XXHighEnd - The Ultra HighEnd Audio Player
July 17, 2019, 08:18:07 am *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: August 6, 2017 : Phasure Webshop open ! Go to the Shop
Search current board structure only !!  
   Home   Help Search Login Register  
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 [13] 14 15  All
  Print  
Author Topic: Investigating a full linear PSU for the XXHE PC and NOS1  (Read 192142 times)
0 Members and 1 Guest are viewing this topic.
acg
Audio Addict
***
Offline Offline

Posts: 749


View Profile
« Reply #180 on: October 05, 2015, 10:47:38 am »


Hi, Anthony

Could you  please share what kind of controler you're going to use for the ATX start and shut-down ?
So I can leave my DC/DC converter and the relays I am using now?
Thanks,
Ed



Hi Ed,

At this stage the controller will be a little SMD board which will do three primary things:

  • monitor voltage and current on three or four power supply rails downstream of the individual rails
  • communicate with motherboard to start and stop the computer
  • control one or more mains relays to control operation of the various power supply rails

Of course I am unfamiliar with your DC-DC converters and relays, but you will be able to connect the output of your supplies to this pcb so they can be monitored and if your relays have 5V coils, are non-latching and NC type there may be enough current available to switch them, but there are no guarantees in that regard.  If you like you can post a link to the datasheet of your relays and I may be able to consider them when I prototype the board shortly.

Regards,

Anthony
Logged

Audio PC
Stealth Mach II with Xeon E5 2698 v4 20 Core 2GHz  with Hyperthreading On [40 cores]/ 32GB Ram / RAM-OS / mobo USB port

XXHighEnd 2.09 RAM-OS (W14393 RAM)
Engine#4 Adaptive Mode / Buffer 4096 / Q1/3/4/5 = 14/1/1/1 / xQ1 = 1 / Processor Core Appointment Scheme = Core 1-3 / PlayerPrio = Low / ThreadPrio = RealTime / ClockRes = 1ms / Not Switch during playback = off / Xtweaks Balanced Load = 43 / Nervous Rate = 100 / Cool when idle = 1 / Provide stable power = 0 / Utilize cores always = 1 / Time Stability = Stable / Time Performance Index = Optimal / SFS 0.90MB Max 120MB / Straight Contiguous / Include Garbage Collect = on / Start Playback during conversion = off / Do not start playback at all = off / Copy to XX-drive by standard = on / Always clear Proxy before Playback = on / Stop Remaining Desktop Services = on / Stop Desktop Services = on / Stop Remaining Services = on / Stop WASAPI Services = on / Stop W10 Services = off / Keep LAN Services = on / Persist = off / Use Remote Desktop = on/ Arc Predict / Minimize OS = on / Peak Extend = off / Unattended

Audio Chain
Stealth MachII PC >> Lush USB 1.5m >> NOS1a G3 B75, Driver v1.0.4 (4ms) >> Placette Active Linestage >> 6 way active horn speakers (Single Ended Triodes)
acg
Audio Addict
***
Offline Offline

Posts: 749


View Profile
« Reply #181 on: October 05, 2015, 10:51:50 am »

So ...

Yesterday was day 6 with the LPSU and it is glory all over.
N.b.: Right now we are packing it for shipment to the person who is going to mod it.

I already feel sad and I want one myself ... (the one I used is for a customer).



Yesterday, knowing that my standard round of 5 days passed in well fashion, I thought to try once again what I felt from the start to be better ... Windows 8.

Ohhhhhhh
was I right.

So far, only in between the lines I have told that Windows 10 isn't the better one at all. However, it performs better because it makes more music. But there's a grey-ish "rounding" flavour.

Big fun : with the LPSU this disappears largely. Maybe even completely. "Maybe" because it is quite hard to pay attention when things sound so good in the first place; as I told - nothing jumps out or disturbs and albums play and play.

Back to a week ago : remember that I started out with the LPSU not in mains earth, and that *that* completely failed. I started with using Windows 8 because I felt or hoped or just knew it should be the better one with the LPSU. But it was misery all over and I shoveled in Windows 10;
Was a bit better but still "non-sense" and then I found the mains earth thing (needs to be used). And then Windows 10 stayed in.

Up to yesterday;
The PSU has to go today, so I sure wanted to try Windows 8 once more.
What did I say ?

Ohhhhhhh

This time I received the "total accuracy" together with music.
Big fun again : I am not even so sure the distance with Windows 10 was large, but Windows 8 just plays "better". Really difficult to explain, but the performers are "better" when played through Windows 8.



So there's a difference of night and day between W8 and W10 when the normal switching PSU is used, and there's hardly a real difference between the two when used with the LPSU.

Now I wonder what this tells ...
Anyone with ideas ?

Peter


Thanks for your further comments Peter.  Looks like the ATX LPS is a definite prospect then. 

I have no idea why it brings W8 closer to W10, or vice versa, but it is an interesting thing to speculate.  Are you running both W8 and W10 operating systems from ramdisk without a hard drive of any type in the PC?  I guess I am asking if the hardware is all the same in these comparisons?

