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

[bodiebill]

Does anybody know whether a passive PoE (power over ethernet) solution like

https://www.amazon.com/Cable-Matters-Passive-Injector-Ethernet/dp/B01J6JP8XE

could also work for AC?

The reason I am asking is that I feed my DIY Stereo Coffee LDR preamp with 12V AC rather than with DC. 

Quoting Stereo Coffee designer Chris Daly:

"The main board provides are rather unique bridge rectifier, comprised of two conventional bridges in parallel

placed inside a third rectifier comprised of 2 mosfets for the negative and two thyristors for the positive.   

Connecting a linear DC supply sadly misses out what the bridge rectifier provides. So a AC 12V supply

applying to the stereo coffee board, is really the best for sound quality."

[bodiebill]

49 minutes ago, dctom said:

I have an i7 NUC end point and i3 itx server - does anyone know of a source of apacer or similar ram in the UK?

 

https://www.mouser.co.uk/Apacer/Embedded-Solutions/Memory-Data-Storage/Memory-Modules-Memory-Cards/Memory-Modules/_/N-9pk2p?P=1yzu785

[bodiebill]

2 hours ago, austinpop said:

1. HDPlex's bigger, badder 800W DC-ATX adapter is a real SQ champ. It sounds significantly better than the 400W DC-ATX adapter. Thanks again to Roy for this tip. While it does take up more space, and really needs a bigger case than the HDPlex H3, it makes up for it in SQ. The other potential concern is inrush current, if you're trying to use it with a lower current rail. In my case, my 19V/10A rail didn't even break a sweat.

 

Could someone explain what are the options to power the HDPlex 800W DC-ATX adapter?

s it possible to use a conventional (nonlinear) PC power supply to do this? Such as my Seasonic 400W fanless platinum PSU?

[bodiebill]

Yesterday I replaced the RAM in my AudioLinux server with HQPlayer embedded with 2x 8GB Apacer DDR4 RAM 2666 (non-ECC). I was sceptical at first. Earlier I compared ECC with non-ECC in my audio pc (endpoint with NAA), and I could not hear the difference. However this time the improvement was immediate and not trivial, which I did not expect for the optically isolated server PC.

 

Then, today I noticed that the server only 'sees' 1x 8GB of RAM. It appeared I did not sufficiently tighten the second module. After doing so, 2x 8GB appeared. However, with two modules the sound was a little less open and musical. So now I run it with 1x 8B, thanking myself for my mistake 🙂

 

Anyone noticed such a difference?

[bodiebill]

I installed the HDPlex 800W DC-ATX converter that arrive today in my low-footprint audio PC. It only uses the first DC-in for 24 pin ATX and 8 pin EPS. (SATA is de-activated in this PC.)

The sound difference is immediate. A much broader spectrum with more dynamics. Alas also more strident (irritants), so I hope that breaking it in for a week or so will make a difference. 

[bodiebill]

16 minutes ago, elan120 said:

Good to know it is making such a positive impact.  What do you use to feed the DC-ATX converter?

 

Currently SBooster BOTW ECO 19V, but I ordered an SR4. (The 800W replaced a HDPlex 250W DC-ATX.)

[bodiebill]

Indeed, here even the SBooster works flawless with the audio PC, although it delivers only 1.85A. But it just has to power the Supermicro mobo with 8GB ECC RAM, a 35W TDP i3 CPU, m.2 optane drive and a fiber network card. No SATA.

I guess the HDPlex only draws what it needs based on this.

The heavy lifting (HQPlayer DSD256 EC7) is done by a server with i7 9700K.

[bodiebill]

I find it hard to tell the actual power usage from the TDP.

Currently I am using an SBooster ECO MkII 19V / 1.75A (!) to power an endpoint PC with a HDPlex 800W DC-ATX converter, Supermicro board, Apacer 8GB ECC RAM, an i3-6100T CPU which is rated at 35W TDP and an Intel PCIe fiber network card. It has worked flawless for months.The PC runs AudioLinux with NAA from an optane M.2 and is otherwise diskless (SATA deactivated).

I have ordered a Paul Hynes SR2 (19V / 2A) to find out if it makes sense to use that instead of the SBooster.

[bodiebill]

Inspired by the last few pages of this thread I experimented a bit with powering my AL+NAA endpoint audio PC which has an HDPlex 800W DC-ATX converter.

