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

[Tatomek7]

13 minutes ago, HumanMedia said:

What is your favourite SFP from your list?

I will revise my preferences, as I’ve changed the LPSU for Buffalos. It has only 48h, so sound is changing. With HDPlex and Keces P3 Finisar 1321 sounded dark (with AD1 tubes in DAC), with new LPSU I have to wait a week more,  to have more reliable opinion.

[Tatomek7]

27 minutes ago, nonesup said:

Mine arrived on Thursday.

Congratulations 😗 Consider to apply footers under LPSU, could surprise you, really.

[elan120]

 

This is a short update on my earlier post about adding one more PF Buffalo switch in series.

 

I had some time over a week ago to finish modifying my second Buffalo switch, after close to 200 hours of burn-in, the result is very much aligned with others that have done this before me.  Not only are all of the improvements with just one PF Buffalo still there, it now adds even more of the clarity, dynamic, and air between the instruments.  Soundstage gets to a new level, the music flows even more enjoyable.  After a while, I just listen to the music instead of comparing, and I start to lose some motivation making more changes.

 

In addition to the modifications I did for the first switch, I changed the mounting for Pink Faun clock board from hard mount to soft mount by using 4 of the vibration damping standoffs, added vibration damping sheets to the chassis, and bypass capacitors at the main board DC input.  I will also be experimenting with adding EMI/RFI absorbing sheets to see if it will help.

 

Below are pictures showing the soft mount and finished switch.

 

Soft mount standoffs:

 

 

 

Finished switch overview:

 

 

 

 

[Superdad]

1 hour ago, elan120 said:

Below are pictures showing the soft mount and finished switch.

Very clean and tidy work! (Looks nicer than some commercial products we have seen. )

[cool_chris]

2 hours ago, elan120 said:

 

This is a short update on my earlier post about adding one more PF Buffalo switch in series.

 

I had some time over a week ago to finish modifying my second Buffalo switch, after close to 200 hours of burn-in, the result is very much aligned with others that have done this before me.  Not only are all of the improvements with just one PF Buffalo still there, it now adds even more of the clarity, dynamic, and air between the instruments.  Soundstage gets to a new level, the music flows even more enjoyable.  After a while, I just listen to the music instead of comparing, and I start to lose some motivation making more changes.

 

In addition to the modifications I did for the first switch, I changed the mounting for Pink Faun clock board from hard mount to soft mount by using 4 of the vibration damping standoffs, added vibration damping sheets to the chassis, and bypass capacitors at the main board DC input.  I will also be experimenting with adding EMI/RFI absorbing sheets to see if it will help.

 

Below are pictures showing the soft mount and finished switch.

 

Soft mount standoffs:

 

 

 

Finished switch overview:

 

 

 

 

Great !

I have 2 standard 2016 Nenon moded Buffalows and 2 Pink Faun modded 2016 Buffalows .

The only exception is that I don't have any of the multi electrolytic caps you do.

Are those from Nenon ?

How much of a good was that capacitor bank addition to the switch ?

 

[elan120]

1 hour ago, cool_chris said:

Are those from Nenon ?

How much of a good was that capacitor bank addition to the switch ?

Correct, the capacitor bank was acquired from @Nenon

 

My overall impression on the switch is a bit on the edgy side of neutral without the capacitor bank, adding this capacitor bank help remove that small bit of edginess away.

 

[ASRMichael]

3 hours ago, Superdad said:

Very clean and tidy work! (Looks nicer than some commercial products we have seen. )

Like Adrians at TLS! 😂

 

Still works mind you as a NAS running Euphony smb server! 

[FredM]

@ASRMichael Good that you’ve found a purpose for the device. Oh boy, Adrian and TLS.. what a disrespect to its customers and complete waist of time

[sunny_time_99]

12 hours ago, ASRMichael said:

Like Adrians at TLS! 😂

 

Still works mind you as a NAS running Euphony smb server! 

