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

[RickyV]

Hi all,

I have been mailing with May from Sotm. She said that it is possible to replace a crystek 575 clock with one of the sCLK-ex tabs given the correct voltage and frequency. 

Thats all the info they need for all xo's you want to replace, frequency and voltage.

 

[seeteeyou]

Interesting to see how the owners of Singxer SU-1 were replacing regulators with much better ones

 

https://www.computeraudiophile.com/forums/topic/29553-my-very-mini-review-of-the-singxer-su-1-ddc/

 

Could we also reap the benefits if we're willing to void the warranty of other audiophile products?

 

http://www.ldovr.com/category-s/118.htm

https://www.belleson.com/store/index.php

http://sparkoslabs.com/discrete-voltage-regulators/

 

Sometimes it's such a pity to feed a motherboard with DC input while spending quite a bit on a decent linear PSU, then all that nice power would be kinda "ruined" by lousy regulators.

 

It's so much easier to bypass the regulators if something were designed just like RATOC Audio Lab RAL-KCM3MB01 since we could actually feed 1.8V and 3.3V separately

 

http://www.ratocaudiolab.com/product/kit/ral_kcm3mb01/index.html

 

[romaz]

There are simply not enough hours in the day these days for me to have much of an online presence and unfortunately, that will have to continue for the foreseeable future.  Consequently, I haven’t kept up with much that has been going on although presently, as I am on a plane flying home (it seems flying on planes is the only time I can find free time these days), I took liberty of my rare opportunity to catch up.  Lots of interesting posts, especially regarding clocking both here and on the REF10 thread.

 

First of all, I agree that this thread has become very large and diverse, enough so that it has become difficult to navigate through it or to find valuable posts made by many since the thread’s inception on Jan 2.  I plead guilty to not being a very good curator or moderator and am also largely responsible for the many detours this thread has taken but for some time now, I have looked at this thread as a “think tank” of ideas to improve upon what the sMS-200 and microRendu bring to the table.  In other words, I like that this thread is so diverse.

 

Before this “direct connection” pathway (where you connect your Ethernet endpoint directly to your music server instead of directly to your router) that was made possible by the bridging of two LAN ports, like many, I held the belief that what was upstream of the sMS-200 or microRendu didn’t matter.  For better or worse, what this “direct connection” provided me was a much greater sense of transparency, not just to the quality of the recording but also to the quality of the components upstream of these endpoints and so I felt it was fair game to discuss power supplies, motherboards, CPUs, operating systems, cabling, storage media, clocking, etc., because with this “direct connection” pathway, all of these things suddenly made much more of a difference.

 

Having said all of this, I like what Rajiv is trying to do.  I think that to be able to create some sort of an index or FAQ would be very helpful for all.  Personally, I wish there was also some way to bookmark a post.  Regarding creating an index or FAQ for this thread, even though I started this thread, I do not presently have the time to do it although I would be happy to relinquish this authority to anyone who would care to take it on.  Frankly, I would nominate someone like Rajiv who has been a part of this thread from the beginning and as he has implemented many of the things discussed on this thread, he also has firsthand knowledge of their impact.  Just as important, I think he has all of our respect. 

 

With the limited time I have on hand, I thought I’d address some of the clocking comments made by some, including the article written by Hugh Robjohns that was referenced by @Johnseye.  I am in agreement with some of what Hugh wrote but I disagree with his conclusion.  I am NOT a clocking expert and I have never claimed to be one but I have heard enough in the recent months to be able to state that my observations regarding clocking are definitely in contrast to some of these claims made by supposed experts.  To quote Duke Ellington, “if it sounds good, it is good.”  Putting my own spin on it, “if it sounds better, it is better…..theories and measurements be damned.”  

