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

[Johnseye]

5 hours ago, Alan_J said:

Thank you for your valuable report, Johnseye.

Am I correct in understanding that the external power into tX-USBexp actually powers tX-USBexp, sCLK-EX, and tX-USBultra USB module?

If so, can you test if LPS-1 into tX-USBexp can power the three (tX-USBexp, sCLK-EX, and tX-USBultra USB module)?

It will be greatly appreciated.

 

You are correct, that's how I have it today and it must be 9v because the tX-USBexp can only pass 9v along from it's internal connector.  I could have the tX-USBexp run at 12v from an internal molex power input, but it will not pass along that power source through its onboard connector to the sCLK-EX.

 

The way I have things setup right now my LPS-1 is powering my hard drive.  I can test what you asked, but will need to open things up.  I'm on vacation now so it might be a week or two.

[esimms86]

Question. I have an sms200Ultra which, of course, has an sCLK-EX Clock. I don’t have the switch mod or the master connector, nor do I have the IsoRegen or the tXUsbUltra. Given the set up listed above, is there any benefit to be realized by adding the upcoming SOtM Reference Clock? Thanks.

 

 

[austinpop]

19 hours ago, d_elm said:

@Austinpop, I think you tried the Netgear GS10? switch before the Zyxel in your system and found an improvement in SQ.  Would you have time to replace the Zyxel with the Netgear and power the latter at 7V with an LPS-1 ?  I am wondering if the improved magnetics in the Netgear eliminate the need for any switch mods.

 

Short answer: yes the switch mods do make a difference, even with a GS105 upstream, properly grounded, and powered by a good LPS.

 

Long answer: I have been powering the GS105 with an el cheapo Chinese LPS set at 12v. For this experiment, I replaced the el cheapo with a VR Mini (set at 9V), still using the grounding shunt. Also, several months ago, I had posted my finding that powering my modded switch and  modded sMS-200 with the same LPS-1 using a Y-cable had almost the same SQ as separate LPS-1s. With that background, I compared:

1. baseline config:
   • GS105 (VR Mini) > modded switch (shared LPS-1) > modded sMS-200 (shared LPS-1) > ISO-R (LPS-1) > tX-USBultra (LPS-1) > DAC
2. No modded switch config:
   • GS105 (VR Mini) > modded sMS-200 (shared LPS-1) > ISO-R (LPS-1) > tX-USBultra (LPS-1) > DAC

The baseline config was very noticeably better, in all the ways previously described.

 

This is not surprising to me, as the Cybershaft-quality clocking in the switch is clearly beneficial - a finding that has been consistent for many of us. I regard the shunted GS105 upstream as a very effective isolator, and the modded switch as a regenerator.

[austinpop]

3 hours ago, esimms86 said:

Question. I have an sms200Ultra which, of course, has an sCLK-EX Clock. I don’t have the switch mod or the master connector, nor do I have the IsoRegen or the tXUsbUltra. Given the set up listed above, is there any benefit to be realized by adding the upcoming SOtM Reference Clock? Thanks.

 

You cannot exploit the upcoming SOtM Reference Clock unless you have a master clock input on sMS-200ultra.

[austinpop]

@Johnseye, @afrancois, et al. who have dabbled with the 3M or  SOtM EMI/RF sheets.

 

Do you put these on the inside of the chassis - of whatever, server, sMS, tX, etc? Or do you open each component up, and cover individual components?  Also, is there a use case for using these on the outside surfaces of a component?

 

I'm not sure I want to open up any trifecta components, but just gathering info!

[auricgoldfinger]

35 minutes ago, austinpop said:

@Johnseye, @afrancois, et al. who have dabbled with the 3M or  SOtM EMI/RF sheets.

 

Do you put these on the inside of the chassis - of whatever, server, sMS, tX, etc? Or do you open each component up, and cover individual components?  Also, is there a use case for using these on the outside surfaces of a component?

 

I'm not sure I want to open up any trifecta components, but just gathering info!

 

 

 

The photo shows where to place the eABS-200 on the upper side of the chassis inside the sMS-200ultra and tX-USBultra.  May sent me this photo so that I would understand the correct placement.  You need a T8 torx screwdriver to open the case.

