Jump to content

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


austinpop
 Share

Message added by The Computer Audiophile

Important and useful information about this thread

Posting guidelines

History and index of useful posts

Most important: please realize this thread is about bleeding edge experimentation and discovery. No one has The Answer™. If you are not into tweaking, just know that you can have a musically satisfying system without doing any of the nutty things we do here.

Recommended Posts

16 minutes ago, ElviaCaprice said:

Received back my NUC, SCLK EX, tX bexp pcie card from SOtM.   Everything was very well packed and labled, no instructions.   Three of the 4 clock leads were labled 1 thru 3 coming out of the NUC.  I proceeded to hook up everything to the sclk ex board.  Ran 12v power to the sclk ex.  Started up the NUC, just loud beeps, different patterns often ending in a long continous beep, unable to boot.   Checked all the leads, resat the memory styx, pulled the mobo, etc.  The NUC in my opinion is lost.  Emailed SOtM, waiting on engineer reply, time is running out.   

Fortunately the last, 4th clock to the txbexp pcie card works great.  So I quickly modified my pico-itx to accommodate the pcie slot, no optimizations, external usb hdd for library.  Smps on the 2qute and mobo.  Lps1 on the iso regen, sclk ex and txbexp.  Sounds really good,  the txbexp in conjunction with the sclk ex adds great sound stage, detail.  

Paul Hynes ps should be here in two weeks.

 

That's a real bummer. Hope you get it sorted out. 

 

I have to to admit I don't quite understand what your setup is, from your description. Maybe once you have it all working, give us a clearer view. 

 

For example: are the sCLK-EX and the tX-USBexp boards supposed to go into slots in the NUC? Or are we talking 2 different computers here?

 

Finally, I'm only used to the SMB attachments on the end points. Those are labeled with the frequency (24 or 25 MHz). Sounds like your 3 NUC clock leads were not labeled? That's nuts!

Link to comment
Share on other sites

46 minutes ago, austinpop said:

 

That's a real bummer. Hope you get it sorted out. 

 

I have to to admit I don't quite understand what your setup is, from your description. Maybe once you have it all working, give us a clearer view. 

 

For example: are the sCLK-EX and the tX-USBexp boards supposed to go into slots in the NUC? Or are we talking 2 different computers here?

 

Finally, I'm only used to the SMB attachments on the end points. Those are labeled with the frequency (24 or 25 MHz). Sounds like your 3 NUC clock leads were not labeled? That's nuts!

I use a pcie adapter that attatches to the mini pcie slot on the NUC or PICO-ITX motherboards, separate computers.  The adapter allows the pcie 1x lane to reside outside the computer, attatched by a long ribbon cable.  The sCLK-EX sits separate on its own, it's just a card mouted on a metal tray.   The SMB attachments are labled with frequency and clock number corresponding to the sCLK-EX.  Having replaced three clocks on the NUC mobo, the problem could be any one of those clocks.  It's so easy to destroy a mobo when replacing parts on those boards.  I would have assumed SoTM tested the NUC before shipping?  Mislabled clock?  All we can do is try some simple tests on my end.

 

 

1503163145106356789390.jpg

The SCLK EX residesat the top of the photo,  in the middle resides pico-itx with the pcie 1x adapter connected to the txbexp card (they are both upside down), uspcb connects the txbexp card to the iso regen,  another uspcb connects the iso regen to the 2Qute.  Speaker cables from 2qute to the Omegas.