 

As many who have followed this thread from the beginning know, I was surprised by the impact that a Paul Pang reclocking switch made in my system.  This is a switch that I bought used from another CA member for about $100 that incorporates a TCXO clock.  When I first implemented this switch after my router and just before my music server, it made a small difference but hardly worth the $100 that I paid for it, even with the switch powered by an LPS-1.  However, when I placed this switch in the more revealing “direct pathway” between my music server and my sMS-200, I was quite startled by its much greater impact with regards to air and soundstage.  Within this direct path and implemented after my noisy server and much closer to my DAC, this $100 switch was now a real bargain.  Interpret it any way you wish but it was clear to me that where in your chain you position a well-clocked component matters.  With further testing, my observations further showed me that a good clock was more likely to reveal its qualities when followed by a clock of equal or greater performance and was more likely to be blunted or negated if followed by a clock of lesser performance.  In other words, it would be ideal if all the clocks in your chain were of an equivalent high standard and if this isn't possible, the clocks in your chain at the very least should progressively get better and better where your worst clock is at the beginning of your chain and your best clock is at the end of your chain and closest to your DAC.

  

Around this time, Chan King Girand Michel (owner of Pachanko cables), who was making for me some specialized cables commented on how his new sCLK-EX clock board was making his Paul Pang OCXO clocks sound like child’s toys and so, of course, my interest was piqued.  This clock board had been officially released by SOtM some months before but had not yet made its way into any of their endpoints including the sMS-200 although SOtM was very game to sell me their sCLK-EX clock board and incorporate it into their dX-USB HD, a USB-to-SPDIF converter that I neither needed nor wanted.  Because it was their only vehicle at the time that could accommodate the large size of their new clock board, I went with it and used it to also clock my sMS-200.  Because I had 1 free clock remaining, on a whim, I also asked SOtM to replace the clock on a cheap $20 TrendNet switch I had laying around and this move proved to be fortuitous because it allowed me to assess the step changes brought about by a series of synchronously clocked components.

 

As I reported, the cumulative impact of this “trifecta” that actually amounts to 4 reclockings was much more than I was expecting.  To say that it was the finest digital front end I had heard to date would be completely accurate and by a very significant margin.  Having spoken with Lee, SOtM’s lead engineer who is responsible for this clock board, the 4 clocks on this board are essentially slaves to an internal master oscillator.  More than just the phase noise properties of this internal “master” clock is the fact that the 4 slave clocks are synchronized, perhaps the strongest argument for using a master clock according to Lee.

 

Prior to Munich, based on the already spectacular sound I was getting from my SOtM stack, based on disparaging  articles such as the one written by Hugh Robjohns that was referenced by @Johnseye, and based on my own lukewarm experience with dCS’s external master clock last year, I had no intention of ever buying an external master clock and I even posted this sentiment back in April.  Of course, part of this sentiment was also based on wishful thinking because I didn’t relish the idea of paying thousands of dollars for a master clock and all the hassle that went with it.  Munich, however, changed all of that.  

 

When I first came upon the REF10 at the Mutec booth, because of the considerable noise at the MOC in Munich, I noted that Mutec elected to use a headphone setup although it was a rather modest setup (a $300 Meze 99 headphone connected to an $1,200 Lehmann Audio Linear D DAC/headphone amp).  This REF10 master clock was clocking Mutec’s own MC-3+USB, their highly regarded USB-to-SPDIF converter but what was strange was they chose to output to the headphone DAC/amp via Toslink.  It would seem to me that if a reasonably minded person wanted to impress me with the performance of their new master clock, they would choose something other than a high-jitter optical connection as this would surely mask some of the performance of the master clock and so my first thought was that this was either a super ballsy move to show just how wickedly good this new clock was or else Mutec would use it as an excuse after I told them I couldn’t hear much difference.  To further stack the odds against it, if someone tells me their new master clock is going to cost about $4,300 and that the 0.5m clock cable being used is going to cost another $800, you can bet that I’m going to have proportionately lofty expectations that will be difficult to meet.

 

I’ll cut to the chase.  This master clock is wickedly good.  In fact, I am still dumbfounded by just how good it sounded.  While I have not had a chance to directly A/B the MC-3+USB against any of SOtM’s Ultra line, I was always impressed by it every time I had heard it in the past and I was impressed with it again when I heard it at the MOC.  Never would I characterize it as sounding coarse, dry, shallow or closed in and yet, when directly compared against the REF10, that is exactly how it sounded.  In fact, the REF10 made the stock clock on the MC-3+USB almost unlistenable.  Over the span of 3 days, I heard many wonderful things in Munich and this audio show, which is the largest of its kind in the world, is truly the place where the best of the best come to gather.  Yet, with all the wonderful things to see and hear, the “Best of Show” for me was this external master clock.