 

Unshielded Clockcable at high frequencies like 24/25mhz and 10-15cm length? OMG!

I hope that there is a good and clean external Supply for the 3 Clocks - otherwise the effect of "better" clocks is not that big.

 

happy listening

 

Sunny

 

imho a unshielded clockcable should not be longer than 4-5cm - with a shielded one, 15-20cm is not a big Problem for the Signal Quality. with unshielded cable you loose as much, as you try to win. only my 2 Cent. in my Setup are al lot of Neutron Stars inside, one in the Buffalo, one in my Netgear Switch, one in my Ideon 3R, 4 in my DAC...

[RickyV]

4 hours ago, sunny_time_99 said:

 

Unshielded Clockcable at high frequencies like 24/25mhz and 10-15cm length? OMG!

I hope that there is a good and clean external Supply for the 3 Clocks - otherwise the effect of "better" clocks is not that big.

 

happy listening

 

Sunny

 

imho a unshielded clockcable should not be longer than 4-5cm - with a shielded one, 15-20cm is not a big Problem for the Signal Quality. with unshielded cable you loose as much, as you try to win. only my 2 Cent. in my Setup are al lot of Neutron Stars inside, one in the Buffalo, one in my Netgear Switch, one in my Ideon 3R, 4 in my DAC...

 

Nice build, can you tell something about the SQ gain. I am thinking of using the same clock. Why don't you use the square wave output? PM me if it is too technical.

[sunny_time_99]

1 hour ago, RickyV said:

 

Nice build, can you tell something about the SQ gain. I am thinking of using the same clock. Why don't you use the square wave output? PM me if it is too technical.

The Sound Quality is not comparable with the stock Buffalo. This is a complete different Switch now. As i said some time before the Buffalo was not my first choise. I have 7 different Switch and there was allways something in the sq that did not please me. after waiting outside of my System for a few months i start to give it a second Chance. i use my new power supply in latest Version and the newclassd.com Neutron star. if you ever try a Neutron star you know how much this perfect clock-concept increases sq. i have modified about 10 -15 different device. ist allways a big step. there is no way back. the clock have many Advantages: Isolator to prevent ground Loops, heater to keep the xo in perfect temperature, adjustibe (for me;-) - so no aging drift ..2 isolated Outputs and many more.  there is only one disadvantage... the  power supply, thats included in the set is not the one i would like to use.

Youre right,the active Output is the better one, if you keep the cabledistance short. i dont want drill in the original buffalo pcb,  so i take the coax-output.  at the Moment the modified buffalo is the best sounding switch i have. there are 2 cisco Switch without clockmod, so in a few weeks i will compare them. i have some additional Neutron stars waiting for a good Job inside of a Switch...

[RickyV]

1 hour ago, sunny_time_99 said:

The Sound Quality is not comparable with the stock Buffalo. This is a complete different Switch now. As i said some time before the Buffalo was not my first choise. I have 7 different Switch and there was allways something in the sq that did not please me. after waiting outside of my System for a few months i start to give it a second Chance. i use my new power supply in latest Version and the newclassd.com Neutron star. if you ever try a Neutron star you know how much this perfect clock-concept increases sq. i have modified about 10 -15 different device. ist allways a big step. there is no way back. the clock have many Advantages: Isolator to prevent ground Loops, heater to keep the xo in perfect temperature, adjustibe (for me;-) - so no aging drift ..2 isolated Outputs and many more.  there is only one disadvantage... the  power supply, thats included in the set is not the one i would like to use.

Youre right,the active Output is the better one, if you keep the cabledistance short. i dont want drill in the original buffalo pcb,  so i take the coax-output.  at the Moment the modified buffalo is the best sounding switch i have. there are 2 cisco Switch without clockmod, so in a few weeks i will compare them. i have some additional Neutron stars waiting for a good Job inside of a Switch...

You could shield the flat cable of the active output.

[barrows]

1 minute ago, RickyV said:

You could shield the flat cable of the active output.