 

setup A  😊

SBooster BOTW ECO MkII 19v/1.75A => HDPlex 800W DC-ATX group 1 => ATX + EPS

 

setup B  😐

HDPlex 200W 19V/10A => HDPlex 800W DC-ATX group 1 => ATX + EPS

 

setup C  😀

SBooster BOTW ECO MkII 19v/1.75A => HDPlex 800W DC-ATX group 1 => ATX

HDPlex 200W 19V/10A => HDPlex 800W DC-ATX group 2 => EPS

 

SQ is C > A > B.

 

C sounds more loose and open.

Not sure if it is the LPSU make the difference or their DC cables. (The cable for the HDPlex seems flimsy.)

 

I may try other combinations (swap LPSU's, trying another one, such as the SR4 which will hopefully arrive by the end of the month, or using a 12V LPSU directly into the EPS/CPU).

 

Setup C seemed slightly better (it made me turn up the volume) than the same with LPSU's swapped, but the difference is either small or imaginary.

 

When should we expect the SQ to benefit most? Using the best LPSU for the motherboard or for the (35W TDP) CPU?

[bodiebill]

8 minutes ago, elan120 said:

I have not get the chance to try it myself, but working on it.  I speculate better SQ will be from better/cleaner power to EPS(CPU).  The combination below should reveal either SBooster is better than HDPlex 200W 19V/10A or the other way around, and hopefully, not the same.

 

SBooster BOTW ECO MkII 19v/1.75A => HDPlex 800W DC-ATX group 2 => EPS

HDPlex 200W 19V/10A => HDPlex 800W DC-ATX group 1 => ATX

 

Already tried that, and not much difference, maybe slightly less good SQ, not sure. But if true, it would indicate that the motherboard benefits more from the best LPSU than the CPU? Which would somewhat surprise me.

[bodiebill]

26 minutes ago, rickca said:

If you can buy or make an appropriate cable, try your HDPLEX 200W 12V rail for EPS (not going though the DC-ATX).  I haven't done this myself, but I'd expect the sound quality would be even better than HDPLEX 200W 19V => HDPLEX 800W DC-ATX ==> EPS.

 

Thanks, I do have such a cable and will try.

 

In that case it may even be preferable to use the HDPlex's upper variable LT3045 2A output, which should be cleaner. Or I could go a step further:

HDPlex 15V => MPAudio DLS-HPULN 12V => EPS

to make it even cleaner.

[bodiebill]

58 minutes ago, bodiebill said:

 

Thanks, I do have such a cable and will try.

 

In that case it may even be preferable to use the HDPlex's upper variable LT3045 2A output, which should be cleaner.

 

Now setup as above, powering the CPU directly from HDPlex's upper variable LT3045 12V/2A output, and the dynamics are improved. However a tad strident so I will fabricate a better DC cable to see whether that will improve things further... 

[bodiebill]

38 minutes ago, rickca said:

I doubt that 2A will be anywhere near enough for your EPS12V.  But I see you already tried it!  You must be running a really light workload on that 35W CPU.

 

I'm mystified.  I would expect better dynamics from 12V/10A with lots of extra headroom.  But of course amperage is just one factor contributing to how a power supply responds to load.

 

The modest demands of this PC surprised me also. However it is a Spartan setup with only AudioLinux + NAA running without hyperthreading from an optane M.2, with SATA deactivated. The whole system (ATX + EPS) already ran well powered by the SBooster with 1.75A, so the HDPlex's 2A dedicated to the CPU only should be sufficient. 

[bodiebill]

10 hours ago, bodiebill said:

 

Now setup as above, powering the CPU directly from HDPlex's upper variable LT3045 12V/2A output, and the dynamics are improved. However a tad strident so I will fabricate a better DC cable to see whether that will improve things further... 

 

I made a short DC cable using Supra CAT8 to replace the stock HDPlex cable, and things have improved. Exit tweaking (for now) and enter the music...

[bodiebill]

2 minutes ago, drjimwillie said:

This is interesting and counter to what Austinpop reported. That higher amperage sounded better, than a better quality and less amperage PSU, when powering the CPU.

I wonder if he was powering the server or the endpoint?  Or a one box system?

I wonder if increasing the amperage to the CPU of the server makes an improvement when sending data over ethernet to the endpoint?

 

For a heavy-lifting server I can imagine a lot of headroom will help. The PS supply of my HQPlayer server (Seasonic fanless 400W) for instance can deliver 33A.