 

It's actually pretty simple to add to the tX-USBultra.  The sMS-200ultra is more challenging because of the difficulty of removing and replacing both boards.  The sMS-200 may be simpler since it lacks the sCLK-EX.

 

I heard a discernible improvement after adding the eABS-200 to the tX-USBultra.  My sMS-200ultra came with it already installed, so I have no basis for comparison.

[d_elm]

39 minutes ago, austinpop said:

 

Short answer: yes the switch mods do make a difference, even with a GS105 upstream, properly grounded, and powered by a good LPS.

 

Long answer: I have been powering the GS105 with an el cheapo Chinese LPS set at 12v. For this experiment, I replaced the el cheapo with a VR Mini (set at 9V), still using the grounding shunt. Also, several months ago, I had posted my finding that powering my modded switch and  modded sMS-200 with the same LPS-1 using a Y-cable had almost the same SQ as separate LPS-1s. With that background, I compared:

1. baseline config:
   • GS105 (VR Mini) > modded switch (shared LPS-1) > modded sMS-200 (shared LPS-1) > ISO-R (LPS-1) > tX-USBultra (LPS-1) > DAC
2. No modded switch config:
   • GS105 (VR Mini) > modded sMS-200 (shared LPS-1) > ISO-R (LPS-1) > tX-USBultra (LPS-1) > DAC

The baseline config was very noticeably better, in all the ways previously described.

 

This is not surprising to me, as the Cybershaft-quality clocking in the switch is clearly beneficial - a finding that has been consistent for many of us. I regard the shunted GS105 upstream as a very effective isolator, and the modded switch as a regenerator.

Thank you @austinpop for trying the Netgear switch only.  Clear now that I should get my FS105v2 moded.

[esimms86]

2 hours ago, austinpop said:

 

You cannot exploit the upcoming SOtM Reference Clock unless you have a master clock input on sMS-200ultra.

Thanks for your reply @austinpop. I get that it would require the master connector. What I’m questioning is whether or not it would be worth my while to get the master connector added and then add the Reference Clock given that the sms200Ultra already has the sCLK-EX clock.

[austinpop]

44 minutes ago, esimms86 said:

Thanks for your reply @austinpop. I get that it would require the master connector. What I’m questioning is whether or not it would be worth my while to get the master connector added and then add the Reference Clock given that the sms200Ultra already has the sCLK-EX clock.

 

That's a tough one. Define "worth my while!"

 

Starting with a base sMS-200ultra, each of these additions improves SQ:

• modded switch
• tX-USBultra (or modded tX-USBhubEX)
• reference clock.

Of course, each step costs money. My suggestion is to go back to the index and search the thread for people's experiences. Ultimately, only you can decide if it is worth your while.

 

For me, I have added all of these, and absolutely cannot go back!

[Johnseye]

3 hours ago, austinpop said:

 

Short answer: yes the switch mods do make a difference, even with a GS105 upstream, properly grounded, and powered by a good LPS.

 

Long answer: I have been powering the GS105 with an el cheapo Chinese LPS set at 12v. For this experiment, I replaced the el cheapo with a VR Mini (set at 9V), still using the grounding shunt. Also, several months ago, I had posted my finding that powering my modded switch and  modded sMS-200 with the same LPS-1 using a Y-cable had almost the same SQ as separate LPS-1s. With that background, I compared:

1. baseline config:
   • GS105 (VR Mini) > modded switch (shared LPS-1) > modded sMS-200 (shared LPS-1) > ISO-R (LPS-1) > tX-USBultra (LPS-1) > DAC
2. No modded switch config:
   • GS105 (VR Mini) > modded sMS-200 (shared LPS-1) > ISO-R (LPS-1) > tX-USBultra (LPS-1) > DAC

The baseline config was very noticeably better, in all the ways previously described.

 

This is not surprising to me, as the Cybershaft-quality clocking in the switch is clearly beneficial - a finding that has been consistent for many of us. I regard the shunted GS105 upstream as a very effective isolator, and the modded switch as a regenerator.

 

Based on your last paragraph I assume you used your Cybershaft master clock in all these testing scenarios. Did you happen to test without the Cybershaft in line? 

[Johnseye]

3 hours ago, austinpop said:

@Johnseye, @afrancois, et al. who have dabbled with the 3M or  SOtM EMI/RF sheets.