 

For those who are wondering what the REF10 looks like, here is a photo of the REF10 stacked underneath the MC-3+USB:

 

Here is a photo of the internals of the REF10.  There are two 50-ohm outputs and six 75-ohm outputs and each output is galvanically isolated from each other.  Power to the OCXO is supplied by a well thought out internal linear PSU of Christian Peter’s own design (Christian is the CEO of Mutec).  

 

 

Here are photos of the rear of the REF10.  The Habst 0.5M 75-ohm clock cable that is used is extremely heavily shielded and utilizes 5N cryo’d silver as a conductor.  You will notice that it occupies a 2^nd clock output from the REF10 and this serves as an extra ground.  If this type of cable is used, it would cut the number of clock outputs in half.

 

Of course, this begs the question of whether the REF10 would improve SOtM’s sCLK-EX by the same magnitude.  My hunch is that the sCLK-EX utilizes a better clock than the stock clock in the MC-3+USB and that the magnitude of improvement won’t be as large.  At the same time, as I will be utilizing 2 sCLK-EX boards and its 8 clocks, the thought of being able to synchronize all 8 clocks to a higher level of performance (even if the margin might be smaller) was a salivating thought.  I was able to convince Christian to provide me the measurements of his OCXO which I showed to Lee.  While I have the actual plot tracings in my possession, I’m not sure I have the authority to post them and so I won’t but here is what Lee told me.  He said that the REF10’s measurements were probably the best measurements for a 10MHz clock he had ever seen and its performance was better than the sCLK-EX.  Pretty high praise from a competitor clock maker and this was all the validation that I needed to hear.  My order has been placed and so I will be able to decide for myself in a few short weeks just how good this clock is in my system.

 

As I read Hugh Robjohns’ article, “Does Your Studio Need A Digital Master Clock,” I read with interest his concluding statement.  “Overall, it should be clear…that employing an external master clock cannot and will not improve the sound quality of a digital audio system.”  My response?  Don’t listen to people like this who speak in such absolute terms and define audio performance purely on technical grounds.  Just because he compared 4 external master clocks and couldn’t measure an improvement doesn’t make it a universal truth.  If you read further into his conclusion, he suggests that anyone who prefers what a master clock provides is preferring the subjective qualities of that clock and not its technical performance.  Obviously, this article was written from a professional studio perspective and not from the viewpoint of most audiophiles because I don’t know any audiophile who would want it any other way.  While it’s comforting for me when things that sound good also measure well, I think most of us can appreciate that superior measurements don’t always equal better sound.  When things that sound better don't measure better, as I see it, either the measuring equipment isn't sensitive enough to discern a difference or else the wrong things are being measured.

 

While my fledgling viewpoint on clocks is still evolving, here is how I see it as of now.  

1. It’s not just about the clock.  Looking at phase noise plots are a good thing to do but they don’t tell the full picture.  The environment that surrounds the clock that allows that clock to function at its best is just as important and may be the greater limiting factor in most cases.  If you look carefully at SOtM’s sCLK-EX board, you’ll notice much more than just clocks but also a well thought out clock circuit to insure that each clock is isolated and cleanly powered.  If you review my photo of the internals of the REF10, you’ll see essentially the same thing.  The REF10’s tiny OCXO sits in the center of its large chassis but most of the chassis is filled with other components that serve no purpose other than to insure that the clock signal is as clean and unadulterated as possible.  While I have not yet had a chance to evaluate the ISO Regen, it appears that John and Alex also put intelligent thought and effort into optimizing the environment of its clock. 
   
