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

[seeteeyou]

Thank you, seeteeyou!

Hi Roy, you're very welcomed! Actually I'm only interested in sCLK-EX for improving the SQ of something called SDTrans384 from Japan, I don't really listen to that music myself and that's why having a few SD cards plus a super simple player should be more than enough for me.

 

So basically I'm just more interested in the technical side and then anything related to Chord Electronics would float my boat as well. Rob's M-Scaler with over a million taps could be potentially revolutionary since that should be able to show us the genuine quality of RBCDs. Of course that's also great news for those of us who are big fans of TIDAL etc.

 

---

Since you're asking SOtM to mod Gigabyte GA-H110TN-GSM Plus already, maybe it's a good idea to find out if the same bypass surgery were applicable for SO-DIMM or otherwise

 

http://www.tirnahifi.org/forum/viewtopic.php?f=15&t=3267

http://aktives-hoeren.de/viewtopic.php?f=30&t=6015

 

If that were actually feasible, you're still getting some benefits whether you're going CF route or RAM route. As I mentioned in another thread, adding whole bunch of supercapacitors to Hugo's power source turned out to be a game changer while that memory power bypass procedure mentioned above also happened to involve something similar

 

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/chord-hugo-re-examined-20739/index18.html#post433203

 

---

So far I haven't seen any Linux distros that are already running Roon Server right off the bat, though there's something worth mentioning and it's designed for Synology NAS

 

https://community.roonlabs.com/t/roon-running-in-docker-on-synology/9979

https://hub.docker.com/r/ronch/roon-server/

 

Maybe you could get that running on your Gigabyte GA-H110TN-GSM Plus as well

 

http://xpenology.me

http://xpenology.com/forum/

 

If not, here's something similar that's based on Alpine Linux

 

https://hub.docker.com/r/josepba/roonserver/

https://github.com/josepba/dockerroonserver/

https://hub.docker.com/r/frolvlad/alpine-glibc/

https://github.com/frol/docker-alpine-glibc

 

 

---

Since the upcoming Wi-Fi-only Chord Poly should have ARM instead of x86 processors, it won't be able to run Roon Server but I'm still looking forward to a streamer of some sort with an integrated M-Scaler in the future.

[romaz]

Actually I'm only interested in sCLK-EX for improving the SQ of something called SDTrans384 from Japan...

This is an interesting SD card player. It appears to be using a TCXO clock and so I'm fairly certain the sCLK-EX would improve it.

 

So basically I'm just more interested in the technical side and then anything related to Chord Electronics would float my boat as well. Rob's M-Scaler with over a million taps could be potentially revolutionary since that should be able to show us the genuine quality of RBCDs. Of course that's also great news for those of us who are big fans of TIDAL etc.

None of my posts on this thread have had anything to do with Chord or with Rob Watts' designs, that is really a different topic altogether. With that said, all of the improvements I am talking about here with respect to this direct connection and replacing clocks pale in comparison to the SQ impact I have personally experienced with his soon-to-be-released M-scaler but because his M-scaler will be all the more revealing of what is upstream, I find it important and have found it very worthwhile to try and improve what is upstream as much as possible.

 

Since you're asking SOtM to mod Gigabyte GA-H110TN-GSM Plus already, maybe it's a good idea to find out if the same bypass surgery were applicable for SO-DIMM or otherwise

MoBo Memory Power Bypass - Tír Na HiFi

aktives-hoeren.de • Thema anzeigen - Separate Stromversorgung für RAM

 

Very interesting and worth exploring as I'm sure there would be SQ gains to be had for sure by externally powering your RAM. The problem is that devising a high quality 1.2-1.5V/4-8A PSU would probably not be easy.