Clock cables should be impedance controlled SMB or u.fl coax cables, using dedicated SMB or micro BNC connectors when possible.  A couple of CM of wire is likely OK, but anything longer should be proper coax with proper terminations.  It is counterproductive to spend $$$$ on a fancy clock, and then connect it by a single wire flying lead.

[ASRMichael]

Not related to above cables post. But I was thinking, if Ferrite material suppresses EMI & RF, why are cable clamps not used on USB/Ethernet/Ground cables, like Entreq. I think it was @barrows that said cable act like antennas. 
 

What would happen if you covered the whole cable length in ferrite sheet? Are there negative here? 
 

Just never seen any DIYers discussing this. 

[RickyV]

1 hour ago, barrows said:

Clock cables should be impedance controlled SMB or u.fl coax cables, using dedicated SMB or micro BNC connectors when possible.  A couple of CM of wire is likely OK, but anything longer should be proper coax with proper terminations.  It is counterproductive to spend $$$$ on a fancy clock, and then connect it by a single wire flying lead.

Just little clarification the coax output of the newclassd clock is a sinewave the second output is square wave and is meant for oscillator replacement. A 3 wire flat cable is supplied by the manufacturer and should be very short.

So what is better sine wave or square wave?

[sunny_time_99]

Every Cable is an antenna for all kind of pollution. That can be hf, lowf, and common mode noise. Only to name three. Ferrite balls have nothing against CM noise

[davide256]

Was there ever a consensus on whether NUC's should be powered by 12 or 19V? I'm running i7NUC + pre amp off the two 12v taps of a JS-2,

thinking about buying a Paul Hynes supply so that both devices have their own supply, can't interact. A 19V supply would rule out

swapping between the two for comparison on the pre-amp side.

[sunny_time_99]

22 minutes ago, RickyV said:

Just little clarification the coax output of the newclassd clock is a sinewave the second output is square wave and is meant for oscillator replacement. A 3 wire flat cable is supplied by the manufacturer and should be very short.

So what is better sine wave or square wave?

The active output is for "square can oscillator replacement" - if you want. It has a voltage input from the circuit you want to feed. So the NS2 and 3 don't need to be set to a special peak to peak voltage like 3.3V or 5V. (for the coaxial output you can choose 1.8V Pk-PK, 3.3V and 5V).the actice output is doing that for you. Both outputs are isolated but its not square wave.

For very short distances and unkown pk-pk voltage and a 4pin Oszillator, use the active output.

The coaxoutput can do the 4pin oscillator replacement also, but its good to check it with an oscilloscope if the signal is exactly from 0-3.3V e.g. Sometimes you need a variable resistor to adjust it, if its not perfect. Lars is showing it on the NSInstall page.

 

 

The impact on the sq is not that big. The active is a bit clearer if you really have 4-5cm distance to the place where the clock was before.

 

With one neutron Star, a 20 Bucks mini fanless PC from Ebay you can build the NUC7i7/Euphony Killing device. Much higher resolution, higher dynamics, better music flow and definition of instruments. The step up in sq is higher than using a Buffalo switch, ER instead of not using it. If someone is interested, i can post it here. I am using that device for over 6 years now. for 2 years nuc/euphony was the better configuration. But with that clockmod, my actual psu and a freeware...i sold all my 3 NUC 7i7 that i got for comparison.

 

Regards sunny

[lmitche]

52 minutes ago, davide256 said:

Was there ever a consensus on whether NUC's should be powered by 12 or 19V? I'm running i7NUC + pre amp off the two 12v taps of a JS-2,

thinking about buying a Paul Hynes supply so that both devices have their own supply, can't interact. A 19V supply would rule out

swapping between the two for comparison on the pre-amp side.

A 19 volt LPS is best for NUC sound. I have a spare PH SR4 19 volt. If you are interested just PM me.