Regards,

Anthony
Logged

Audio PC
Stealth Mach II with Xeon E5 2698 v4 20 Core 2GHz  with Hyperthreading On [40 cores]/ 32GB Ram / RAM-OS / mobo USB port

XXHighEnd 2.09 RAM-OS (W14393 RAM)
Engine#4 Adaptive Mode / Buffer 4096 / Q1/3/4/5 = 14/1/1/1 / xQ1 = 1 / Processor Core Appointment Scheme = Core 1-3 / PlayerPrio = Low / ThreadPrio = RealTime / ClockRes = 1ms / Not Switch during playback = off / Xtweaks Balanced Load = 43 / Nervous Rate = 100 / Cool when idle = 1 / Provide stable power = 0 / Utilize cores always = 1 / Time Stability = Stable / Time Performance Index = Optimal / SFS 0.90MB Max 120MB / Straight Contiguous / Include Garbage Collect = on / Start Playback during conversion = off / Do not start playback at all = off / Copy to XX-drive by standard = on / Always clear Proxy before Playback = on / Stop Remaining Desktop Services = on / Stop Desktop Services = on / Stop Remaining Services = on / Stop WASAPI Services = on / Stop W10 Services = off / Keep LAN Services = on / Persist = off / Use Remote Desktop = on/ Arc Predict / Minimize OS = on / Peak Extend = off / Unattended

Audio Chain
Stealth MachII PC >> Lush USB 1.5m >> NOS1a G3 B75, Driver v1.0.4 (4ms) >> Placette Active Linestage >> 6 way active horn speakers (Single Ended Triodes)
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #182 on: October 05, 2015, 11:22:00 am »

Hi Anthony,

W10 runs fully from RAM. No additional RAMDisk (eh, obviously ? <-- I considered that and the space in RAM *is* there (together with the OS), but I just thought it could worsten only). Only device connected is NOS1a and LAN cable.
Video card is in.

W8 Runs from 2.5" HDD. XXHighEnd and Playback Drive on RAMDisk. Only device connected is OS HDD, NOS1a and LAN cable.
Video card is in.

Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #183 on: October 05, 2015, 11:55:59 am »

It will be way too much of stipulating, but I can think of this :

Earlier on (week ago) I thought of the Linear Supply being "slow". I did not measure that and actually I don't know, but I derive it from common good behaviour of large buffer capacitors we tend to use in power supplies, which "obviously" are always linear.
And in the switching supply the are not there (only a relative small one).

Now let's sort of back fire this :

When the SMPS is very fast and sufficiently fast, where is it "able" to get its juice from ? ... from the mains itself.
This means that it surges the mains in spikey fashion.
Maybe our DAC does not like that so much. Maybe our amplifier doesn't.

With a trained eye I can see how the playback software behaves, although the speed of the power meter is only 2 samples per second. When all is tuned in well fashion (see my sig) then the variation can be 0.5W only.
(side note : d*mn it, I now realize that I forgot to look at the LPSU through the power meter - and now it is packed already. Grrrr !)
The 0.5 W emerges when an SFS round is in order (load a piece of track into memory).

That is the very ROUGH measurement. Super rough, because with 2 samples per second it just *is* super rough.

The meter shows the usage; it does not show any surge (or surging). So I mean : it does not show dropping of voltage or whatever. I just don't know that.

While over the longer term I see a variation in Wattage (0.5), in reality this will be way way waay spikey. So, would I be able to sample 100.000 times per second, then at each sample I would see a difference (in usage). And well, if you ask me, nothing in there is speedy enough to follow (thus surging occurs). Of course no power plant a 100Km away is following my spiky needs, and all what's in there is the "elasticity" of the length of the flow of electrons (which also is 100 Km - or maybe it's from my door step to the stereo only) which could be named "capacitance" I think.

With my back fire I imply an upside down operation to begin with, but I will do that again by stating that no SMPS is spitting noise on to the mains ... no ... it is surging spikes from the mains.

While a week ago I said that the LPSU possibly was too slow (because of not enough capacitance storage), I obviously said that because it sounded bad. But that is history now, and so I put that the other way around : it has so much sufficient capacitance that the mains won't notice much of it, except for the very irregular (randomized) re-charge of all the capacitors (which are 40-60 or so (I still did not count really)).
It's a smooth operation.

The NOS1a is 100% isolated. Still the Q1 etc. stuff makes a difference. Actually crazily easy audible, just because everything went better again (over a year ago I could not hear differences).
So as I often said already : the influence must be creeping in (into the DAC and/or amplifiers) via the backdoor : the mains.

I don't need to say more, because I just already said it all.
But do notice that the influence via the SMPS is almost direct (hardly any buffer) while the influence via the LPSU is very very indirect (buffers all over the place).