As I mentioned in another thread, adding whole bunch of supercapacitors to Hugo's power source turned out to be a game changer while that memory power bypass procedure mentioned above also happened to involve something similar

 

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/chord-hugo-re-examined-20739/index18.html#post433203

I am not surprised by this finding. I have found all of Rob's DACs to excel in digital conversion and I have found that even his $550 Mojo performs digital conversion better than DACs that cost >$25k but at the price point of Mojo and Hugo and because these DACs were designed to be portable, there were necessary and intentional compromises made with respect to the quality of the analog stage and the power supply and so that is where these DACs fall short. I have even explored a PSU upgrade for my DAVE and have had discussions with Paul Hynes about this matter. I'm sure many share my belief that good power is foundational to all good things in audio but what is still very surprising to me is just how well something like the mR or sMS-200 (which consume less than 5w each on average) responds so favorably to something like Paul Hynes' SR7 and its 250VA mains transformer designed for 125w continuous use and 480w of transient power.

 

Maybe you could get that running on your Gigabyte GA-H110TN-GSM Plus as well

 

I will now be abandoning this Gigabyte motherboard. Before sending it off to SOtM for clock modification, I felt it was important to build it first to make sure my selected components were compatible and I'm glad I did. Unfortunately, none of the newest boards that incorporate Intel's 1151 chipset is capable of SATA in pure IDE mode (only AHCI). Consequently, I am finding that it is not possible to use a CF card as a boot drive which is a deal breaker.

 

I will have to go backward to an older motherboard that utilizes an 1150 chipset and I have found several that I like although I presently have my eye on an Asrock server-class enhanced mini-ITX motherboard, the MT-C224, that has even fewer unnecessary ICs such as onboard audio and can still be powered from a single 12V lead from my SR7.

 

Using this older chipset will also force me to abandon my 7th generation Intel 7700T but the advantage of this server-class motherboard is that it can utilize a Xeon class CPU and ECC RAM. This motherboard is specifically compatible with an even lower power dual-core Xeon, the E3-1220L V3, with only a 13w TDP and yet still has a fairly generous 4MB of L3 SmartCache (meaning the entire cache is usable for a single core).

 

At this point, I don't know how much energy I have to want to compare SQ among different RAM with various timings and different brands of SLC CF cards (Apacer, Delkin, Silicon Tech) and so I have decided to move ahead with the recommendations of others, specifically Phil Hobi of AO and Marcin of JPLAY:

 

https://www.highend-audiopc.com/PDF/reference-system.pdf

 

SSD Sound differences - eSATA SSD vs. PCI-E SSD vs. mSATA SSD - Computer Audio - JPLAY - hi-end audio player for Windows

[lmitche]

Roy, your writing inspired me to research alternative connection points for my sata hard disk cable to the motherboard. My motherboard supports 4 SATA 6 and 2 SATA express connections. Today I installed a $5 m.2 ngff to SATA board into the m.2 slot next to the CPU, hooked it up, and it is the best sounding connection point I can find for the drive. I suspect this is due to the much shorter path to the CPU.

 

On another topic, for over a year I ran my machine via ISCSI boot from my NAS to avoid any drive in my HQplayer upsampling machine. This works well. However, in the end, placing the OS and music in a simple local hard disk sounds better as the network background tasks can be turned off entirely. FWIW.

 

When you tested the 2.5 hard drive, was it externally powered? My guess is that the benefit of the 3.5 inch hard disk is that it uses the 12 volt bus, not the 5 volt, keeping crap away from the latter. 2.5 inch drives use 5 volts.

 

Thanks,

 

Larry

[greenleo]

At this point, I don't know how much energy I have to want to compare SQ among different RAM with various timings and different brands of SLC CF cards (Apacer, Delkin, Silicon Tech) and so I have decided to move ahead with the recommendations of others, specifically Phil Hobi of AO and Marcin of JPLAY:

 

https://www.highend-audiopc.com/PDF/reference-system.pdf

 

SSD Sound differences - eSATA SSD vs. PCI-E SSD vs. mSATA SSD - Computer Audio - JPLAY - hi-end audio player for Windows

 

Hi Romaz,

 

I've read the JPlay thread and it states that the Pang's modified CF SATA card is no longer available. What CF SATA card will you use to connect your CF card to the mono?