   Now imagine the dozen or so clocks in your audio system.  Your router has at least 2 clocks.  Each switch used has a clock.  Your Ethernet ports will each have its own clock.  Within your motherboard, there is a system clock that serves as a master to a variety of slave clocks (CPU, RAM, SATA, IDE, USB, Firewire, Thunderbolt, PCIe, VGA, integrated audio, RS-232, etc).  Basically, any time the signal passes from one part of the audio “production line” to the next, a reclocking occurs.  If it took the extraordinary measures that SOtM, Mutec, and Uptone Audio went to for their respective clocks to perform at their best, imagine how much harm all these other clocks in our chain must be causing to the signal that aren’t in the same optimized environments?  Someone posted that as they replaced every clock in their audio path from the router forward with better clocks that were cleanly powered, what they noticed was a progressive reduction in glare.  This has been my observation as well.  Those of us who think our stock systems sound great have just gotten used to the sound of this glare, at least that’s what I believe, but once you hear how your music opens up and becomes smoother and less fatiguing with better clocks, it’s hard to go back.
2. Trust your ears.  Forget what the experts tell you if what they say contradicts what you are hearing.  Listening trumps theories or measurements, at least it should.  It’s amazing how an authority figure makes a claim, it gets propagated by the masses, and before long, this claim somehow becomes an irrefutable law of physics.  Sometimes, these experts aren’t measuring the right things.  Many times, they’re as clueless as we are.  If it measures better but you can’t hear the difference, don’t buy it.  If it measures poorly but you love how it sounds, that’s all that matters.

 

 

[guillaume31]

On 5/30/2017 at 3:34 PM, lmitche said:

If you are running the qnap os, the recent versions support serup of a network bridge from the gui.

 

https://www.qnap.com/en/how-to/tutorial/article/how-to-use-and-set-up-your-network-virtual-switch

Thank you very much lmitche !

 

I tried to set-up a "virtual switch" and at first it looked good, until I played with some settings and the NAS disappears all together from my network and I was unable to access it. I did a reset and it was sorted. Somehow I am actually not able to enable the "virtual switch", it does not appears in my menus even after I follow the recommended procedure. I need to check with QNAP if I'm supposed to have this functionality with the HS-251+.

 

On the sMS-200 side, do you need to fix the IP address ? with the latest FW, this functionality does not exist.

 

Thanks

[beautiful music]

Welcome back @romaz 

[magnuska]

Hi!

Received my sms-200 ultra today and tx-usb ultra.

 

A quickl listening to only the sms-200 ultra  ( powered with 12v  HD plex 200) and WOW I can`t believe how much better it is than my microRendu with LPS-1.

Its pure music to my ears even from cold.

 

Look forward to go slowly in steps and after a couple of days add the tx-usb ultra in the chain.

 

Thanks May and Crux audio for a smoth a quick delivery.

 

Regards Magnus

[k-man]

Was it possible for the SMS-200 ultra to be powered by LPS-1, or was it the optimum to have it powered by HD plex? (Or maybe its reserved for the tx-USB ultra ). Plenty to play with 

[austinpop]

Great to hear from you again, Roy!

 

5 hours ago, romaz said:

Having said all of this, I like what Rajiv is trying to do.  I think that to be able to create some sort of an index or FAQ would be very helpful for all.  Personally, I wish there was also some way to bookmark a post.  Regarding creating an index or FAQ for this thread, even though I started this thread, I do not presently have the time to do it although I would be happy to relinquish this authority to anyone who would care to take it on.  Frankly, I would nominate someone like Rajiv who has been a part of this thread from the beginning and as he has implemented many of the things discussed on this thread, he also has firsthand knowledge of their impact.  Just as important, I think he has all of our respect. 

 

Thanks for those kind words. They mean a lot to me.

 

OK - I will take on the job of creating this index, but it is a daunting task, to be sure. I am traveling for the next week, so I will respond when I get back. Part of my approach will surely be to ask for volunteers to help me with sections, especially those dealing with areas in which I didn't dabble - like computer system mods (mobo, SSDs, etc) and fiber/Adnaco. I'll make a better list when I get back, but please reach out to me if you can help.

 

5 hours ago, romaz said:

He said that the REF10’s measurements were probably the best measurements for a 10MHz clock he had ever seen and its performance was better than the sCLK-EX.  Pretty high praise from a competitor clock maker and this was all the validation that I needed to hear.  My order has been placed and so I will be able to decide for myself in a few short weeks just how good this clock is in my system.

 

I've been lurking in that Ref 10 thread too, because it is so intriguing, even though the price point is too rich for my blood. Your experiences will be fascinating once you receive it. One area that remains to be explored is when you have a DAC that can also accept a 10MHz reference clock. I know your DAVE is not setup for this, and there are only a handful of super-expensive DACs from Esoteric and Antelope currently. Although ... the TEAC NT-503 was also on the list. Sadly, I missed the Massdrop sale on this that just ended. If anyone snagged this, consider using this with the upcoming SOtM master reference clock, and reporting back.