(JRiver) Jetway barebones NUC (mod 3 sCLK-EX, Cybershaft OP 14)  (PH SR7) => mini pcie adapter to PCIe 1X => tXUSBexp PCIe card (mod sCLK-EX) (PH SR7) => (USPCB) Chord DAVE => Omega Super 8XRS/REL t5i  (All powered thru Topaz Isolation Transformer)

Link to comment
Share on other sites

9 minutes ago, ElviaCaprice said:

I use a pcie adapter that attatches to the mini pcie slot on the NUC or PICO-ITX motherboards, separate computers.  The adapter allows the pcie 1x lane to reside outside the computer, attatched by a long ribbon cable.  The sCLK-EX sits separate on its own, it's just a card mouted on a metal tray.   The SMB attachments are labled with frequency and clock number corresponding to the sCLK-EX.  Having replaced three clocks on the NUC mobo, the problem could be any one of those clocks.  It's so easy to destroy a mobo when replacing parts on those boards.  I would have assumed SoTM tested the NUC before shipping?  Mislabled clock?  All we can do is try some simple tests on my end.

Elvia, How are you playing anything without the NUC?

Pareto Audio aka nuckleheadaudio

Link to comment
Share on other sites

17 minutes ago, lmitche said:

Elvia, How are you playing anything without the NUC?

Larry, I am using the pico-itx in replacement.  Obviously minus the three other clocks on the sclk ex.

 

1503163908177-1236906178.jpg

1503163976301675051539.jpg

 

I must say, the s-clk ex (2nd photo) is the key to the magic of the txbexp pcie card (1st photo)

(JRiver) Jetway barebones NUC (mod 3 sCLK-EX, Cybershaft OP 14)  (PH SR7) => mini pcie adapter to PCIe 1X => tXUSBexp PCIe card (mod sCLK-EX) (PH SR7) => (USPCB) Chord DAVE => Omega Super 8XRS/REL t5i  (All powered thru Topaz Isolation Transformer)

Link to comment
Share on other sites

49 minutes ago, ElviaCaprice said:

Larry, I am using the pico-itx in replacement.  Obviously minus the three other clocks on the sclk ex.

 

1503163908177-1236906178.jpg

1503163976301675051539.jpg

 

I must say, the s-clk ex (2nd photo) is the key to the magic of the txbexp pcie card (1st photo)

Got it! LOL, my eyesight is changing.  I read that you were were using a pico-atx and I was wondering how a power supply made music. Your m.2 pcie cable solution is very clever.

Pareto Audio aka nuckleheadaudio

Link to comment
Share on other sites

@austinpop PC mod shops carry heatsinks. In US you can look at FrozenCPU for example, they have hundreds of heatsinks of all sizes.

 

I would use a thermal pad. Thermal paste would be even better but it is messy and doesn't hold the heatsink in place.

 

 @Ciukas  - thermal pads do increase the thermal conductivity. Simply placing the heatsink on top of a surface is not efficient. Thermal pads are actually quite efficient conductors and their conductivity increases with temperature. In order to get good conductivity only by contact, the surface finish must be very good and joined with pressure. Thermal paste is even better. 

 

Link to comment
Share on other sites

3 hours ago, MikeyFresh said:

 

I use two of these sourced from eBay, took only 4 days from Hong Kong to NY. That same seller has many other available options.

 

Not the prettiest, but very effective, the LPS-1 is only warm to the touch now powering the combo of a microRendu and bus powered DAC:

 

 

PA290896.jpg

 

Nice one MikeyFresh! I personally think it looks awesome! ?

Auralic Aries Mini, Ariston RD40 & Mytek Brooklyn DAC system with Fostex TH900 & Gibson Les Paul 8 reference monitors.                           

Learn from the mistakes of others. You can never live long enough to do them all yourself. Graucho Marx

Link to comment
Share on other sites

3 hours ago, MikeyFresh said:

 

I use two of these sourced from eBay, took only 4 days from Hong Kong to NY. That same seller has many other available options.

 

Not the prettiest, but very effective, the LPS-1 is only warm to the touch now powering the combo of a microRendu and bus powered DAC:

 

 

PA290896.jpg

 

1 hour ago, unbalanced output said:

@austinpop PC mod shops carry heatsinks. In US you can look at FrozenCPU for example, they have hundreds of heatsinks of all sizes.

 

I would use a thermal pad. Thermal paste would be even better but it is messy and doesn't hold the heatsink in place.