[romaz]

Roy, your writing inspired me to research alternative connection points for my sata hard disk cable to the motherboard. My motherboard supports 4 SATA 6 and 2 SATA express connections. Today I installed a $5 m.2 ngff to SATA board into the m.2 slot next to the CPU, hooked it up, and it is the best sounding connection point I can find for the drive. I suspect this is due to the much shorter path to the CPU.

Thanks for sharing this finding, Larry. I have been curious myself about whether it is possible to tap more directly to the CPU via M.2 if you are using a SATA drive instead of an NVMe drive and your finding suggests that it is possible although this could be motherboard dependent. Here is the block diagram of a typical modern ASUS motherboard that utilizes a Z170 chipset:

 

As you can see, in this particular motherboard, M.2 doesn't connect directly to the CPU but through the Platform Controller Hub which then connects to the CPU via DMI. With M.2, which is essentially just a connection interface, you then have the option of connecting to the hub either via PCIE bus or the SATA bus depending on what type of drive is connected. Obviously, the SATA bus is a slower, higher latency bus and to avoid it would be ideal.

 

I had assumed that if I used a SATA to M.2 NGFF adapter like you have used for your SATA drive, it would be no different than plugging the drive directly into a SATA port but with your motherboard, it appears that this might not be the case. In the ASUS block diagram above, this particular motherboard seems to use SATA Express as opposed to just plain SATA with it's M.2 connector. The advantage of SATA Express is that it seems to ride on a faster PCIEx2 bus and that may be what is going on with your motherboard.

 

Looking at the above block diagram, as you can also see, the two PCIEx16 slots have direct access to the CPU without having to go through the hub and so this opens up a new opportunity. It would seem to me that if I used a SATA to PCIE adapter board and plugged it into one of these PCIEx16 slots, any SATA drive I used would now have the most direct access to the CPU possible and I may need to explore this further as that may be a way I could use a CF card to have direct access to the CPU although I fear there may be compatibility issues.

 

On another topic, for over a year I ran my machine via ISCSI boot from my NAS to avoid any drive in my HQplayer upsampling machine. This works well. However, in the end, placing the OS and music in a simple local hard disk sounds better as the network background tasks can be turned off entirely. FWIW.

 

I hear you. I have tried this with AO and for sure, SQ improves if you shut down network services. Unfortunately, this is an option (or rather a convenience) that I can't seem to live without. Because I prefer to use an iPad Pro or remote computer for control and like to stream Tidal often enough, I find the convenience a necessity.

When you tested the 2.5 hard drive, was it externally powered? My guess is that the benefit of the 3.5 inch hard disk is that it uses the 12 volt bus, not the 5 volt, keeping crap away from the latter. 2.5 inch drives use 5 volts.

Yes, my 2.5 inch hard drive was powered with my LPS-1 at 5v. I basically swapped out my SSD and swapped in my hard drive.

[romaz]

Hi Romaz,

 

I've read the JPlay thread and it states that the Pang's modified CF SATA card is no longer available. What CF SATA card will you use to connect your CF card to the mono?

I'll refer you back to my post #972:

 

http://www.computeraudiophile.com/f10-music-servers/novel-way-massively-improve-sq-sms-200-and-microrendu-31110/index39.html#post643108

 

Paul discontinued this adapter since all new motherboards seem to no longer do IDE mode and because some were complaining of how much more sluggish their OS had become. Obviously, this is something that only makes sense with a dedicated "audio only" headless computer. You wouldn't want to do any work on it or surf the internet. It works well with Roon because control takes place on a separate machine or tablet and so with my Mac Mini that is running off an SD card, I am not detecting any significant sluggishness at all with my iPad or Mac Pro, which I use for control.