 

Speaking of which, that is the other intriguing question. Will Lee take the "best he's ever seen" Mutec data and go create a competing master reference clock in a price range we mere mortals can afford?  

 

5 hours ago, romaz said:

At the same time, as I will be utilizing 2 sCLK-EX boards and its 8 clocks, the thought of being able to synchronize all 8 clocks to a higher level of performance (even if the margin might be smaller) was a salivating thought. 

 

I don't believe we ever really heard how your experiments with the second sCLK-EX board went. Or maybe I missed it. Is the first set of 4 for the switch/sMS/tX, and the next set of 4 upstream in your server? What was the relative effect of the second sCLK-EX?

 

At some point, when next you have the time, please summarize!

 

5 hours ago, romaz said:

Trust your ears.  Forget what the experts tell you if what they say contradicts what you are hearing.  Listening trumps theories or measurements, at least it should.  It’s amazing how an authority figure makes a claim, it gets propagated by the masses, and before long, this claim somehow becomes an irrefutable law of physics.  Sometimes, these experts aren’t measuring the right things.  Many times, they’re as clueless as we are.  If it measures better but you can’t hear the difference, don’t buy it.  If it measures poorly but you love how it sounds, that’s all that matters.

 

Hear, hear!

[Johnseye]

5 hours ago, romaz said:

Prior to Munich, based on the already spectacular sound I was getting from my SOtM stack, based on disparaging  articles such as the one written by Hugh Robjohns that was referenced by @Johnseye, and based on my own lukewarm experience with dCS’s external master clock last year, I had no intention of ever buying an external master clock and I even posted this sentiment back in April.  Of course, part of this sentiment was also based on wishful thinking because I didn’t relish the idea of paying thousands of dollars for a master clock and all the hassle that went with it.  Munich, however, changed all of that.  

 

 

 

As I read Hugh Robjohns’ article, “Does Your Studio Need A Digital Master Clock,” I read with interest his concluding statement.  “Overall, it should be clear…that employing an external master clock cannot and will not improve the sound quality of a digital audio system.”  My response?  Don’t listen to people like this who speak in such absolute terms and define audio performance purely on technical grounds.  Just because he compared 4 external master clocks and couldn’t measure an improvement doesn’t make it a universal truth.  If you read further into his conclusion, he suggests that anyone who prefers what a master clock provides is preferring the subjective qualities of that clock and not its technical performance.  Obviously, this article was written from a professional studio perspective and not from the viewpoint of most audiophiles because I don’t know any audiophile who would want it any other way.  While it’s comforting for me when things that sound good also measure well, I think most of us can appreciate that superior measurements don’t always equal better sound.  When things that sound better don't measure better, as I see it, either the measuring equipment isn't sensitive enough to discern a difference or else the wrong things are being measured.

 

 

Trust your ears.  Forget what the experts tell you if what they say contradicts what you are hearing.  Listening trumps theories or measurements, at least it should.  It’s amazing how an authority figure makes a claim, it gets propagated by the masses, and before long, this claim somehow becomes an irrefutable law of physics.  Sometimes, these experts aren’t measuring the right things.  Many times, they’re as clueless as we are.  If it measures better but you can’t hear the difference, don’t buy it.  If it measures poorly but you love how it sounds, that’s all that matters.

 

 

 

What you are discovering could be a path leading to the master clock architecture being the best solution for home audio.  You and others started down the path of improving a clock, then multiple clocks unlinked, then multiple clocks linked (quasi mastered), then what you've just heard was a full master clock.  There are other variables involved, specifically the quality of the clock itself as that continues to improve in each scenario.  There are still many variables to consider, and that the REF10 was heard at a convention, not in your own home to compare with your previous system, but if what you heard was enough to prompt the REF10 purchase there could be good things to come.  I'm looking forward to your side by side comparison with the sCLK-EX.

 

I mean this with all sincerity, you should write Hugh Robjohns.  He wrote that article 7 years ago but he may be interested in your discoveries.  Maybe his opinions or findings have changed over the years.  You write that one shouldn't blindly trust one person making a statement regardless of whether they're an expert in their field with measurement tools and I agree.  Surprising to heat this from a doctor, but I'm sure that sentiment is valid in your field as well.  One person's findings are just that.