[tgb]

6 hours ago, sunny_time_99 said:

With one neutron Star, a 20 Bucks mini fanless PC from Ebay you can build the NUC7i7/Euphony Killing device. Much higher resolution, higher dynamics, better music flow and definition of instruments. The step up in sq is higher than using a Buffalo switch, ER instead of not using it. If someone is interested, i can post it here

Hi Sunny !

Please do !

Feel free to post & present yours mods ! 👍

 

I use a Neutron Star 2 on the spdif part of my DC/DC converter (Singxer SU-1) : great job => sound is smooth... Crystek clock (although PS is OK) sounds harsh in comparison.

 

Regarding your photo of your Buffalo (above) : 12VDC to power & just the Neutron Star 3 ?

I've tested a triple LPS on my Buffalo : mistake ! I thought it would help the internal clock but finally it does not help much. The Buff SQ stays behind my other source (see sign.)

BRgds

[jbparrish]

Quick question, if I mount an Intel Optane ssd in a housing and connect it by USB to my NUC, i.e. as an external drive, have I lost all of that Optane goodness because of the USB circuitry? Or does the Optane ssd need to be mounted directly on the main board to contribute to a sound quality advantage? Assume quality LPS for external drive.

 

Thanks

[sandyk]

11 minutes ago, jbparrish said:

if I mount an Intel Optane ssd in a housing and connect it by USB to my NUC, i.e. as an external drive, have I lost all of that Optane goodness because of the USB circuitry?

YES !

 https://www.intel.com.au/content/www/au/en/products/memory-storage/solid-state-drives/consumer-ssds/optane-ssd-9-series.html

[davide256]

1 hour ago, jbparrish said:

Quick question, if I mount an Intel Optane ssd in a housing and connect it by USB to my NUC, i.e. as an external drive, have I lost all of that Optane goodness because of the USB circuitry? Or does the Optane ssd need to be mounted directly on the main board to contribute to a sound quality advantage? Assume quality LPS for external drive.

 

Thanks

Optane mounted internally is at PCI express speeds. As close as you can get to RAM memory speed. USB is glacially slow in comparison.

[sunny_time_99]

7 hours ago, tgb said:

I use a Neutron Star 2 on the spdif part of my DC/DC converter (Singxer SU-1) : great job => sound is smooth... Crystek clock (although PS is OK) sounds harsh in comparison.

 

Hi,

the NS will do a great Job there.

crystek is building very good clocks. sometimes i use them or Abracon Clocks on the Neutron Star  (its not easy to do that - you need to remove some Parts from the NSII).

The main thing is, if you can hear harschness in your Setup, the cause is normally not what you think - the clock e.g or the chipset of the device etc. The cause of the Problem is the power supply. if the ripple and noise rate is too high (like 95% of all Audio devices), the Sound will become harsh. Clocks are very sensitive - i think the most sensitive parts in our audio devices. in the most Audio devices the clock power supply is very poor. some mV ripple.... every clock in my Setup is feed by a power supply with about 12-15µV ripple - that means 0,000 012V @100mA - plus the extra Regulation on the Neutron Star...that is, what does change the Sound. (Ripple in this Dimension is hard to measure. you need a times 1000 amplifier to make it visible.)

If you are going to do that, you will hear no harshness in your crystek clock anymore. The Isolator on the NS will also prevent ground Loops. 

Audio designer need more practical lessons and practical knowledge - stop designing your psu on ltspice... the result will  look the same, and sound awful.

 

happy listening

 

sunny

[jean-michel6]

On 11/30/2020 at 8:24 PM, sunny_time_99 said:

 

With one neutron Star, a 20 Bucks mini fanless PC from Ebay you can build the NUC7i7/Euphony Killing device. Much higher resolution, higher dynamics, better music flow and definition of instruments. The step up in sq is higher than using a Buffalo switch, ER instead of not using it. If someone is interested, i can post it here. I am using that device for over 6 years now. for 2 years nuc/euphony was the better configuration. But with that clockmod, my actual psu and a freeware...i sold all my 3 NUC 7i7 that i got for comparison.

Please do , it seems very interesting!