 

The only difference between Paul's adapter and the one I am buying is his adapter utilizes either his TCXO or OCXO. I plan to use one of the clocks on the sCLK-EX board. It remains unclear to me how putting a clock here makes a difference when I am already planning on replacing the motherboard's system clock but Paul insists there is a worthwhile difference and so all the drives that he sells come with his clocks.

[gadgetman]

SOtM hasn't listed the specs for the sPS-500. It states it is a hybrid PSU and so I don't yet know what that means but my guess would be that it involves batteries or supercaps in some fashion. If this is the case, it would be unlikely to be able to power a Mac Mini. Even Vinnie Rossi's Mini Pure DC-4-EVR supply can't power a Mac Mini.

 

EDIT: I stand corrected! You are right. I just noticed there are now specs listed and indeed, 12V/5A is listed as an option which is quite surprising considering how small this PSU is. I am definitely interested to know how well this performs.

 

What I will say is that my experience with PSUs with switchable voltages is that they come with challenges. With Paul Hynes' PSUs, to do this requires cascading regulators (which is a good thing in terms of added line noise rejection) but as you go from 12V down to 7V, for example, to maintain 5A of output results in more heat generation that could result in stability issues. Exactly how the sPS-500 will address this will be interesting to see.

 

If sPS-500 is truely a hybrid and a small PSU then it's highly possible a switching PSU with linear regulators. Under this configuration the heat dissipation may won't a big issue. As to my knowledge Chord Electronics is one of the players who's using such mechanism in most of its products. As long as the RFI and electric noise can be controlled at a minimum I don't think it can't be a good audiophile grade PSU.

 

I own a mid 2010 Mac mini, thought it may be quite different from the latest version but they are all using a 85W PSU. I had measured my good old Mac mini and it never goes over 3A in normal use. So, it's possible that sPS-500 can power Mac mini as long as no heavy load is running.

[seeteeyou]

If I were actually building something for running Roon Server for myself, I wouldn't count on the built-in DC-ATX power supply that's provided by manufacturers of server boards. Most likely I'd rather go for the external ones since I could even opt for better wiring materials while still utilizing something nice such as SR7 from Paul Hynes

 

https://item.taobao.com/item.htm?id=41232541497

https://item.taobao.com/item.htm?id=545477585125

 

Audiophiles in China seemed to like his stuff very much

 

https://shop106855878.taobao.com/category-1049180713.htm

 

HDPLEX et al. should have similar offerings as well, maybe Hynes could also look into that kinda direction?

 

---

Those server boards with Xeon processors could provide extra benefits such as Intel Node Manager

 

http://www.intel.com/content/www/us/en/data-center/data-center-management/node-manager-general.html

http://www.intel.com/content/www/us/en/power-management/intelligent-power-node-manager-3-0-specification.html

 

Granted we're talking about 1.1GHz processor @ 13W TDP already, it's still interesting to see what could be accomplished with differernt power management options

 

https://software.intel.com/en-us/articles/power-management-states-p-states-c-states-and-package-c-states

[romaz]

Roy, your writing inspired me to research alternative connection points for my sata hard disk cable to the motherboard.

 

Larry, I'm going to need to return the compliment. Hopefully, a I've come across a serious breakthrough thanks to your post.

 

I have found another Asrock motherboard, the E3C222D4U, but this one is now slightly bigger (microATX). I will need to move up to the slightly larger Streacom FC9 Alpha (still fanless) chassis and will now need to go to a Mini-Box picoPSU to power it although I should still be able to power it with a single 12V lead from my SR7. Like the other Asrock motherboard, this one does true IDE mode and can run a Xeon and ECC memory.

 

The advantage of this motherboard is that it has 2 PCIE slots (x16 and x8) that directly connect to the CPU. It has a 3rd PCIE slot (PCIE 2.0x8) that connects via the Platform Controller Hub. What this allows is the opportunity to plug my Intel dual LAN card into one of the directly connecting PCIE slots and the following dual port SATA 3 to PCIE controller into the other directly connecting PCIE slot:

 

The great thing with this SIIG card is that it has a replaceable clock for 2 SATA ports and so I can clock both the OS CF drive and a 2nd adapter made by SYBA that conveniently stores 2 CF cards (potentially 1TB of music) as follows:

 

The nice thing about this SYBA dual CF to SATA adapter is that 2 CF cards can now be powered by a single SATA power cable.