 

 @Ciukas  - thermal pads do increase the thermal conductivity. Simply placing the heatsink on top of a surface is not efficient. Thermal pads are actually quite efficient conductors and their conductivity increases with temperature. In order to get good conductivity only by contact, the surface finish must be very good and joined with pressure. Thermal paste is even better. 

 

 

1 hour ago, George Hincapie said:

Thermal cement might be a better, but permenant, choice. I believe Artic Silver make such a product. 

 

Thanks for all the info.

 

I hear you on the thermal efficacy of pads/cement etc, but for now, I'll be happy if just surface contact reduces the temperature, even if not in the most efficient way.

 

I have this butt-ugly one on order for now, just to see how it does: https://www.amazon.com/gp/product/B073QS338W

 

I'll take before and after thermal readings to see how well it does.

Link to comment
Share on other sites

8 hours ago, ElviaCaprice said:

Larry, I am using the pico-itx in replacement.  Obviously minus the three other clocks on the sclk ex.

 

1503163908177-1236906178.jpg

1503163976301675051539.jpg

 

I must say, the s-clk ex (2nd photo) is the key to the magic of the txbexp pcie card (1st photo)

To me this is very interesting, it seems like the ( s-clk ex clocked) txbexp card is an improvement, even though it is earlier in the signal path than the iso regen ... ?

Link to comment
Share on other sites

@unbalanced output  Again, I mentioned thermal pads are useful if a heatsink does not sit perfectly flush w a component as you've also reiterated in your post. I do not think the LPS-1's scenario requires such extreme measures. Thermal paste is not applicable in this scenario as well so point moot. Why not include liquid metal while we're at it?

 

Thermal pad efficiency increases as temperature rises, yes, but the best commercial pads only have thermal ratings up to 17w/mK (the 17s gets brittle and dry after a while). In contrast, aluminum has a rating of 250w/mK. There is no contest. Anyway, I do not wish to harp on this any further as this is an audio forum and not a overclocker's sanctuary. 

Link to comment
Share on other sites

2 hours ago, motberg said:

To me this is very interesting, it seems like the ( s-clk ex clocked) txbexp card is an improvement, even though it is earlier in the signal path than the iso regen ... ?

That is correct.  Others have found the tx ultra also better preceding the iso regen.  What I like better about the txusbexp is that it is friendly for using the USPCB in the stream, preceding the ISO Regen.  I'm sure if the NUC would have worked, that I would have found even better SQ and these clocks a prime candidate for a master clock upgrade.

That being said, I'm quite impressed with what the single SOtM component, txusbexp, has brought in improved SQ.in conjunction with the ISO Regen.   I'm sure the Paul Hynes SR7 will solidify that once received.  Then the server, 2qute and sclk ex will each get a clean quality 12V supply.

 

 

 

(JRiver) Jetway barebones NUC (mod 3 sCLK-EX, Cybershaft OP 14)  (PH SR7) => mini pcie adapter to PCIe 1X => tXUSBexp PCIe card (mod sCLK-EX) (PH SR7) => (USPCB) Chord DAVE => Omega Super 8XRS/REL t5i  (All powered thru Topaz Isolation Transformer)

Link to comment
Share on other sites

6 hours ago, Ciukas said:

@unbalanced output  Again, I mentioned thermal pads are useful if a heatsink does not sit perfectly flush w a component as you've also reiterated in your post. I do not think the LPS-1's scenario requires such extreme measures. Thermal paste is not applicable in this scenario as well so point moot. Why not include liquid metal while we're at it?

 

Thermal pad efficiency increases as temperature rises, yes, but the best commercial pads only have thermal ratings up to 17w/mK (the 17s gets brittle and dry after a while). In contrast, aluminum has a rating of 250w/mK. There is no contest. Anyway, I do not wish to harp on this any further as this is an audio forum and not a overclocker's sanctuary. 

