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

[hurka]

4 hours ago, Cornan said:

 

Great stuff! Be sure to report back your findings. The independant u cap psu is surely in for a cat fight! ?

just want reduce the psu`s quantity.Every time I need charge up ,and connect -reconnect .Not the big problem,but this new smps little interested, new challange.

[mourip]

On 12/30/2017 at 9:29 PM, bobfa said:

What an amazing collection of information. It took me over three days of “spare” time to read the whole thing.  Thanks to EVERONE.

 

I am evolving my systems and have upgraded my DAC/AMP  to the new Brooklyn DAC+ and AMP.  My current front end is Sonic Transporter i5 and an microrendu.  I have SOTM SMS-200 ultra and a modded switch on the way. I have held off on the USB re-clock part for just a bit.  I am looking at the grounding and shielding ideas that I can do on my own.  The idea of sleeving the USB with a shield and grounding the Switchs, etc.  will be fun.

 

i have a couple of server questions.

 

Has anyone used the new Sonic Transporter i7 bridged server?

 

I am looking to try my i7 Intel NUC with Windows 10 to see how that works for sound quality and bridging networks.  There was a whole section in the thread on out manually optimizing windows.  Does anyone have any thoughts about that or one of the methods?

 

—-RJF

 

Welcome. You are in for a lot of fun and much improved sonics.

 

Be sure to read John Swenson's posts on grounding.

[sig8]

11 hours ago, lmitche said:

Moussa,

 

Startech USB 3 cables sound best here.  I've been through 4 to 5 brands, but no audiophile versions.  I'm happy with the Startechs  and not going there.

Are you referring to the one which came with the StarTec box or something else?

[Cornan]

1 hour ago, hurka said:

just want reduce the psu`s quantity.Every time I need charge up ,and connect -reconnect .Not the big problem,but this new smps little interested, new challange.

 

I hear you! I got tired of using battery supplies for the very same reason. Now I have only got one BPS powering the BluWave’s GI output. Anyway, I am interested to know how sPS-500+LT3045 turns out, since this is one of my opt’s.

[austinpop]

12 hours ago, Johnseye said:

 

I'm surprised by the preliminary results of the SR4.  If after your burn it in further, and it's the same result, I suspect the sPS-500 the better supply and says a lot for Paul's new line.  Disappointing and I'm glad I waited as I was about to order one.

 

That was @limniscate's view. I have not done any comparisons yet, but will in the next few days.

 

12 hours ago, Johnseye said:

Did you buy a Zenith?  That's a lot of coin.  Congrats.

 

The Zenith SE's long-term future in my system depends on how much of the trifecta it displaces, to offset the cost. We shall see.

 

12 hours ago, Johnseye said:

I just started a thread on breaker to outlet AC wiring.  Would appreciate your input if you've been doing some research.  6awg is some thick wire.  I'm guessing the outlet is designed for it.  Looks like we continue down the same path.

 

Link?

[Johnseye]

1 minute ago, austinpop said:

Link?

 

 

[LTG2010]

16 hours ago, mozes said:

3- HDD (OS+Data)+SOTM SATA II filter: 94 (very warm and lush at the expense of air and extended highs). My interpretation here is that we may have too much filtering between the tX-USBexp and SATA II filter.

Thanks for the information.

The HDD (os+ Data and sotm filter 2) I assume was plug into computers SATA port in scenario 3? and powered externally ? If so did you unplug the filter and try it simply (HDD os + Data)? I think this 'MAY' be how @ Elvia has it.

Just trying to figure out wether its the swap from HDD to SSD or the interface which has yielded the slight improvement between '1 & 2'

The SATA 2 filter is not an ideal match for the sata 3 HDD it also filters the data line and could account for the loss of dynamics, whereas filtering the power line alone might have improved matters.

In effect - scenario 1 you are converting SATA to USB then filtering this through the SOTM input USB card to pCIE interface then re- filtering this again via the SOTM output USB.  Again here you have 2 sCLK EX clock taps in action for both USB cards. _ Intriguing stuff, Its great that you're getting improvements in sound from your efforts.

 

[mozes]

1 hour ago, LTG2010 said:

In effect - scenario 1 you are converting SATA to USB then filtering this through the SOTM input USB card to pCIE interface then re- filtering this again via the SOTM output USB.  Again here you have 2 sCLK EX clock taps in action for both USB cards. _ Intriguing stuff, Its great that you're getting improvements in sound from your efforts.

That’s correct, Regarding SSD bs HDD I did play music from the SSD +Sata II filter and from the HDD plugged to the tX-USBexp when the OS was hosted on the SSD. I preferred by a small margin the HDD. Without a careful A/B testing, it really doesn’t matter. I am finding that the impact of PSUs and its DC leads is much more significant than the whole arena of SSD/HDD/SATA/OS location.