Nice.
Now I'm at the beginning of having some clue, I don't have the LPSU any more. I suppose I must get me one. swoon

Peter


PS: No, I never touched any SQ dial after things started to work (connect to mains earth). So I also have no idea whether any influence is still there. All I know is that I did not want any influence. Same as that I never turned off *any* album for 5 days (only yesterday I played random tracks because I wanted to hear as many varying stuff as possible). And never turning off any album in 5 days of time ? no, that seriously never happened to me.

Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #184 on: October 05, 2015, 12:03:56 pm »


One more thing, because you may not have noticed, or I did not rave about it sufficiently enough :

I regard myself to be the loudest shouting person on the Net that a Linear PSU for an audio PC, WORKS. It works so d*mn good that now I don't understand why nobody is shouting about it. And yea, Mani, I know, you "admitted" that it works for you too. But you are the same as the so (really so) many others who obtained the Teradak - initially you are just too quiet. Or anyway quiet enough to not remember any hurays.

Personally I think it is the largest leap by one appliance I ever experienced.

Now I start to be afraid of how tonight's sound will be. I got used to something and ... scratching

Regards,
Peter


PS: But it is true ... I too put the results of the LPSU somewhere in between some lines. I am not sure why, but I decided for it explicitly. yes
Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
acg
Audio Addict
***
Offline Offline

Posts: 749


View Profile
« Reply #185 on: October 05, 2015, 12:50:29 pm »

Peter, are you able to experiment with capacitor banks after the SMPS?  Those low voltage caps are pretty cheap.

I really like your thinking regarding capacitance and mains surges, but don't have time tonight to explore it further.  One thing to think about is that there are people that have been playing around with LPS for some time that suggest that a choke filtered supply is the best of the best.  Probably something to do with the shape of the ripple and the choke filtering out the higher frequencies in the noise.

Another thing to think about are snubbers between the transformer and rectifiers.  I have not had a look inside the NOS1a to see if these exist, and I certainly don't know if they are employed in the Teradak, but transformer ringing may be quite an issue for the low noise performance of the NOS1a...I was going to experiment with snubbing down the track but I thought I should mention it now.
Logged

Audio PC
Stealth Mach II with Xeon E5 2698 v4 20 Core 2GHz  with Hyperthreading On [40 cores]/ 32GB Ram / RAM-OS / mobo USB port

XXHighEnd 2.09 RAM-OS (W14393 RAM)
Engine#4 Adaptive Mode / Buffer 4096 / Q1/3/4/5 = 14/1/1/1 / xQ1 = 1 / Processor Core Appointment Scheme = Core 1-3 / PlayerPrio = Low / ThreadPrio = RealTime / ClockRes = 1ms / Not Switch during playback = off / Xtweaks Balanced Load = 43 / Nervous Rate = 100 / Cool when idle = 1 / Provide stable power = 0 / Utilize cores always = 1 / Time Stability = Stable / Time Performance Index = Optimal / SFS 0.90MB Max 120MB / Straight Contiguous / Include Garbage Collect = on / Start Playback during conversion = off / Do not start playback at all = off / Copy to XX-drive by standard = on / Always clear Proxy before Playback = on / Stop Remaining Desktop Services = on / Stop Desktop Services = on / Stop Remaining Services = on / Stop WASAPI Services = on / Stop W10 Services = off / Keep LAN Services = on / Persist = off / Use Remote Desktop = on/ Arc Predict / Minimize OS = on / Peak Extend = off / Unattended

Audio Chain
Stealth MachII PC >> Lush USB 1.5m >> NOS1a G3 B75, Driver v1.0.4 (4ms) >> Placette Active Linestage >> 6 way active horn speakers (Single Ended Triodes)
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #186 on: October 05, 2015, 01:52:36 pm »

Anthony,

Actually it is hard to see how filtering the mains (supply) is going to work out. Did you ever look at it (scope) ?
It's monsters in there. Well, over here it is.

Look below; that's not going to be filtered out by anything. It is very regular though.
But careful, this is fooling you because it is 50Hz pick-up through air. It is from this topic : Our great mains. If you scroll down a little, you see this picture coming by.

Notice that *because* of this low level (200mV peak instead of 220V) you can see what is going on. So it is my estimate that on the way larger scale these (regular) excursions are the same, but then relative to 220V (and you won't be able to see them). Anyway, this can be looked at as 50Hz noise (noise with a frequency of 50Hz). Not such a good idea to filter that out ... haha.

I don't know.
Assumed that such a thing (and worse) is reality, this still implies "an RMS voltage" (mind the RMS). It also will imply a very same cycle-to-cycle same RMS voltage because the distortions are completely equal (well, seen at this level). However, when we "surge" somewhat, the voltage will drop ... (and this maybe 100,000 times per second and whatever oscillation (new frequencies) may occur together with the noise you see (below).
Don't ask me, but possibly a mV or even way less of drop on 220V (btw officially 230 and in your country most often 240 I think) *is* audible.

Now you can choke that but we're not even talking about noise as such (and the noise you see can not be filtered IMO). Can it be "snubbed" ? perhaps. But it looks like a too difficuly way to attack something which could be attacked by ...