 

Do you put these on the inside of the chassis - of whatever, server, sMS, tX, etc? Or do you open each component up, and cover individual components?  Also, is there a use case for using these on the outside surfaces of a component?

 

I'm not sure I want to open up any trifecta components, but just gathering info!

 

3 hours ago, auricgoldfinger said:

 

 

 

The photo shows where to place the eABS-200 on the upper side of the chassis inside the sMS-200ultra and tX-USBultra.  May sent me this photo so that I would understand the correct placement.  You need a T8 torx screwdriver to open the case.

 

It's actually pretty simple to add to the tX-USBultra.  The sMS-200ultra is more challenging because of the difficulty of removing and replacing both boards.  The sMS-200 may be simpler since it lacks the sCLK-EX.

 

I heard a discernible improvement after adding the eABS-200 to the tX-USBultra.  My sMS-200ultra came with it already installed, so I have no basis for comparison.

 

I put the stuff anywhere there isn't a vent. I've not found any negative impact, only positive.  Only to your wallet. In @auricgoldfinger's picture I put it in the middle using tweezers. 

[esimms86]

For me, “worth my while “ refers to a “WOW experience,” as opposed to an incremental improvement. I suppose I will have to wait on reviews of the yet to be released Reference Clock before making my decision.

[Johnseye]

30 minutes ago, esimms86 said:

For me, “worth my while “ refers to a “WOW experience,” as opposed to an incremental improvement. I suppose I will have to wait on reviews of the yet to be released Reference Clock before making my decision.

 

That's relative. For me, the SR7 and sCLK-ex were wow experiences but still not as impact full as my speakers. Maybe as impact full as my preamp. 

 

[austinpop]

2 hours ago, esimms86 said:

For me, “worth my while “ refers to a “WOW experience,” as opposed to an incremental improvement. I suppose I will have to wait on reviews of the yet to be released Reference Clock before making my decision.

 

I think you’ll get a sense from reading @romaz‘s impressions with the Ref 10, @hols‘s and mine with the Cybershaft, and others I’m probably forgetting.

 

Go back to the index and look.

[austinpop]

2 hours ago, Johnseye said:

 

Based on your last paragraph I assume you used your Cybershaft master clock in all these testing scenarios. Did you happen to test without the Cybershaft in line? 

 

You're killing me here! Once you have a reference clock like that in the path, there is no way you want to remove it.

 

But I think the result I reported would still hold. Even the sCLK-EX-clocking on the switch improves the SQ substantially.

[Johnseye]

49 minutes ago, austinpop said:

 

You're killing me here! Once you have a reference clock like that in the path, there is no way you want to remove it.

 

But I think the result I reported would still hold. Even the sCLK-EX-clocking on the switch improves the SQ substantially.

 

I had to ask and will take your word for it. 

 

[tims]

Not sure if this has mentioned here before but the Audiobacon website has a new comprehensive review of the sMS-200Ultra. 

They compare different power supplies (the SOtM MBPS-d2s,  Uptone Ultracap LPS1 and Paul Hynes SR7) and also SQ differences between the sMS-200Ultra and the Sonore UltraRendu.

 

What I found interesting is the comparisons between the sMS-200Ultra,  a stock sMS-200 and an sMS-200 that has been externally super clocked:

 

"Comparisons

vs. SOtM sMS-200
The sMS-200ultra is a completely different beast versus the original. Surprisingly, the venerable sMS-200 sounds soft and confused in comparison. It’s just not as refined, articulate, or dynamic. The sMS-200ultra is quite a few shades more natural and organic sounding with a more proper density to voices and all types of instruments. I didn’t suspect including a super clock would be this transformative. The difference is truly astonishing.

vs. SOtM sMS-200 (Externally clocked)
If you’re curious, the tone of the original internally clocked sMS-200 versus its externally clocked counterpart is virtually the same. The difference is better timbre, dimensionality, blacker background, better contouring/layering, pinpoint imaging, and an overall realism to the music.
That said, I was surprised at sonic variance between the externally super clocked sMS-200 and the internally super clocked sMS-200ultra. They’re both being clocked by the same sCLK-EX but the sMS-200 uses the super clock residing in the tX-USBultra via two 6″ clock cables. So technically the only difference is the clock cables and the fact that the sMS-200 is synced with the switch and tX-USBultra.
The externally clocked sMS-200 has a lower noise floor, is airier, and just has more proper phase timing. Soundstage is also deeper and wider. The isolated sMS-200ultra, on the other hand, was supremely rich and euphoric while still maintaining a good amount of dimensionality.
I really enjoyed the tone and timbre of the sMS-200ultra, even though its clock wasn’t synced with the tX-USBultra or modified switch. There was a point where I actually enjoyed the isolated sMS-200ultra more. However, not being synced at the ethernet and USB layers does have its, possibly deal-breaking, implications.