 

I could then use a 2nd SIIG SATA to PCIE adapter for the 3rd PCIE slot that connects via the Platform Controller Hub which then gives me 2 more clockable SATA ports that at least connect "semi-directly" to the CPU and so I now have the option of connecting 4 more CF cards for a further 2TB of music storage (or a total of 3TB). Despite having to power possibly 7 CF cards, since only the OS CF card and only one of the music CF cards will be active at one time, I should be able to power everything with a single LPS-1.

 

Of course, the big question is whether this will work? Since my intended boot drive will not be using the SATA bus at all, I'm not sure if IDE mode capability in BIOS will have an impact.

[romaz]

If sPS-500 is truely a hybrid and a small PSU then it's highly possible a switching PSU with linear regulators. Under this configuration the heat dissipation may won't a big issue. As to my knowledge Chord Electronics is one of the players who's using such mechanism in most of its products. As long as the RFI and electric noise can be controlled at a minimum I don't think it can't be a good audiophile grade PSU.

 

I own a mid 2010 Mac mini, thought it may be quite different from the latest version but they are all using a 85W PSU. I had measured my good old Mac mini and it never goes over 3A in normal use. So, it's possible that sPS-500 can power Mac mini as long as no heavy load is running.

Quite true. A switching PSU with linear regulators can work very well and indeed, Chord is famous for doing this. I would have to guess this is the topology of the sPS-500 as well.

 

My 2014 Mac Mini that runs off an SD card utilizes a 15w CPU and only 4GB of RAM with no internal hard drive and with the wifi card removed. At idle, it consumes less than 5w and never runs warm even when active. Because the fan never has to ramp up, I never hear it. Of course, when it boots, it will require more current but I suspect a 12V/3A PSU could easily power mine.

[romaz]

If I were actually building something for running Roon Server for myself, I wouldn't count on the built-in DC-ATX power supply that's provided by manufacturers of server boards. Most likely I'd rather go for the external ones since I could even opt for better wiring materials while still utilizing something nice such as SR7 from Paul Hynes

 

https://item.taobao.com/item.htm?id=41232541497

https://item.taobao.com/item.htm?id=545477585125

 

Audiophiles in China seemed to like his stuff very much

 

https://shop106855878.taobao.com/category-1049180713.htm

 

HDPLEX et al. should have similar offerings as well, maybe Hynes could also look into that kinda direction?

 

---

Those server boards with Xeon processors could provide extra benefits such as Intel Node Manager

 

Intel® Node Manager

Intel® Intelligent Power Node Manager 3.0: Specification

 

Granted we're talking about 1.1GHz processor @ 13W TDP already, it's still interesting to see what could be accomplished with differernt power management options

 

https://software.intel.com/en-us/articles/power-management-states-p-states-c-states-and-package-c-states

Thank you. You're absolutely right and because I have decided to move to microATX, I am now thinking this way. Most recommend higher voltage PSUs (24V) to enable lower current draw but I want to avoid an ATX PSU at all costs. Nothing against HDPlex but I just don't want to go there. Because I expect this server to consume barely 15w at the most, I am still hoping to be able to use a single 12V/6A lead from my SR7. Regarding the MiniPox Pico PSU, I will approach SOtM first. I am hoping they will have the in-house capability to transform it further including replacing all the cabling. If not, I would certainly trust Paul Hynes to do it.

 

As for the Intel Node Manager, this is something that is foreign to me but certainly looks interesting. As a rule, with Process Lasso, you let it handle CPU power management on the fly rather than set a fixed power management scheme and I have found it works quite effectively with regards to being efficient and responsive when it needs to be but also being able to shut down cores when not in use but perhaps, there is a better way.