Quote
Peter, are you able to experiment with capacitor banks after the SMPS?

well, that. So *if* I am right, simply that would be the solution.
The other solutions should merely be about the "spitting noise" again, and while we (or I) always thought about that as being a culprit, look again at the picture. How can it get even worse ?
lol

Anthony, I think it was before your time in this forum, but at some stage many people bought a PCI card with a capacitor bank on it. Various capacity banks actually. I did too. Here : PCI supply rail noise. I forgot what was (sounded) wrong with it, but I rather quickly removed it. The reason will be in that topic somewhere (what about close to the end).
Anyway, it mattered for sure and possibly (as I can reason out only today) these caps were (dis)charging too slowly (they're not under load really). So it is blob here and blurp there and with all of them together just *because* it happens to slow, it wasn't random enough. Say a 100 times per second with all of them, and then not randomly (5 "fire" withing 10us of each other, then a few single with not equal intervals, etc.).
And of course back at the time we only thought about influence via USB. Per today they just surge the mains. Haha.

It's all a bit of a crazy thought.

Peter


PS: No AC snubbing in the NOS1a (but try if you like).

PPS: Don't forget what we look at below. This is through air. So that in itself is bad of course. But just practice (over here at least).



* Mains03.png (17.9 KB, 955x573 - viewed 1000 times.)
Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
manisandher
Crazy Audiophile
****
Offline Offline

Posts: 2070

from-first-principles.com


View Profile WWW Email
« Reply #187 on: October 05, 2015, 07:00:12 pm »

I regard myself to be the loudest shouting person on the Net that a Linear PSU for an audio PC, WORKS. It works so d*mn good that now I don't understand why nobody is shouting about it. And yea, Mani, I know, you "admitted" that it works for you too. But you are the same as the so (really so) many others who obtained the Teradak - initially you are just too quiet. Or anyway quiet enough to not remember any hurays.

Nah Peter, you're wrong. I didn't shout about the Teradak because there were so many other changes in my system at exactly the same time and I couldn't attribute the obvious increase in SQ to just the LPS. Although I didn't have much time to listen to the system over the summer (just after all the changes), one thing I was sure about was that I was finally content with the sound I was getting. I mean, how long did it take me to upgrade to 2.02?

Oh, and I'd been using a Paul Hynes linear power supply with the PC I was using to make recordings (with the PMII) for I suppose over 4 years. At the time, I definitely shouted about the difference between the PH LPS and a SMPS.

Mani.
Logged

Main System:
Phasure Mach III (Win 14393.0 on RAM-OS / controlled by RDC, / connected directly to music server / XXHighEnd 2.10 / Minimize OS / Engine#4 Adaptive / DB=4096 / Q1=10 / xQ1=15 / Q3,4,5=1 / SFS=0.69 / XTweaks = 34, 10, 0, 0, 0 / Straight Contiguous / Clock Resolution = 15ms / Scheme 3-5 (low/realtime) / 8x Arc Prediction / switch #5 'up/off' / Unattended) mobo USB3 port -> Lush^2 -> Phasure NOS1a B75 G3 -> 8m Blaxius^2 -> First Watt F5-cloned mono amps -> Tune Audio Anima horn speakers
Office System:
Phasure Mach II (with Nick's clock upgrade) -> Lush^2 -> Phasure NOS1a G3 -> BD-Design BD30-SPR mono amps -> Tune Audio Marvel horn speakers
CoenP
Audio Addict
***
Offline Offline

Posts: 812


View Profile Email
« Reply #188 on: October 05, 2015, 10:45:14 pm »

Hi,

I would definitely NOT put a capacitor bank after the SMPS. Only the wiring to and from the bank in conjuction with capacitor parasitics would make a resonance horror scenario. The SMPS is certainly NOT designed to take a large capacitance as load. I think that exactly happened with the 'capacitor buffer' PCI card that totally worked counterproductive.

Large banks are ok for dc smoothing in linear supplies provided that the transformer and diodes can keep up charging them. Mind you that all regulated supplies assume a distinct capacitance at their outputs for proper operation. That means that sq change as result of extra capacitance at the regulators output is likely yet an improvement Less so.

This is not a new territory. In the pre cambrium of computing all computers used only linear supplies. You can still find the big, big, big 'computer grade' capacitors you cound at the dumpstores. Also overclockers are fond of linear supplies because they are rumoured to allow for even more cpu juice at lower temperatures.

I've personally made a simple wiring change in the computer that made my LPSU (HDPlex) setup leap forward in SQ to an unsuspected level. I'd rather focus on the aspect of proper currentloops and creating the lowest noise by optimizing wiring than extra capacitance.

Regards, Coen

P.s. i found win8 and w10 also to be quite similar, but that was on the same pc with identical parts and setup. Yet I still favour win10 over 8, and win7 over them both.

Logged

Settings: Qn: , SFS: , timeres: XT tweaks: , buf: 4096, driver: 8 ms,

Audio PC (jan 19): XXHE PC v1 with RAMdisk w.o. videocard and 1 of 2 cpu fans + BRIX/USB3 storage musicserver. ETN to Fibre converters (linear supplies), 500m SFP modules & 5m OM4 cable. Power cable PE not connected, together with nos1 and poweramp in separate "audio" powerstrip.