 

In the end, as you point out, all that matters is what sounds good to you.  This is a defacto standard in this hobby.  Everyone's a scientist or engineer in discovery, but everyone's system, listening environment, budget and personal tastes are different.  While listening may trump measurements, Lee's comments on his jitter measurements has and impact on both opinion and his designs.  Measurements shouldn't be discarded, even if you don't think you can hear it.  I'm glad you're putting the investment into this endeavor and sharing your opinions.  I'll be keeping an eye out for your final conclusions once they've been made on the clock.  A lot has been learned, and a lot has changed in the course of a few months.

[lmitche]

10 hours ago, guillaume31 said:

Thank you very much lmitche !

 

I tried to set-up a "virtual switch" and at first it looked good, until I played with some settings and the NAS disappears all together from my network and I was unable to access it. I did a reset and it was sorted. Somehow I am actually not able to enable the "virtual switch", it does not appears in my menus even after I follow the recommended procedure. I need to check with QNAP if I'm supposed to have this functionality with the HS-251+.

 

On the sMS-200 side, do you need to fix the IP address ? with the latest FW, this functionality does not exist.

 

Thanks

Over a year ago I had the virtual switch running with my qnap hs-251+.  Make sure you are running the latest firmware.

 

I have no idea about the sms-200.  However if it's looking for a DHCP server it will find it in either your router behind the qnap, or from the qnap if you enabled the internal DHCP server. Remember to power it up only after the qnap is fully booted.

[limniscate]

I'm getting my stack today and will report back later.

[magnuska]

5 hours ago, k-man said:

Was it possible for the SMS-200 ultra to be powered by LPS-1, or was it the optimum to have it powered by HD plex? (Or maybe its reserved for the tx-USB ultra ). Plenty to play with 

Hi k-man,

 

I choose the 12 volt version for the sms-200 just because It is supposedly sounds a litte better. And yes you are correct the uptone LPS-1 will power the tx-USB ultra. 

My Hd plex 200 is a new updated design compared to the original and it sounds very good actually.

 

 

[limniscate]

I got the 12V versions of the sMS-200 Ultra and dX-USB Ultra as well and will be using the previous generation HD-PLEX to power them until I get my Paul Hynes SR7DRMR2XL.

[austinpop]

15 minutes ago, limniscate said:

I'm getting my stack today and will report back later.

 

Awesome! I'm on a plane, but looking forward to checking these out in a couple weeks at chez liminiscate.

 

BTW - regarding "master" clocks: this post from the Ref 10 thread sheds some light on the differences between master/slave, and reference/word clocks.

 

 

[k-man]

Thanks for your reply @magnuska- I somehow confused myself that LPS-1 had 12V output, but actually it would accept 12V input instead. I forget these things even if I own a LPS-1 

 

 

[k-man]

21 minutes ago, limniscate said:

I got the 12V versions of the sMS-200 Ultra and dX-USB Ultra as well and will be using the previous generation HD-PLEX to power them until I get my Paul Hynes SR7DRMR2XL.

Congrats getting a Paul Hynes! I had a first generation SR3, but it has moved on to a nicer place. Do you mind telling me what the 'DRMR2XL' bit is?

If you had spoken to Lee or May, please tell me the rationale behind the 12V improvement over 6-9V. I only previous thought the PSU choice would be the biggest factor. 

 

[limniscate]

DR = cascaded voltage regulators to give a supply line and rectification interference rejection exceeding 150 dB from DC to 100Khz

MR = Multi Rail (galvanic isolation between the supply rails)

XL = ultra low impedance connectors and fine silver internal wiring between capacitor banks, regulator modules, and the output connectors

 

[romaz]

5 minutes ago, limniscate said:

DR = cascaded voltage regulators to give a supply line and rectification interference rejection exceeding 150 dB from DC to 100Khz

MR = Multi Rail (galvanic isolation between the supply rails)

XL = ultra low impedance connectors and fine silver internal wiring between capacitor banks, regulator modules, and the output connectors

 

 

My new DR SR7 has arrived and the difference against my single regulated SR7 is easily noticeable and very worthwhile.  Improved dynamic contrasts and an even smoother, more relaxed and effortless presentation.  I wasn't sure how my SR7 could get better and now I know.  This is a highly recommended upgrade and I have now decided to replace my standard SR7MR2XL with the DR version.

