occasions where there was no linear power to the pc that I could not pick a difference with the expensive sata cable that I have here.  However, tonight, with the linear psu feeding the usb card I could pick an extra 'crispness' and 'clarity' to the sound I was hearing.  It is important to note that I am not convinced of the change as yet and need to do some more listening over the next day or two to confirm it.  It could simply be placebo or psychoacoustic.

I don't know the technical aspect of why the Adnaco did not work for myself and others.  Maybe it is jitter: I understand that the conversion can induce jitter.  But the 'steeliness' I heard did not sound like jitter I think I have identified before.  Like many I am but an armchair expert!

I will let you know how I go with my ssd experiments tomorrow.

Cheers,

Anthony

[AlainGr]

I will do some reading. Thanks for the link and contribution

Regards,

Alain

[PeterSt]

Sorry PeterSt, but I have another question for you, and please forgive me if you have answered this elsewhere.

Are you able to offer any insight into why a linear psu feeding a pc may improve sound quality in a dac?  I have been asking this question on another forum and nobody there seems to be able to provide an answer other than "your dac is broken" or "it just does" (neither answers are particularly helpful but they are insightful).

By the way Peter, I only stumbled across your NOS1 project a few days ago but it seems very forward thinking to me.  I would love to hear one. 

Hey Anthony - warm welcome here !

Well, as you will have seen, I hasted to say that I have no real judgement on that to begin with.

But which all doesn't tell me really that this helps SQ. I just don't know. It can for sure, but don't blame me because of some enthusiastic Wow when it does not.

So how to improve upon a "your DAC is broken" etc. ?
Well, the response can't be simple and I'm afraid it will be too brief and will show that I really don't know.

Let's take a USB (asynchronous) connection;
This won't show any switching PSU sh*t that I have seen. It will show its own transfer-current noise (tough to solve in general) but that's it. So, no linear PSU will help there.

Notice though that I assume a good grounding scheme for which no rules exist. Have that wrong, and you may be hearing your mouse through the speakers. But, through what connection that goes ?

Now we have a random switching supply (could be from a phone charger, could be from the PC). Still no problems in the USB link. BUT, problems on the mains. So, all sorts of nasty things can be seen there.
Do we see this in the DAC ? again, with proper grounding we do not. And I mean overall "house" grounding around the stereo.

And this is about it; when all is grounded correctly, no problems will be seen anywhere. Still this sh*t is on the mains. But now the point : you won't be able to see this sh*t unless you had an analyser for the job ($$$$).
And NOW avoiding the problem in the first place is crucial.
So yes, it can help. But no, it won't help me. Well, that is what I think, because influences need to be very minimal only to imply SQ changes, and they will be in the area of unmeasurable. So, same like XXHighEnd influences SQ - which can not be measured as well.

So, YMMV ?
Peter


PS: For fun look here : Buzzing noise at plauback
The culprit (as known by now) : A too old tube in some amp. Do we get that ? ok, I don't.

[PeterSt]

One question for PeterSt.  Do you think that a common mode filter between the PC and NOS1 will help at all?  I am not familiar with the USB input on the NOS1, but the maker of this filter has posted jitter measurements here. (see post #9)

Assuming Async USB, it can't. Well, it can, but you'd be filtering in-band audio.
USB noise expresses at 8KHz, 16KHz (and 24KHz and beyond) and nothing is able to filter it out. Smart grounding schemes do though (applied in the NOS1).

Note that USB noise is about the current needed to get in (!) the packets of data.
This is also how no glass connection will help that (and how an Adnaco can make it worse by first a poor PSU and next more current needed to get the job done).

Peter

[acg]

Thanks for your time Peter.

Note that USB noise is about the current needed to get in (!) the packets of data.
This is also how no glass connection will help that (and how an Adnaco can make it worse by first a poor PSU and next more current needed to get the job done).

I have learned something today!!  Can you tell me where in the computer the packets are formed?  Is it in the host controller (EHCI)?  If we were to use a PCIe to USB card, for example, would that shift the location that the packets are formed to the PCIe card itself?  Could this be why I hear a difference with my modded usb card that isolates itself from the pc with a transformer and relies on external power?  I have also noticed a _huge_ difference in sound with just a normal $15 PCIe card in an busy Xeon based workstation: the sound went from 'crackles and pops' to actually listenable!!  I put that one down to better power supply to the usb controller, but of course I may be wrong.

So many questions for a newbie. 

Anthony

[Robert]

Well I replaced a switched mode power supply for a linear one on my M2tech Hiface Evo and whamo what a difference in sound. Vast in fact. POWER SUPPLY IS EVERYTHING never changes!!!!

[PeterSt]

I have learned something today!!  Can you tell me where in the computer the packets are formed?

I don't know, but let's say it is in that particular hardware.
Two reasons though why I regards this as less important :

1. Implied software activity is I think more important (think Driver hence OS);
2. The sh*t happens at the other side merely.

So, in the DAC all sh*t happens, or that is where all comes together if you like. In-PC as much noise and as many spikes can happen what you want - it is about that arriving in the DAC what is the only place where it can be harmful. And indeed, the path through the mains is the most complicated (and 100% USB related), let alone the radiation coming through air where you think you could kill it (like an off line UPS could imply that).