The audible benefits of syncing the clocks across devices:
Lower noise floor
Quicker and proper transients
Much better soundstage depth and width
More precise imaging
Better resolution

In the end, I stuck with the synced setup. Although incredibly musical, not being synced across the chain just doesn’t sound as lifelike. I didn’t find this tradeoff worthwhile after extensive listening. This is the reason why I believe the Ultra Trifecta would be the ultimate combination. With the tX-USBultra and sMS-200ultra synced to an even better master clock…who knows what might happen. Hopefully, I’ll find out soon."

 

sMS-200Ultra Review

 

 

[austinpop]

8 hours ago, d_elm said:

Thank you @austinpop for trying the Netgear switch only.  Clear now that I should get my FS105v2 moded.

 

Yes, knowing what we know now from John’s research, the FS/GS 105/8 switches are the ideal candidates for modding. I personally would pick the GS because unlike the FS it appears to work with an LPS-1 set at 7V, whereas the FS needs at least a 7.5v supply. This is all based on anecdotal evidence from the other threads.

[austinpop]

1 hour ago, tims said:

Not sure if this has mentioned here before but the Audiobacon website has a new comprehensive review of the sMS-200Ultra. 

They compare different power supplies (the SOtM MBPS-d2s,  Uptone Ultracap LPS1 and Paul Hynes SR7) and also SQ differences between the sMS-200Ultra and the Sonore UltraRendu.

 

What I found interesting is the comparisons between the sMS-200Ultra,  a stock sMS-200 and an sMS-200 that has been externally super clocked:

 

"Comparisons

vs. SOtM sMS-200
The sMS-200ultra is a completely different beast versus the original. Surprisingly, the venerable sMS-200 sounds soft and confused in comparison. It’s just not as refined, articulate, or dynamic. The sMS-200ultra is quite a few shades more natural and organic sounding with a more proper density to voices and all types of instruments. I didn’t suspect including a super clock would be this transformative. The difference is truly astonishing.

vs. SOtM sMS-200 (Externally clocked)
If you’re curious, the tone of the original internally clocked sMS-200 versus its externally clocked counterpart is virtually the same. The difference is better timbre, dimensionality, blacker background, better contouring/layering, pinpoint imaging, and an overall realism to the music.
That said, I was surprised at sonic variance between the externally super clocked sMS-200 and the internally super clocked sMS-200ultra. They’re both being clocked by the same sCLK-EX but the sMS-200 uses the super clock residing in the tX-USBultra via two 6″ clock cables. So technically the only difference is the clock cables and the fact that the sMS-200 is synced with the switch and tX-USBultra.
The externally clocked sMS-200 has a lower noise floor, is airier, and just has more proper phase timing. Soundstage is also deeper and wider. The isolated sMS-200ultra, on the other hand, was supremely rich and euphoric while still maintaining a good amount of dimensionality.
I really enjoyed the tone and timbre of the sMS-200ultra, even though its clock wasn’t synced with the tX-USBultra or modified switch. There was a point where I actually enjoyed the isolated sMS-200ultra more. However, not being synced at the ethernet and USB layers does have its, possibly deal-breaking, implications.

The audible benefits of syncing the clocks across devices:
Lower noise floor
Quicker and proper transients
Much better soundstage depth and width
More precise imaging
Better resolution

In the end, I stuck with the synced setup. Although incredibly musical, not being synced across the chain just doesn’t sound as lifelike. I didn’t find this tradeoff worthwhile after extensive listening. This is the reason why I believe the Ultra Trifecta would be the ultimate combination. With the tX-USBultra and sMS-200ultra synced to an even better master clock…who knows what might happen. Hopefully, I’ll find out soon."