 

Here is what Paul Hynes had to say:

 

"The DR version effectively expands the dynamic range of clean power delivery by reducing output voltage sag with larger load transient current delivery and reducing the overall noise level of the output voltage rail over the range of load current change. There are other benefits but I would prefer you to have the experience of these before I say anything further, as I do not want to affect your conclusions in any way."

[Johnseye]

4 minutes ago, romaz said:

 

My new DR SR7 has arrived and the difference against my single regulated SR7 is easily noticeable and very worthwhile.  Improved dynamic contrasts and an even smoother, more relaxed and effortless presentation.  I wasn't sure how my SR7 could get better and now I know.  This is a highly recommended upgrade and I have now decided to replace my standard SR7MR2XL with the DR version.

 

Here is what Paul Hynes had to say:

 

"The DR version effectively expands the dynamic range of clean power delivery by reducing output voltage sag with larger load transient current delivery and reducing the overall noise level of the output voltage rail over the range of load current change. There are other benefits but I would prefer you to have the experience of these before I say anything further, as I do not want to affect your conclusions in any way."

 

My understanding is that the DR version is single rail only, and not available in multirail.  Can you confirm if this has changed?

[limniscate]

2 minutes ago, Johnseye said:

 

My understanding is that the DR version is single rail only, and not available in multirail.  Can you confirm if this has changed?

I sure can because I ordered a DRMR2XL.  It fits in the MR4 chassis.

[romaz]

12 minutes ago, Johnseye said:

 

My understanding is that the DR version is single rail only, and not available in multirail.  Can you confirm if this has changed?

 

John, here is what Paul had to say.  My preference has always been to have each rail have its own dedicated transformer and so my SR7s have at most 1-2 rails:

 

"The SR7DR can only support 1 off 6A rail and 1 off 3A rail due to the 

thermal requirements of the four modules used in this configuration. The 

SR7MR4 can support 4 rails but with one transformer or two rails with 

two transformers. 4 rails with 4 transformers would only be possible for 

low current rails of up to 2A. Space is the main issue here. The SR9MR4 

chassis was good regarding size but I was less than happy with the 

structural integrity of the larger chassis using the chassis materials 

of the SR7 for the expanded width. I decided not to pursue this chassis 

style for this application and have designed a heavy-duty custom 

extrusion for the chassis sides capable of supporting up to 10mm 

aluminium plate for the base and front panel and 5mm plate for rear 

panel and top plate. This will be sized to allow 19” rack mounting by 

the addition of rack mounting wings to the extrusion or it can be free 

standing, with or without these rack wings, of course. This more 

substantial structure will allow up to 4 high current rails with 

individual transformers or a combination of high and low power rails 

with various transformer configurations."

 

If you have an order in place, it doesn't mean you couldn't modify it.

[Johnseye]

21 minutes ago, limniscate said:

I sure can because I ordered a DRMR2XL.  It fits in the MR4 chassis.

 

7 minutes ago, romaz said:

 

John, here is what Paul had to say.  My preference has always been to have each rail have its own dedicated transformer and so my SR7s have at most 1-2 rails:

 

"The SR7DR can only support 1 off 6A rail and 1 off 3A rail due to the 

thermal requirements of the four modules used in this configuration. The 

SR7MR4 can support 4 rails but with one transformer or two rails with 

two transformers. 4 rails with 4 transformers would only be possible for 

low current rails of up to 2A. Space is the main issue here. The SR9MR4 

chassis was good regarding size but I was less than happy with the 

structural integrity of the larger chassis using the chassis materials 

of the SR7 for the expanded width. I decided not to pursue this chassis 

style for this application and have designed a heavy-duty custom 

extrusion for the chassis sides capable of supporting up to 10mm 

aluminium plate for the base and front panel and 5mm plate for rear 

panel and top plate. This will be sized to allow 19” rack mounting by 

the addition of rack mounting wings to the extrusion or it can be free 

standing, with or without these rack wings, of course. This more 

substantial structure will allow up to 4 high current rails with 

individual transformers or a combination of high and low power rails 

with various transformer configurations."