 

You cannot get the same thermal conductivity of pure aluminium in an interface. Since on top of that the surface is anodised, you have to divide that number by 20 or 30 depending on the thickness, then reduce it again by a factor due to surface finish. So yeah, just sitting the heat sink on top of a device is pretty much a waste, best case scenario you're getting half of the heat transfer a cheap thermal pad would give. Sitting the LPS1 on its side will probably be more efficient than that (good suggestion by Superdad btw).

 

The LPS1 runs just warm for me since my dac uses its own power supply. 

Link to comment
Share on other sites

7 hours ago, Ciukas said:

@unbalanced output  Again, I mentioned thermal pads are useful if a heatsink does not sit perfectly flush w a component as you've also reiterated in your post. I do not think the LPS-1's scenario requires such extreme measures. Thermal paste is not applicable in this scenario as well so point moot. Why not include liquid metal while we're at it?

 

Thermal pad efficiency increases as temperature rises, yes, but the best commercial pads only have thermal ratings up to 17w/mK (the 17s gets brittle and dry after a while). In contrast, aluminum has a rating of 250w/mK. There is no contest. Anyway, I do not wish to harp on this any further as this is an audio forum and not a overclocker's sanctuary. 

 

Thermal pads are a poor choice in my experience; they are too thick (you want only a tiny amount of transfer material in the contact area) and are not made of the most efficent material. PC Overclockers have used silver based compounds for over a decade for a reason.

 

I don't accept your final comment either; the disciplines are overlapping more all the time and 'audiophiles' should always seek to use knowledge from the PC world to their advantage.

Link to comment
Share on other sites

@unbalanced output  Point taken. Let's call it a difference of opinions. The important thing is that Austin got his fix :)

 

@George Hincapie  Not sure why you quoted me? That's what I've been saying all along about pads, you're preaching to the choir. Also, you misunderstand me - my final comment was meant to discourage derailing this thread as the matter was already resolved and not as you indicated to spurn knowledge from other 'disciplines'.

Link to comment
Share on other sites

42 minutes ago, Ciukas said:

@unbalanced output  Point taken. Let's call it a difference of opinions. The important thing is that Austin got his fix :)

 

@George Hincapie  Not sure why you quoted me? That's what I've been saying all along about pads, you're preaching to the choir. Also, you misunderstand me - my final comment was meant to discourage derailing this thread as the matter was already resolved and not as you indicated to spurn knowledge from other 'disciplines'.

 

The thread wasn't being derailed; it's all relevant. If not to the OP then to others reading. Regardless, let's move on.

Link to comment
Share on other sites

This is going to be a long 2-3 weeks.

 

@Bamber's ISO-Regen is on its way back. My Ultra trifecta is on its way to May to be retrofitted with the SMB connectors, and a master reference clock input on the tX-USBultra.

 

In the meantime, I am using "just" this chain:

- Roon core bridged > mR > Lush USB > DAC.

 

This was state of the art just a year ago. Now - wow, how much further we've come!

 

Hope @Superdad hits his Aug 24 target for my own ISO-R that's on order, so I can at least get some of the magic back.

Link to comment
Share on other sites

Has anyone with a 12v sMS-200ultra tried the 6.5V~8.4v setting in this option table on their sCLK-EX board?

 

image.thumb.png.b0bc93a938d50fd15d93a6e0e6f56bd4.png

SOtM sMS-200ultra (12v, silver internal DC cable, 75Ω external clock input),

SOtM tX-USBultra (12v, silver internal DC cable, 75Ω external clock input),

Singxer SU-1 (7v SOtM sCLK-EX, silver internal DC cable, 75Ω external clock input),

W4S DAC2v2 SE (SOtM sCLK-EX silver internal DC cable, 75Ω external clock input),

Mutec REF 10, Bryston BP6, Trinnov Altitude 32, 9 x Bryston 7B³s, B&W 800D3s, B&W HTM1 D3, 2 x Arendal Sub 3,

SOtM dCBL-CAT7 (ISO-CAT6 SE), SOtM eABS-200 installed on all components, JCAT NET Card FEMTO, mRendu, dCS Rossini DAC, Chord Hugo 2, Schitt Yggy, Chord Qutest (ordered for home office), UpTone ISO Regen, 5 x UpTone LPS-1s, 4 x PH SR4s, 5 x UpTone LPS-1.2s

Link to comment
Share on other sites

15 hours ago, austinpop said:

@ElviaCaprice

 

Just curious - what frequency(ies) are the 3 mobo sCLK taps set at?