USB2 vs USB3 is also such a thing. We are all sure that USB3 sounds better (for the NOS1 at least). Still my own reasoning can only go in the other direction, and it should sound worse. Well, with all the tweaks and things applied the last couple of years, maybe someone must try USB2 again.

This whole thing is so super complex that at least I may never find the real merits of it all.
But hey, aren't we with many more ?
*That* is our chance ...

Peter

[acg]

Hi Alain (and others),

I've finished my listening impressions with powering my SSD with a good quality power supply and also with an expensive 'audiophile' sata cable.  Please note that I neither run XXHighEnd nor a NOS1 (although plans are underway to change that) so of course take this with a grain of salt.

I could hear no sound quality difference from a ssd regardless of whether it was powered by a good quality linear psu or it was fed smps pc power.  There may be a difference in a more revealing system than mine, but I could hear nothing.  Likewise, when the pc was devoid of power from any linear psu i.e. was powered entirely by smps I could detect no difference between the expensive and generic sata cables.  But, when the linear psu was applied to either the PPAstudio usb card or the ssd and the expensive sata cable was in place I could hear a difference: a general crispness across the entire spectrum.  Did this sound better?  Maybe, I am undecided.  Was this difference significant?  Not in my system.  The difference was pretty faint really and considering the cost of the linear psu and the cable certainly the expense is not warranted for me.  I've heard much bigger differences tweaking the operating system or changing cables, even digital cables.

The interesting thing in all this is that the SATA sounded different when linear power was applied to the pc and it did not seem to matter if the linear power was supplied to the ssd or not.


Anyway, I'm glad I could contribute a little.

Cheers,

Anthony

[AlainGr]

[quoteA linear power supply for the PC ? I hope there will one powerful enough someday][/quote

http://www.coreaudiotechnology.com/products/music-servers/atx-linear-power-supply/

Be sure to check out the disclaimer at the bottom of the page.

Hi Brian,

Reading through the posts again, I realized I missed your post. Well a linear power supply exists... The price is steep, but expectable... I just wonder if it would be powerful enough...

There are some who suggest to put nice filtered DC to DC regulators (and filters) to reduce the frequencies effects, but I wonder if that would improve the ripple itself ?

Alain

[AlainGr]

Hi Alain (and others),

I've finished my listening impressions with powering my SSD with a good quality power supply and also with an expensive 'audiophile' sata cable.  Please note that I neither run XXHighEnd nor a NOS1 (although plans are underway to change that) so of course take this with a grain of salt.

I could hear no sound quality difference from a ssd regardless of whether it was powered by a good quality linear psu or it was fed smps pc power.  There may be a difference in a more revealing system than mine, but I could hear nothing.  Likewise, when the pc was devoid of power from any linear psu i.e. was powered entirely by smps I could detect no difference between the expensive and generic sata cables.  But, when the linear psu was applied to either the PPAstudio usb card or the ssd and the expensive sata cable was in place I could hear a difference: a general crispness across the entire spectrum.  Did this sound better?  Maybe, I am undecided.  Was this difference significant?  Not in my system.  The difference was pretty faint really and considering the cost of the linear psu and the cable certainly the expense is not warranted for me.  I've heard much bigger differences tweaking the operating system or changing cables, even digital cables.

The interesting thing in all this is that the SATA sounded different when linear power was applied to the pc and it did not seem to matter if the linear power was supplied to the ssd or not.


Anyway, I'm glad I could contribute a little.

Cheers,

Anthony

Hi Anthony,

Thanks for the output. Maybe the wall wart powering the SSD is good enough, so the linear power supply does not provide an audible improvement ? As I mentionned in my first post, the iUSB did not provide audible results ni comparison with the original wall wart... That SSD contains both the OS and XXHE. With the music drive though, the difference was "stunning", but it could be because the wall wart with that enclosure is really cheap - I don't know.

Though the "recipe" to access a silent system would depend a lot on the parts and their interactions, I suppose that the results will vary from one system to the other.

On my side, there is a frequency I can hear at times and I have not been able to get rid of it yet. I can hear it when the music finishes. I don't know where it comes from. Maybe disconnecting a few things will help to pin point this...

Alain

[PeterSt]

Get out your iPad, look for the FFT program SpectrumView from OxfordWaveResearch (free), and hold it around / in your PC;

I know of a situation that some on-board oscillator (or whatever it actually is) produces the same frequency as USB transmits, and this is 8KHz. So look for that.
The USB transmitting stops a few seconds after playback.

But of course it can be something else ...

Peter

[AlainGr]

Get out your iPad, look for the FFT program SpectrumView from OxfordWaveResearch (free), and hold it around / in your PC;

I know of a situation that some on-board oscillator (or whatever it actually is) produces the same frequency as USB transmits, and this is 8KHz. So look for that.
The USB transmitting stops a few seconds after playback.

But of course it can be something else ...

Peter

Ah !... Will try Thanks !

Alain