Clarixa set + Intona (or Lush 1m), Phasure NOS1a-75B G3 USB (buf 16 ms)-> Blaxius ->SE EL95 (0,8W triode) + cheap link to Abaqus 300W plateamps> Bastanis cable-> Bastanis Sagarmatha Duo ("DIY").

[other sources: TD124/3009SII-i/Grace F9/lounge LCR phono; Rega Planet 1997 vintage]
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #189 on: October 06, 2015, 09:53:44 am »

Thank you for your input, Coen.

Btw, I think you purposedly mentioned the capacitor bank AFTER the SMPS, which in my view would be in the DC path. But I envision something "in front of" (AC) and now I am not even sure whether that can work.

Anyway it should be clear that if the SMPS responds too fast to our likings, that nothing will change about that when the smoothing happens after it, hence now looking backwards : from the demand (MoBo etc.) towards the supply. All what will happen in that case is something for the far worse, assumed that the caps imply larger surges (at recharging) than the MoBo (etc.) does (though via many small "input" caps there, of course).

Otherwise I did not bring up the idea really, because I feel it just won't work because all is too slow now. However, not sure at all. But envision a processor in idle state (say consuming 20W) which within 0.000001 (just a very small number) seconds goes to 100+W. The SMPS is made for that quite explicitly with all kind of sensing whether it can cope (and when not, the SMPS shuts down the system (roughly put : cuts the power)).
I think this is also why such a supply runs at a high "sampling" rate (control loop) of 100KHz or whatever, so it can respond fast enough (with the notice that I really don't know much about it - which you already noticed Happy).

Both solutions (SMPS and LPSU) work, but when with the SMPS the large buffers are in between the supply itself (say that control loop) and the consumer (MoBo) then all will be quite out of order because the control loop logic will look at the charge rate / demand of the capacitors. And the larger they are, the worse (slower) it will work (for proper control). It is just not made for that kind of operation.

So put the caps in front (AC) ? maybe that can't work either. I mean, I can imagine that the demand the SMPS can create (with it's speed of doing so) is too high for a too slow capacitor(s). But here again, I just don't know much about the working of the SMPS.

Feel free to debunk this all !
Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #190 on: October 06, 2015, 10:15:49 am »

Something else :

Yesterday the only thing what changed was the LPSU being removed of the chain and the SMPS back in. I did not change anything else, meaning that also the same wall outlet was used, thus with mains earth connected.
W8 also remained.

Horrible far away flat sound.
Luckily I didn't have more than 20 minutes of listening time anyway.

Of course this "test" was important, because it could have theoretically been so that suddenly my PC required mains earth in general (because of whatever other unnoticed change) and the whole rave about the LPSU was a hoax(ed thing).

But I have a very important message for all, again :

Any normal human being, never mind him or her being an audio freak, would right away state by means of this test that the SMPS can go to the trashbin and the LPSU is Walhalla.
Well, non-sense !
All I need to do is put the PC out of mains earth and all will be fairly normal. Use W10 instead of W8 (in the testing throughout) and the difference will be minimal, though the LPSU situation will win.

Anyway, that hard is comparing.

And was I ready ?
No, because while with the SMPS it is W10 which wins, with the LPSU it is W8 although again the difference is marginal.

... And that counts for my system ...
(although often people will be able to mimic the situation as I have it - as history testifies)

Tonight I will be using W10 again in the situation I was used to, and see whteher I can survive without LPSU.

Regards,
Peter

Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
acg
Audio Addict
***
Offline Offline

Posts: 749


View Profile
« Reply #191 on: October 07, 2015, 07:01:59 am »

Well, how did it go Peter?  Can you do without a LPS?
Logged

Audio PC
Stealth Mach II with Xeon E5 2698 v4 20 Core 2GHz  with Hyperthreading On [40 cores]/ 32GB Ram / RAM-OS / mobo USB port

XXHighEnd 2.09 RAM-OS (W14393 RAM)
Engine#4 Adaptive Mode / Buffer 4096 / Q1/3/4/5 = 14/1/1/1 / xQ1 = 1 / Processor Core Appointment Scheme = Core 1-3 / PlayerPrio = Low / ThreadPrio = RealTime / ClockRes = 1ms / Not Switch during playback = off / Xtweaks Balanced Load = 43 / Nervous Rate = 100 / Cool when idle = 1 / Provide stable power = 0 / Utilize cores always = 1 / Time Stability = Stable / Time Performance Index = Optimal / SFS 0.90MB Max 120MB / Straight Contiguous / Include Garbage Collect = on / Start Playback during conversion = off / Do not start playback at all = off / Copy to XX-drive by standard = on / Always clear Proxy before Playback = on / Stop Remaining Desktop Services = on / Stop Desktop Services = on / Stop Remaining Services = on / Stop WASAPI Services = on / Stop W10 Services = off / Keep LAN Services = on / Persist = off / Use Remote Desktop = on/ Arc Predict / Minimize OS = on / Peak Extend = off / Unattended