[seeteeyou]

FYI - someone in China was comparing quite a few SMPS and LPS with somewhat surprising results in terms of the ripple

 

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/power-supply-8-group-test-linear-power-supply-and-switchedmode-power-supply-31879/

 

Other audiophiles in Taiwan actually tried LPS with picoPSU but didn't like that combo very much. Then they just went for TWO pieces of ATX power supplies (1200W each) and the sound turned out to be very pleasing to their ears because of the much improved dynamics

 

http://www.andaudio.com/phpbb3/viewtopic.php?f=25&t=128680

 

It's a big no-no to power the CPU and other components with the same PSU, that's the whole point of having 2 separate ones instead of "putting all your eggs in one basket" so to speak

 

http://www.andaudio.com/phpbb3/viewtopic.php?f=25&t=128680&start=25#p1122025

 

---

Now we've got two schools of thought. First one is all about minimizing the noise while reducing the power consumption.

 

Another one is supplying 1200W (or twice with two PSUs) to fanless computers that could be fed with 25W or so, obviously cooling fans will never kick in since we simply won't reach that kinda utilization percentage. Once the LPS was replaced by that "monster" ATX power supply, they even got rid of stuff like USB DDC, dual-headed USB cables as well as the UpTone Regen (not sure if that were Green or Amber?) since those gadgets only provided marginal improvements or even knocking some points off the SQ department. That's somewhat similar to the experiences who went for the optimized Window 10 in RAM route since they no longer needed USB DDC or expensive LAN cables such as Audioquest Diamond etc.

 

Those audiophiles in Taiwan already tried stuff like WS2012r2 with AO in Core Mode, FidelizerPro, HQPlayer, JPLAY, NAS, Daphile, Raspberry Pi etc. back then but they're still rooting for a pair of WAY overkill ATX power supplies. Maybe it's just a matter of personal preferences since stuff like psychoacoustics could mean different things for different ears, or they could be onto something that we still aren't aware of yet? It's like an entirely new beast since we could very well be talking about a real clash of paradigms here IMHO.

 

It doesn't hurt to give those quality ATX power supplies a shot if we could borrow them somewhere or order them somewhere with good return policies.

[romaz]

FYI - someone in China was comparing quite a few SMPS and LPS with somewhat surprising results in terms of the ripple

 

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/power-supply-8-group-test-linear-power-supply-and-switchedmode-power-supply-31879/

 

Other audiophiles in Taiwan actually tried LPS with picoPSU but didn't like that combo very much. Then they just went for TWO pieces of ATX power supplies (1200W each) and the sound turned out to be very pleasing to their ears because of the much improved dynamics

 

AndAudio.com • 檢視主題 - 80分的CAT之路

 

It's a big no-no to power the CPU and other components with the same PSU, that's the whole point of having 2 separate ones instead of "putting all your eggs in one basket" so to speak

 

AndAudio.com • 檢視主題 - 80分的CAT之路

 

---

Now we've got two schools of thought. First one is all about minimizing the noise while reducing the power consumption.

 

Another one is supplying 1200W (or twice with two PSUs) to fanless computers that could be fed with 25W or so, obviously cooling fans will never kick in since we simply won't reach that kinda utilization percentage. Once the LPS was replaced by that "monster" ATX power supply, they even got rid of stuff like USB DDC, dual-headed USB cables as well as the UpTone Regen (not sure if that were Green or Amber?) since those gadgets only provided marginal improvements or even knocking some points off the SQ department. That's somewhat similar to the experiences who went for the optimized Window 10 in RAM route since they no longer needed USB DDC or expensive LAN cables such as Audioquest Diamond etc.