It could be that tweaking in this arena has more to do with clocking and in every scenario, I always have at least two sCLk clocks in the path, so the gains are expected to be minimal. However, with PSUs, we are probably playing in a completely different arena and have another beast to tame. This brings the thought of what @austinpop mentioned before about clocking and Power supply being orthogonal. Interesting times and for sure a lot out there that needs to be uncovered.

[TJHUB]

This SSD via USB HDD talk got me wondering.  So I did a little experiment this morning.

 

I spent the last couple of days getting Windows Server 2016 Core running.  I use Foobar2000, JPLAY, and Audiophile Optimizer 2.2b5 on a hibernated single PC.  And while 2016 Core has a major increase in clarity, I did not like the sound.  Treble was fantastic in clarity and tone, but the midrange lost tonal density, and the bass was horrible and buzzy sounding.  This all compared to Server 2012 R2 in Core mode.  

 

This morning, I cannibalized an old 320gb 2.5” portable drive to use the HDD as my OS drive.  I did an exact install on the spinner, which is powered by an HDPlex linear PS, and connected to a SATA 2 connector on my motherboard.  

 

The change in sound is staggering.  The entire presentation has changed, and from my first half hour of listening to my test Tracks, everything is much better.  The treble is extremely clean and tone is just great.  The midrange is clean and dense with layers, and the bass is tuneful and punchy.  It’s like everything sort of fell into place.  But the real change is in my connection to the music.  Clarity is not great sound.  I had no connection to the music like this before.  

 

It’s too early to claim awesomeness here, it’ll take me the rest of the week to really know, but so far this is just amazing for me.  I left spinning HDDs behind many years ago.  I find it strange that I’m back to one, only to find what I was missing.  I think my Samsung SSD is going in the garbage...(not literally, as it will find its way into a laptop).

 

BTW, my wav library is stored on a Synology NAS.

 

Thanks for this discussion!!

[sandyk]

1 hour ago, TJHUB said:

It’s too early to claim awesomeness here, it’ll take me the rest of the week to really know, but so far this is just amazing for me.  I left spinning HDDs behind many years ago.  I find it strange that I’m back to one, only to find what I was missing.  I think my Samsung SSD is going in the garbage...(not literally, as it will find its way into a laptop).

 

You would have more than likely got the same improvement if you used the SSD the same way, which means no interaction with the internal PSU. In fact, if you markedly improve the power to an internal SSD you can obtain similar results.

It's actually easier to improve the power supply to an internal SSD than an internal HDD, because they don't draw so much current.They don't use the +12V supply either, unless you use this to regulate the power for the SSD down to the required +5V

as I am doing.

[TJHUB]

13 minutes ago, sandyk said:

 

You would have more than likely got the same improvement if you used the SSD the same way, which means no interaction with the internal PSU. In fact, if you markedly improve the power to an internal SSD you can obtain similar results.

It's actually easier to improve the power supply to an internal SSD than an internal HDD, because they don't draw so much current.They don't use the +12V supply either, unless you use this to regulate the power for the SSD down to the required +5V

as I am doing.

 

Sorry if I wasn’t clear, but the SSD was powered the same exact way.  I was trying to convey that the ONLY change was from SSD to HDD.

[sandyk]

29 minutes ago, TJHUB said:

 

Sorry if I wasn’t clear, but the SSD was powered the same exact way.  I was trying to convey that the ONLY change was from SSD to HDD.

 That suggests that the power supply used was not able to fully clean the power to the SSD where the current demand spikes are much more abrupt. The HDPlex PSUs are not very low noise, and I have made JLH PSU add-on PCBs for a couple of local members to further improve the power from their HDPlex to obtain the best from their SSDs. 

Incidentally, there have been quite a few problems with the HDPlex, where the 5V rail has failed with the voltage regulator going short circuit , applying +19V to the load ! Such a failure destroyed a friend's expensive USB card as well.

This can also happen if you set the adjustable voltage rail close to +5V with a reasonable load. Later models,( and repaired modules) have voltage and current overload installed/retro-fitted..

[TJHUB]

40 minutes ago, sandyk said:

 That suggests that the power supply used was not able to fully clean the power to the SSD where the current demand spikes are much more abrupt. The HDPlex PSUs are not very low noise, and I have made JLH PSU add-on PCBs for a couple of local members to further improve the power from their HDPlex to obtain the best from their SSDs. 

Incidentally, there have been quite a few problems with the HDPlex, where the 5V rail has failed with the voltage regulator going short circuit , applying +19V to the load ! Such a failure destroyed a friend's expensive USB card as well.

This can also happen if you set the adjustable voltage rail close to +5V with a reasonable load. Later models,( and repaired modules) have voltage and current overload installed/retro-fitted..

 

I did try the famous LT3045’s without success on my SSD, so I’m not certain what you stated is exactly the case here.  I have not tried the LT3045 on the spinner, but I might.  I didn’t like anything I got from my LT3045’s boards, so I should just sell them.  

[sig8]

3 minutes ago, TJHUB said:

I didn’t like anything I got from my LT3045’s boards, so I should just sell them.  