Audio Chain
Stealth MachII PC >> Lush USB 1.5m >> NOS1a G3 B75, Driver v1.0.4 (4ms) >> Placette Active Linestage >> 6 way active horn speakers (Single Ended Triodes)
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #192 on: October 07, 2015, 11:16:44 am »

Haha

I must be a complete nerd by now ... yes

As said, Windows 10 shows a tad of grey, although not many people will really notice. Anyway I do, and this is already because it's a character in the sound (of all). This disturbs me as I want all to be perfectly (??) neutral.
So yes, it was back to that again, and it is apparant because that was not the case with the LPSU.

The nerd part springs from here :

Just *because* I now knew how W10 can sound, and which is with more snap, not grey, more sprankle, I could see through my own settings possibilities and what to do to obtain that same sound. Sounds easy, and of course I only hoped for it, at first.
I came to this idea after an hour of playing or so, hoping that the (now) disturbing sound would have gone away with things getting warmer (ok, knowing that 20 minutes of playback is sufficient for that).

I was deliberating my purchase of an LPSU, meanwhile thinking about your efforts, Anthony, and how I could survive or again get used to what I had been used to for by now close to 3 months.

My SFS had been at 12 for the past weeks, coming from my more common 4, but remembering what ~ happens when I'd dial in 0.2. I try this some times, because I found that the Wallpaper Coverart shows normally, which otherwise (SFS at 4 or higher) is sometimes problematic in the RAM version of W10. Yes, somehow.
But that 0.2 never lasted for long although I forgot a bit exactly why (but it can last 30 seconds before sounds gets stable (not "scratching") after starting playback).

Well, I did that know too (set to 0.2), "hearing" that it would resemble the LPSU better, in advance. And God knows why - that worked. No grey anywhere.
Again a 45 minutes later I changed it to 0.1, having in mind the limit (which is for me 0.06 or so).
Haha, way better again (and I never before tried 0.1 or lower with W10).

Watch out please, because I only listened to one certain type of music, and it doesn't tell a thing about more normal music, like with voices etc. (they could be extra nasty). So this again needs more time.

The day after, I could try to reason out what actually happens, and it even seems possible to tell something about it. That is, since I the past few days was able to sort of figure out why the LPS sounds better than the SMPS, assuming the mains itself as the culprit (and not the computer interface or PC itself with now "better" power). So :

Well, when I was correct in my surging and too much spiking on the mains - the SPMS implying it, then lowering the SFS from 12 (or 4) to 0.1 is a simple 120 (40) times less spikey or 120 times more smooth if you like.
Point of course is that this reasoning equals that other reasoning (read : more chance to be correct with it).

This way of thinking is new, at least for me it is. I mean, so far it has always been about a too low SFS is skipping samples and things become more sharp (and pinpointed !) because of it, though while this is true, it is reasoning like a drag because this can only hold true for the very lowest SFS sesttings (say 0.08 and lower) because you start to hear it (I do clearly at 0.05 IIRC, although this is W8). But nobody is going to tell me that at the change from 4 to 12 to 60 you perceive a more and more "dark" sound because of less and less skipping samples. Thus no real explanation (up to the very end, I mean) in the losing samples reasoning.

This surging thing ... I like it and indeed it will work all the way of the SFS range (say from 0.1 to 470). It is only that once the SFS is somewhat higher (like 2) it becomes quite impossible to reason out what the OS is doing internally, hence is going to overrule my own software for being spikey.

Regards,
Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
acg
Audio Addict
***
Offline Offline

Posts: 749


View Profile
« Reply #193 on: October 08, 2015, 10:48:07 am »

As said, Windows 10 shows a tad of grey, although not many people will really notice. Anyway I do, and this is already because it's a character in the sound (of all). This disturbs me as I want all to be perfectly (??) neutral.

I notice that "grey" and it is one of the things that I want to remove with my playback...it needs to be gone!

The nerd part springs from here :

Just *because* I now knew how W10 can sound, and which is with more snap, not grey, more sprankle, I could see through my own settings possibilities and what to do to obtain that same sound. Sounds easy, and of course I only hoped for it, at first.
I came to this idea after an hour of playing or so, hoping that the (now) disturbing sound would have gone away with things getting warmer (ok, knowing that 20 minutes of playback is sufficient for that).

I was deliberating my purchase of an LPSU, meanwhile thinking about your efforts, Anthony, and how I could survive or again get used to what I had been used to for by now close to 3 months.

My SFS had been at 12 for the past weeks, coming from my more common 4, but remembering what ~ happens when I'd dial in 0.2. I try this some times, because I found that the Wallpaper Coverart shows normally, which otherwise (SFS at 4 or higher) is sometimes problematic in the RAM version of W10. Yes, somehow.
But that 0.2 never lasted for long although I forgot a bit exactly why (but it can last 30 seconds before sounds gets stable (not "scratching") after starting playback).