 

Those audiophiles in Taiwan already tried stuff like WS2012r2 with AO in Core Mode, FidelizerPro, HQPlayer, JPLAY, NAS, Daphile, Raspberry Pi etc. back then but they're still rooting for a pair of WAY overkill ATX power supplies. Maybe it's just a matter of personal preferences since stuff like psychoacoustics could mean different things for different ears, or they could be onto something that we still aren't aware of yet? It's like an entirely new beast since we could very well be talking about a real clash of paradigms here IMHO.

 

It doesn't hurt to give those quality ATX power supplies a shot if we could borrow them somewhere or order them somewhere with good return policies.

 

It seems all the PSU comparisons I see discuss only ripple noise. If the testing ends there, I don't know that much has been accomplished.

 

Thanks, I appreciate the advice but I do believe we are talking about a clash in paradigms. I am very much in favor of minimizing noise while reducing power consumption. I am also very much in favor of speed, agility and responsiveness that only a low impedance PSU can provide as opposed to brute force and blunt detail for the sake of dynamics and so I will head in my preferred direction first and see if I like what I hear. Nothing against those who use dual 1200W PSUs for a music server that will consume 15w but unless you've heard Paul Hynes' SR7, I'm not sure you know where I'm at either.

 

The PicoPSU may indeed be the limiting factor and so I have already reached out to both Paul and SOtM for their advice and assistance. At the very least, I agree, I want the cabling on this Pico replaced and I believe it can be done. The only thing I need to power with it is the motherboard, CPU and RAM just like with my Mac Mini, and if anything, I will be underclocking and not overclocking.

 

What I am doing with this build has absolutely nothing to do with USB DDC, dual headed USB cables, USB Regen, LAN cables, etc. Those kinds of things are completely separate and not remotely relevant.

[greenleo]

I'll refer you back to my post #972:

http://www.computeraudiophile.com/f10-music-servers/novel-way-massively-improve-sq-sms-200-and-microrendu-31110/index39.html#post643108

The only difference between Paul's adapter and the one I am buying is his adapter utilizes either his TCXO or OCXO. I plan to use one of the clocks on the sCLK-EX board. It remains unclear to me how putting a clock here makes a difference when I am already planning on replacing the motherboard's system clock but Paul insists there is a worthwhile difference and so all the drives that he sells come with his clocks.

I see, you will buy a CF SATA converter and then do a mod. Loo forward to your findings.

[greenleo]

As a rule, with Process Lasso, you let it handle CPU power management on the fly rather than set a fixed power management scheme and I have found it works quite effectively with regards to being efficient and responsive when it needs to be but also being able to shut down cores when not in use but perhaps, there is a better way.

 

So you just use the default mode of Lasso without any tweaking? Is the improvements in the SQ obvious? Given so many optimizers like AO, Fidelizer, Lasso and JPlay, I've bought AO and find the improvements obvious and am wondering if I should try other optimizers as well.

[Theobetley]

FYI - someone in China was comparing quite a few SMPS and LPS with somewhat surprising results in terms of the ripple

 

 

Other audiophiles in Taiwan actually tried LPS with picoPSU but didn't like that combo very much. Then they just went for TWO pieces of ATX power supplies (1200W each) and the sound turned out to be very pleasing to their ears because of the much improved dynamics

 

 

 

 

 

---

The tests you refer to are very interesting. Also that many of your colleagues use 2 high power atx supplies is also very interesting. Back in the day Cics (of Cplay/cmp fame) also recommended running music server on 2 atx's--one for motherboard and one for cpu. I have faithfully followed this prescription for years since. Nice to see that others also do this.

[lmitche]

It seems all the PSU comparisons I see discuss only ripple noise. If the testing ends there, I don't know that much has been accomplished.

 

Thanks, I appreciate the advice but I do believe we are talking about a clash in paradigms. I am very much in favor of minimizing noise while reducing power consumption. I am also very much in favor of speed, agility and responsiveness that only a low impedance PSU can provide as opposed to brute force and blunt detail for the sake of dynamics and so I will head in my preferred direction first and see if I like what I hear. Nothing against those who use dual 1200W PSUs for a music server that will consume 15w but unless you've heard Paul Hynes' SR7, I'm not sure you know where I'm at either.