 

If you have an order in place, it doesn't mean you couldn't modify it.

 

My rail 1 will be start at 19v @10A so that if I ever want to support a more power hungry PC I have the option.  That may limit me from the DR version since it appears to max at 6A.  If that's the case, do I opt for the DR now and eliminate the option to power a different PC?  He just started working on my build so if I make a change it has to be now. 

 

Here's the spec

 

SR7MR2XL
Rail 1 – 19v @ 10A adjustable down to 12v @ 5A initially set to 12v
Rail 2 – 4v to 14v @ 3A initially set to 9v

DC10FSXL with 2.5mm DC plug
DC6FSXL with 2.1mm DC plug

[romaz]

1 minute ago, Johnseye said:

 

 

My rail 1 will be start at 19v @10A so that if I ever want to support a more power hungry PC I have the option.  That may limit me from the DR version since it appears to max at 6A.  If that's the case, do I opt for the DR now and eliminate the option to power a different PC?  He just started working on my build so if I make a change it has to be now. 

 

Here's the spec

 

SR7MR2XL
Rail 1 – 19v @ 10A adjustable down to 12v @ 5A initially set to 19v
Rail 2 – 4v to 14v @ 3A initially set to 9v

DC10FSXL with 2.5mm DC plug
DC6FSXL with 2.1mm DC plug

 

You might want to ask him for his advice.  You could opt for applying the DR to Rail 2 but to be honest, it may be with your server where the DR has the greatest impact.

 

The XL option is very worthwhile (impedance <1milliohm) and I have chosen this with each of my SR7s.  The problem is with the 2.1mm or 2.5mm DC plugs.  These are not low impedance plugs.  It would be ideal if you could apply the XL connector to your music server and your endpoints and I have thought about this but the XL connector is also very large and heavy.  There's no way, for example, to use this connector with something small like the microRendu.

[Johnseye]

3 minutes ago, romaz said:

 

You might want to ask him for his advice.  You could opt for applying the DR to Rail 2 but to be honest, it may be with your server where the DR has the greatest impact.

 

The XL option is very worthwhile (impedance <1milliohm) and I have chosen this with each of my SR7s.  The problem is with the 2.1mm or 2.5mm DC plugs.  These are not low impedance plugs.  It would be ideal if you could apply the XL connector to your music server and your endpoints and I have thought about this but the XL connector is also very large and heavy.  There's no way, for example, to use this connector with something small like the microRendu.

 

I sent him an email and hopefully it's read before he's done with the build.  I could just go for the 12v DR now but what if I change the motherboard.  Future proofing is tough in this hobby.

 

Another benefit I get from the 19v rail is that I have a turntable and phono stage that each operate just under 19v.  I can power one of those when I'm not powering the server if I'd like.  The flexibility is huge given the cost.  I have a plan to upgrade the turntable and phono stage at some point, otherwise I would have gone with the 4 rail.  Those may get their own dedicated LPSU if what I choose allows for it.

 

Thanks for the advice and breathing some life into this thread again.  It gets stale without the exciting news you bring.

[romaz]

1 minute ago, Johnseye said:

 

I sent him an email and hopefully it's read before he's done with the build.  I could just go for the 12v DR now but what if I change the motherboard.  Future proofing is tough in this hobby.

 

Another benefit I get from the 19v rail is that I have a turntable and phono stage that each operate just under 19v.  I can power one of those when I'm not powering the server if I'd like.  The flexibility is huge given the cost.  I have a plan to upgrade the turntable and phono stage at some point, otherwise I would have gone with the 4 rail.  Those may get their own dedicated LPSU if what I choose allows for it.

 

Thanks for the advice and breathing some life into this thread again.  It gets stale without the exciting news you bring.

 

You could ask him to do what he did for me with my DR.  Even though my DR is a single rail supply, it has 2 outputs.  It has a DR 12V output that can be adjusted down to 2V very easily with the turn of a small knob on a potentiometer within the SR7.   He also installed for me a 2nd output that outputs 19V but if I choose to use this 2nd output, I lose the benefits of double regulation because at 19V, I am now running only off the 1st stage regulator.