Well, I did that know too (set to 0.2), "hearing" that it would resemble the LPSU better, in advance. And God knows why - that worked. No grey anywhere.
Again a 45 minutes later I changed it to 0.1, having in mind the limit (which is for me 0.06 or so).
Haha, way better again (and I never before tried 0.1 or lower with W10).

Watch out please, because I only listened to one certain type of music, and it doesn't tell a thing about more normal music, like with voices etc. (they could be extra nasty). So this again needs more time.

The day after, I could try to reason out what actually happens, and it even seems possible to tell something about it. That is, since I the past few days was able to sort of figure out why the LPS sounds better than the SMPS, assuming the mains itself as the culprit (and not the computer interface or PC itself with now "better" power). So :

Well, when I was correct in my surging and too much spiking on the mains - the SPMS implying it, then lowering the SFS from 12 (or 4) to 0.1 is a simple 120 (40) times less spikey or 120 times more smooth if you like.
Point of course is that this reasoning equals that other reasoning (read : more chance to be correct with it).

This way of thinking is new, at least for me it is. I mean, so far it has always been about a too low SFS is skipping samples and things become more sharp (and pinpointed !) because of it, though while this is true, it is reasoning like a drag because this can only hold true for the very lowest SFS sesttings (say 0.08 and lower) because you start to hear it (I do clearly at 0.05 IIRC, although this is W8). But nobody is going to tell me that at the change from 4 to 12 to 60 you perceive a more and more "dark" sound because of less and less skipping samples. Thus no real explanation (up to the very end, I mean) in the losing samples reasoning.

This surging thing ... I like it and indeed it will work all the way of the SFS range (say from 0.1 to 470). It is only that once the SFS is somewhat higher (like 2) it becomes quite impossible to reason out what the OS is doing internally, hence is going to overrule my own software for being spikey.

Regards,
Peter


I love the way your mind works Peter.  You hypothesis seems valid on the face of things and it will be interesting to see where it goes.  The fact that you can ameliorate much of the influence of the LPS with software is exciting.

With the NOS1a noise should not be getting in via the USB cable, but could potentially airborne or via the mains or both.  But USB cables make a difference (this one I cannot figure out)!  Is there potentially another mechanism at work here via the USB, or are we hearing the effect of the USB via the computer back to the other modes of noise transmission?

 
Logged

Audio PC
Stealth Mach II with Xeon E5 2698 v4 20 Core 2GHz  with Hyperthreading On [40 cores]/ 32GB Ram / RAM-OS / mobo USB port

XXHighEnd 2.09 RAM-OS (W14393 RAM)
Engine#4 Adaptive Mode / Buffer 4096 / Q1/3/4/5 = 14/1/1/1 / xQ1 = 1 / Processor Core Appointment Scheme = Core 1-3 / PlayerPrio = Low / ThreadPrio = RealTime / ClockRes = 1ms / Not Switch during playback = off / Xtweaks Balanced Load = 43 / Nervous Rate = 100 / Cool when idle = 1 / Provide stable power = 0 / Utilize cores always = 1 / Time Stability = Stable / Time Performance Index = Optimal / SFS 0.90MB Max 120MB / Straight Contiguous / Include Garbage Collect = on / Start Playback during conversion = off / Do not start playback at all = off / Copy to XX-drive by standard = on / Always clear Proxy before Playback = on / Stop Remaining Desktop Services = on / Stop Desktop Services = on / Stop Remaining Services = on / Stop WASAPI Services = on / Stop W10 Services = off / Keep LAN Services = on / Persist = off / Use Remote Desktop = on/ Arc Predict / Minimize OS = on / Peak Extend = off / Unattended

Audio Chain
Stealth MachII PC >> Lush USB 1.5m >> NOS1a G3 B75, Driver v1.0.4 (4ms) >> Placette Active Linestage >> 6 way active horn speakers (Single Ended Triodes)
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15476



View Profile Email
« Reply #194 on: October 09, 2015, 10:09:36 am »

Hi Anthony,

Well, I like the fact that you are able to think along my ways. It encourages for some sparring (better : try to give it progress).

In about one sentence you suggest two important things for the possibility of explanation :

1. A similar noise pattern that may escape via the mains backdoor may also show its presence through radiation at the isolator (in-NOS1a);
2. A dedicated USB noise pattern may fold back (reflect back) into the PC environment and from there travels the mains path.

I tried to think of 3 and 4 (and 5 ?) as combinations of 1 and 2, but I think 1 and 2 contain all possibilities.

The point by now is, that once one is able to dig out the nuances and where to be for them, it gets more and more wild, but which for me means that I less and less understand what can be going on;

We can have "music" through our speakers or we can have super stellar sound (Q) exhibitions which are unheard.
Ok, that was not clear. Rephrase :

When we, like last week, go to an audio show and play all the in advance prepared tunes - which for me means a couple of years of collecting them - then none of them work. It just does-not-want-to-work. This means that none of it should be played, because it is total "nonsense". I use this phenomenon more often, because when you play something which doesn't seem to make sense for the reason of playing it, it is nonsense. Mind you, I am not referring to Jazz in the Pawnshop sh*t, which has its own life for some, and which will too play better or worse. But the worse will not be much noticable unless it hurts or something. Anyway, such a type of music always makes sense to some degree. And that I don't like a clarinet - can happen.