 

The PicoPSU may indeed be the limiting factor and so I have already reached out to both Paul and SOtM for their advice and assistance. At the very least, I agree, I want the cabling on this Pico replaced and I believe it can be done. The only thing I need to power with it is the motherboard, CPU and RAM just like with my Mac Mini, and if anything, I will be underclocking and not overclocking.

 

What I am doing with this build has absolutely nothing to do with USB DDC, dual headed USB cables, USB Regen, LAN cables, etc. Those kinds of things are completely separate and not remotely relevant.

You may want to try the HDPLEX DC to DC PSU instead of the Pico PSU. The hdplex sounds better then the 160xt Pico PSU in my low power rig.

[lmitche]

FYI - someone in China was comparing quite a few SMPS and LPS with somewhat surprising results in terms of the ripple

 

 

Other audiophiles in Taiwan actually tried LPS with picoPSU but didn't like that combo very much. Then they just went for TWO pieces of ATX power supplies (1200W each) and the sound turned out to be very pleasing to their ears because of the much improved dynamics

 

 

 

 

 

---

The tests you refer to are very interesting. Also that many of your colleagues use 2 high power atx supplies is also very interesting. Back in the day Cics (of Cplay/cmp fame) also recommended running music server on 2 atx's--one for motherboard and one for cpu. I have faithfully followed this prescription for years since. Nice to see that others also do this.

I use a 1600 EVGA titanium ATX SMPSes here for motherboard and hard disk power and it sounds great, so I can second the results of the Taiwanese test.

 

This is also intriguing:

 

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story6&reid=487

 

A friend with a duplicate of my rig just ordered one, so we will hear soon about results.

[lmitche]

Romaz,

 

Turning off the Windows networking tasks still allows remote control via tightvnc server and Roon access to Tidal. It is only drive sharing that is lost. I turn it off after boot, so it just takes a reboot to turn everything back on.

[Superdad]

My 2014 Mac Mini that runs off an SD card utilizes a 15w CPU and only 4GB of RAM with no internal hard drive and with the wifi card removed. At idle, it consumes less than 5w and never runs warm even when active. Because the fan never has to ramp up, I never hear it. Of course, when it boots, it will require more current but I suspect a 12V/3A PSU could easily power mine.

 

Sorry Roy, but a 3A PS will not boot a Mac mini, no matter how slim you make it. I've done the tests on the bench, and minis will fail to boot with anything less than about 3.9A.

Just FYI.

[Superdad]

To go from 4GB of DDR3 (which is what I currently use and what I expect to use with my new build) to 16GB of DDR3, if this calculator is to be believed, under load with 16G of RAM, you would draw a whopping 4A more current compared against 4G of RAM. DDR4 fares a bit better and the draw is 3.3A more when you jump from 4G to 16G but this is still huge. As for 1 vs 2 sticks of RAM, using 2 sticks of 8G RAM draws no more current when compared against just a single stick of 16G RAM.

 

Whoa! 4 amps for 16GB of RAM?! That calculator you are using is not to be believed... If anywhere near that much current was drawn by RAM (maybe magnetic core memory from the 1950s?) we would not have laptop computers and the traces on RAM modules would just burn up...

 

Just sayin'...

[lmitche]

Actually magnetic core memory didn't take much power as it held it's state without any. We used to load the OS, turn the machine off, ship it, and then just start it up from the same PC (program counter) at the customers site.

 

It was hand woven too.

[R1200CL]

Deleted

[Theobetley]

If I had to vote I would vote this thread with all of its sonic opportunities for computer audiophiles as the greatest thread ...ever!

[rickca]

So now I get it. To build a good audio PC, take everything you know about building a modern gaming PC and do the opposite. High resolution display ... bad. NVMe PCIe SSD ... bad. Tons of fast memory ... bad. OK, I'm kidding you guys but really it's kind of funny.