 

I only know what they (SOtM) use for USB (24MHz) and Ethernet (25MHz) controllers, but I was wondering what frequencies mobo mods need.

 

Again - just curious. Hoping your NUC issues get resolved.

Hi Rajiv,

The 3 clocks are as follows for the NUC,

12MHz 

25MHz

25MHz

I wouldn't know what each clock corresponds to on the mobo.  From past experience, usually the main system clock is 25MHz.   Why the repeat of two clocks at 25MHz, when both clocks on the mobo could have been fed from the same SCLK EX clock?  Different voltage?

I did request NOT to replace any ethernet or usb clocks on the mobo since with this mobo I would use the single sata port only.

I also requested that if one clock was left open to make it ready for my own DIY of the iso regen.  Obviously that did not happen.

There has been no contact concerning the NUC from SOtM.  Considering all the different beep signal patterns I get on powering up the NUC, I can only assume that the mobo is lost.  Now it's a question of options to rectify the situation.  Unfortunately my time is limited here in the states before I depart to my home in Costa Rica.   There it can be impossible to receive and send items in safe, affordable fashion.

 

I await their aknowledgment and solution/options. 

 

(JRiver) Jetway barebones NUC (mod 3 sCLK-EX, Cybershaft OP 14)  (PH SR7) => mini pcie adapter to PCIe 1X => tXUSBexp PCIe card (mod sCLK-EX) (PH SR7) => (USPCB) Chord DAVE => Omega Super 8XRS/REL t5i  (All powered thru Topaz Isolation Transformer)

Link to comment
Share on other sites

i am also interested in replaceing the clock on my mobo but how can i measure the necessary voltage? Do i need a oscilloscope for that or can i do it with a multimeter?

Meitner ma1 v2 dac,  Sovereign preamp and power amp,

DIY speakers with scan speak illuminator drivers.

Raal SR1a

Under development:

NUC7i7dnbe, Euphony Stylus, Qobuz.

Modded Buffalo-fiber-EtherRegen, DC3- Isoregen, Lush^2

Link to comment
Share on other sites

8 hours ago, ElviaCaprice said:

Hi Rajiv,

The 3 clocks are as follows for the NUC,

12MHz 

25MHz

25MHz

I wouldn't know what each clock corresponds to on the mobo.  From past experience, usually the main system clock is 25MHz.   Why the repeat of two clocks at 25MHz, when both clocks on the mobo could have been fed from the same SCLK EX clock?  Different voltage?

I did request NOT to replace any ethernet or usb clocks on the mobo since with this mobo I would use the single sata port only.

I also requested that if one clock was left open to make it ready for my own DIY of the iso regen.  Obviously that did not happen.

There has been no contact concerning the NUC from SOtM.  Considering all the different beep signal patterns I get on powering up the NUC, I can only assume that the mobo is lost.  Now it's a question of options to rectify the situation.  Unfortunately my time is limited here in the states before I depart to my home in Costa Rica.   There it can be impossible to receive and send items in safe, affordable fashion.

 

I await their aknowledgment and solution/options. 

 

 

I do hope you get this resolved.

 

The 25MHz frequency for the main system clock makes sense, as I know the sMS-200 requires a 24MHz clock (for the USB controller) and a 25MHz clock (for the ethernet controller, and presumably, for the the main system clock).

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share



×
×
  • Create New...