 

sMS-200Ultra Review

 

 

 

Yes, his findings were interesting. I have done a similar comparison between my modded sMS-200 and @limniscate‘s sMS-200ultra, but he has the 12V unit, so we have never been able to use the identical PSU. 

 

In our listening comparisons, we were hard pressed to tell the 2 apart in terms of SQ.

 

Maybe once his SR7 or my SR4 arrives, we’ll be able to control this variable.

 

Regardless, I think it is fair to say that this difference is small compared to the overall benefit of the sCLK-EX clock.

[d_elm]

10 minutes ago, austinpop said:

 

Yes, knowing what we know now from John’s research, the FS/GS 105/8 switches are the ideal candidates for modding. I personally would pick the GS because unlike the FS it appears to work with an LPS-1 set at 7V, whereas the FS needs at least a 7.5v supply. This is all based on anecdotal evidence from the other threads.

I have both FS105v3 and the older FS105v2.  Boards are quite different, more LSI on the v3, and the v2 runs at 7v.  Same magnetics in both.

[austinpop]

4 minutes ago, d_elm said:

I have both FS105v3 and the older FS105v2.  Boards are quite different, more LSI on the v3, and the v2 runs at 7v.  Same magnetics in both.

 

Ah yes, I remember now - the older versions accept a lower voltage. Good to go!

[Speed Racer]

16 minutes ago, austinpop said:

 

Yes, knowing what we know now from John’s research, the FS/GS 105/8 switches are the ideal candidates for modding. I personally would pick the GS because unlike the FS it appears to work with an LPS-1 set at 7V, whereas the FS needs at least a 7.5v supply. This is all based on anecdotal evidence from the other threads.

 

The only mod that makes any real sense for these Netgear switches is to ground the DC negative output of the SMPS. Using an LPS seems like an unjustified expenditure. 

[afrancois]

12 hours ago, austinpop said:

@Johnseye, @afrancois, et al. who have dabbled with the 3M or  SOtM EMI/RF sheets.

 

Do you put these on the inside of the chassis - of whatever, server, sMS, tX, etc? Or do you open each component up, and cover individual components?  Also, is there a use case for using these on the outside surfaces of a component?

 

I'm not sure I want to open up any trifecta components, but just gathering info!

I didn't open any SOtM gear. My sMS-200 Ultra and sPS-500 are resting on the 3M paper, as are all 4 LPS-1's. The LPS-1's also have the 3M glued on top. Every LPS-1 has a heat sink on top.

The back of the AQVOX has 3M paper clued on the outside, this is also the case for my pre-amp, power amp and MC3+ USB. The pre-amp, power amp and MC3+ USB also have 3M on the inside of the cover.

The ISO Regen is covered on both sides with 3M. Close to the spaghetti wiring there’s a 3M.

In short I’ve put 3M where I could put it, except on the inside of the SOtM gear.

 

Included are some pictures to give you an idea of my garage setup. Pictures however are quite dark.

[mozes]

4 hours ago, afrancois said:

I didn't open any SOtM gear. My sMS-200 Ultra and sPS-500 are resting on the 3M paper, as are all 4 LPS-1's. The LPS-1's also have the 3M glued on top. Every LPS-1 has a heat sink on top.

The back of the AQVOX has 3M paper clued on the outside, this is also the case for my pre-amp, power amp and MC3+ USB. The pre-amp, power amp and MC3+ USB also have 3M on the inside of the cover.

The ISO Regen is covered on both sides with 3M. Close to the spaghetti wiring there’s a 3M.

In short I’ve put 3M where I could put it, except on the inside of the SOtM gear.

 

Included are some pictures to give you an idea of my garage setup. Pictures however are quite dark.

Your setup is awesome!! I love how neat and tidy it is, unlike mine, all over the place  

Can you pls share a link of where did you get these LPS-1 heat sinks?

[rickca]

8 hours ago, austinpop said:

Yes, knowing what we know now from John’s research, the FS/GS 105/8 switches are the ideal candidates for modding. I personally would pick the GS because unlike the FS it appears to work with an LPS-1 set at 7V, whereas the FS needs at least a 7.5v supply. This is all based on anecdotal evidence from the other threads.

Interesting that the GS305 from Netgear's SOHO line uses 5V.  Unfortunately, we don't know if it blocks high impedance leakage like these other models.