My "music" is that type of exhibiting a thriller. It was decribed by a visior of ours, a few months ago; it was the more rare occasion that we both enjoyed the exact same feeling and senses (bit of pun) while you actually are unable to see how it is done. It is so complex that it feels like it could only have been made possible in the artist's mind very explicitly, who next engages a bunch of synthesizers and hears back his explicit creation, but only because he knows what it oughts to do or be. Another one would not hear it because it is just too difficult to express it by the playback chain. It now is plain stupid stuff although some may bang their head on it.

In order to emphasize even more what technically is going on, without us being able to really see it : At that show we at first perceived sheer very explicit standing waves in parts of the room, especially in the back. You actually could not be there, that bad. But we knew that room and it could not be. So the analyser was pulled out and I saw an 11dB or so of too much noise. Mind you, still at ~ -130dB and it was totally inaudible (200W amps on 115dB sensitive speakers - no preamp or analogue attenuation).
In 30 minutes of time that extra noise was reduced to 5dB more than normal (grounding things a bit diferently) and I thought it was good enough.
No single standing wave anywhere any more and my partner in crime, Bert, could not believe that reducing that noise was able to diminish that enormous amount of standing wave. But it just did and without the experience you indeed are not able to believe it. But it is my hobby horse since about the start of XXHighEnd and a sheer measure. For me.

And still my demo material made no sense.
I just could not play it and instead we (ok Bert) reverted to Jazz in the Pawnshop.
What a drag ...

With some interruption of a Linear Power Supply, back at home all was joy all over, except for the first day when a connection to mains earth (or not) made the same difference of the world (but even worse than at that show).
Rephrase : best ever, once things were connected to the likings of the stuff itself. Happy

Now the LPS is out, and of course I am sad; my reference has changed and I want to go back to that. An ever existing dial is changed and it seems even to have gotten better than with the LPS.
I don't need to play "demo" stuff to enjoy the meaning of ALL what I play randomly. Instead, all what I play just goes to the demo gallery, because it is so crazily good. And amongst them albums I sure played before and which were no good at all previously. Other people would ditch those - I save them for maybe better days.

So yesterday I had enough of listening to the very best and thought to change my SFS of 1.0 to 0.09. I could bare some ...
Restarted the track I just before finisned with the old settings, and already could not recognize it any more.
It got BETTER again.

Why this long story, seemingly unrelated ?
Well, because it is very related. swoon
About 18 months ago I create something with a silver or golden "a" and whatever I do in XXHighEnd, it does not change SQ. Today ? today it seems impossible that I did not perceive changes back then. But we must be careful, because since the beginning of the NOS1a, these things have changed :

1. New Blaxius Interlinks, including a full 75 Ohm impedance path;
2. New Clairixa USB cable;
3. The NOS1 I used for my "a" version was new and has a burn-in of 18 months by now;
4. New digital filters;
5. Windows 10;
6. OS fully in RAM only and no single device connected to the PC;
7. I experienced/learned a new reference (by means of the Linear Poser Supply), so I know what I can theoretically head for.

The effect ? well, where just over a year ago I needed to use some track with a seagull to let it fly through the room, now it hardly matters what I slide in. It is one big amazing experience.

And then after two says with some 1.0 SFS setting I try 0.09 and I can just as well start all over.

Ah, so what was I trying to say ?
Well, that unmeasurable things which even theoretically should not be there, make a difference which seems to be
impossible.

Radiation over an isolation ? can be. But if so, then apparently it can be influenced so wildly that something else (???) is even more wildly influenced in a fashion that is beyond me.

Surging noise on the mains, which is 230V to begin with and then influenced by a processor etc. which is performing a few instructions ? It is all veeeeery nice talk, but one must be totally crazy to believe in such a thing in the first place.

Still it is so that once I can make something consistent, it actually always leads to something for the better (sound) again. So I have to be the believer in the impossible and just go ahead with it, for even a year if required. Thus :

Nothing changed to my PC setup and if I today feel that it is mains surge what is happening, then a year ago that was happening just the same. However, the fact that I can assume this (surging) is happening is new, and with that knowledge I may be able to proceed and make progress again.

So now it is time for some software changes; long time in this realm, actually ...

Regards,
Peter

Logged

For the Stealth III LPS PC :
W10-14393.0 - Oct 31, 2018 (2.10)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 10/20 cores) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 140.19  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = 35 / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^2*A:B-W-Y-R, B:B-W-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
Pages: 1 2 3 4 5 6 7 8 9 10 11 12 [13] 14 15  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1 RC2 | SMF © 2001-2005, Lewis Media Valid XHTML 1.0! Valid CSS!
Page created in 0.058 seconds with 19 queries.