You are not alone.

[TJHUB]

Just now, sig8 said:

You are not alone.

 

Interesting...I thought I was alone.

[LTG2010]

6 hours ago, austinpop said:

The Zenith SE's long-term future in my system depends on how much of the trifecta it displaces, to offset the cost. We shall see.

The zenith SE Seems (from a picture i've seen of it's internal's) to consist of a (carefully chosen) Supermicro motherboard, an ssd connected with a standard sata cable and a premium triple rail power supply. All the attention seems to have gone in providing low noise power, anti vibration, emi rejection etc. No fancy clocks, ( maybe i'm wrong here)  USB  or ethernet cards.

It just shows how clean power can make a difference. I do expect that you will be attaching your tx- USB Ultra and Master clock to its tails.

 

[RickyV]

And i think i saw a pico psu on the atx connector plus a second power connection to the mobo of the zenith.

I don’t know i do not think it will be a keeper.

[LTG2010]

35 minutes ago, RickyV said:

And i think i saw a pico psu on the atx connector plus a second power connection to the mobo of the zenith.

I don’t know i do not think it will be a keeper.

Its a celeron j900 10W motherboard, theres a  DC input to the motherboard, one to the hard drive and one to the 'Pico' type dc - dc connector which seems to power the optical drive. Those are Mundorf Capacitors on the regulator board- very expensive, it apparently ran @ romaz set up pretty close, so it would be very interesting to hear @ austinpop's evaluation.

[LTG2010]

1 hour ago, LTG2010 said:

Its a celeron j900 10W motherboard, theres a  DC input to the motherboard,

Looking at this a bit closer, the designers have decided to use the atx connectors thus bypassing most of the motherboards regulators.

A 4 pin atx connector (this I presume links to the Celeron processor directly) is fed a clean 12V from the linear PSU. Then the 24 pin atx power connector is supplied by the Pico connector (again from a second rail from the PSU) the third rail to the Hard drive. Clever stuff.

http://www.supermicro.com/products/motherboard/celeron/x10/x10sba.cfm

[Dev]

16 minutes ago, LTG2010 said:

Looking at this a bit closer, the designers have decided to use the atx connectors thus bypassing most of the motherboards regulators.

A 4 pin atx connector (this I presume links to the Celeron processor directly) is fed a clean 12V from the linear PSU. Then the 24 pin atx power connector is supplied by the Pico connector (again from a second rail from the PSU) the third rail to the Hard drive. Clever stuff.

http://www.supermicro.com/products/motherboard/celeron/x10/x10sba.cfm

 

The Supermicro manual says that 4-pin 12v is optional power source and doesn't specifies that it can power the CPU directly. Wonder if the Zenith folks have made any mods to it or to the pico.

[sandyk]

4 hours ago, TJHUB said:

I did try the famous LT3045’s without success on my SSD, so I’m not certain what you stated is exactly the case here.  I have not tried the LT3045 on the spinner, but I might.  I didn’t like anything I got from my LT3045’s boards, so I should just sell them.  

 

TJHUB and sig8

 The LT3045 PCBs have been discussed in detail in the Uptone area of the forum.

Unless the amount of capacitance at their input is markedly increased using normal type electrolytic capacitors in parallel, they are likely to sound too detailed and perhaps a bit harsh sounding.

If they were used in a normal PSU which has a large value Filter Capacitor after the rectifier diodes they should sound far more tonally balanced with an improved low end .

 Don't forget too, that some SSDs , especially the much faster types require a great deal more current than others such as the Samsung Evo  840 and 850. Did the SSD work using the LT3045 , or did it just not sound tonally balanced ?

Obviously,  results will differ from System to System.

[lmitche]

4 hours ago, TJHUB said:

 

Interesting...I thought I was alone.

What devices did you power with the lt3045s?  Now we're they powers?  I am more than thrilled with the results here on USB things with clocks, like USB pcie cards, Regens, and USB powered dacs.

[TJHUB]

56 minutes ago, lmitche said:

What devices did you power with the lt3045s?  Now we're they powers?  I am more than thrilled with the results here on USB things with clocks, like USB pcie cards, Regens, and USB powered dacs.

 

I have a 7v, 6v, and a 5v.  I tried them on the SSD, and on my Pro3z SPIDF converter.  Both sounded equally bad.  

[hurka]

14 minutes ago, TJHUB said:

 

I have a 7v, 6v, and a 5v.  I tried them on the SSD, and on my Pro3z SPIDF converter.  Both sounded equally bad.  

feed with lps1 or sps500?

[hieukm]

7 hours ago, TJHUB said:

 

Sorry if I wasn’t clear, but the SSD was powered the same exact way.  I was trying to convey that the ONLY change was from SSD to HDD.

 

I am a bit thick here but might i ask how you power an external SSD with an LPS?

 

Also anyone try testing if external USB SSD connect to a PCI card powered by LPS made a difference?