 

But maybe it is a question of 'enough is enough'? If the CPU of my minimized endpoint PC only draws, say, 1.5A (as it seems to do), why would it care if the PS is able to offer more?

Of cause that is just rational reasoning which should never overrule our ears. 

[bodiebill]

1 minute ago, drjimwillie said:

Bodiebill- how much of that 33A goes to the CPU and how many watts is the CPU, please?

 

I did not measure the amperage drawn in the server, but the CPU is a 95W TDP i7-7900K.

Before the Seasonic I used the 200W/10A HDPlex into a 400W HDPlex DC-ATX converter, and that worked without problem, apart from excessive heat. As I heard no SQ benefits using the latter LPSU, I went bac to the Seasonic.

[bodiebill]

42 minutes ago, drjimwillie said:

@bodiebill is the 400 W Seasonic currently feeding the 400W HDPlex DC – ATX?

What is the amperage limit of the 12V rail or output feeding the CPU/EPS?

 

No, for the server I only use the Seasonic. This is a regular (nonlinear) power supply, but high quality with low ripple and voltage variance.

I wouldn't know how to let the Seasonic power the HDPlex as the latter needs 19V and the Seasonic only delivers 12V and downwards. 

 

39 minutes ago, drjimwillie said:

Also, are you saying that changing the PSU of the server did not affect the sound quality of the endpoint?

 

For SQ I slightly preferred the Seasonic but the difference was hardly noticeable -- perhaps even imaginary. What I do know is that I did not like the extreme heat of the 200W HDPlex.

[bodiebill]

1 hour ago, Nenon said:

If the CPU draws 1.5A, I would feed it with a LPS that can provide 8A, or it least 6A. That from my experience sounds a lot better. It's what other people have reported too. Oversizing the power supply feeding the CPU makes a difference. Don't ask my why... I have a theory but not worth getting into it. 

 

Thanks Nenon, I tried the same setup except using the HDPlex normal 12V/10A output instead of its (variable) 12V/2V LT3045 output. Indeed I think I prefer the 12V/10A. Better slam, dynamics.

 

Downstream I now need to power:

A. motherboard with 19V

B. CPU with 12V

C. iFi micro iUSB3.0 with 9-12V

 

When my Paul Hynes arrives I will have the following power linear power supplies:

P1. Paul Hynes SR4: 9, 12, 15 or 19V / 2A

P2. SBooster BOTW ECO MkII: 19V / 1.75A

P3. HDPlex 200W: 12 or 19V / 10A or 5, 9, 12 or 15V / 2A (latter LT3045 based)

P4. MPAudio DLS-HPULN (LT3045 based): anything up to 15V / 5A (to be powered by anything up to 17V)

 

In addition I have three Studer900 LPSU's (9v/15W, 12V/30W, 15V/25W) and a nondescript Chinese 19V/?W.

 

Currently:

P2 => A

P3 12V/10A => B

P3 12V/2A => P4 => C

 

For best SQ, what to power with which?

 

Of course I will try, compare and report back, but it would be nice to know your and other's expectations...

 

[bodiebill]

1 hour ago, Nenon said:

This is hard to answer or even give you any recommendation. I have only used the SBooster and HDPlex from your set of power supplies at very early stages of experimenting with music servers. I don't have them anymore and would not consider them in my system again. I have not tried the SR4 or the MPAudio. 

 

Instead of trying to map a power supply to component, let me try tell you what I think in general. There are two factors - power and quality. Typically the best quality power supplies cannot provide a lot of power. 

 

If 'quality' was the only factor, it would have been a really easy recommendation:

You put the best quality power supply you have on the USB. The second best goes to the CPU. The third best goes to the motherboard. 

 

If 'power' was the only factor, it would have been a really easy recommendation too:

You put the most powerful power supply you have on the CPU. The second most powerful goes to the motherboard. The third most powerful goes to the USB.

 

The problem is when you have to balance between those two factors and introduce a dozen other factors - then it gets really complex. You would have to experiment and see what works best. A good starting point for those experiments would be to:

• Put the most powerful power supply on the CPU.
• From the remaining power supplies, pick the best quality one and put it on the USB.
• From the remaining power supplies, try the best quality power supply on the motherboard; but you may also try the most powerful, or something in between. 

Good luck with the tests and let us know what works best. 

 

Oh, and one more note - I am increasingly starting to think that building a computer with multiple rails of the same power supply and wire accomplishes more natural sound overall. Look at my Building a DIY Music Server thread for example of what I mean by that. 

 

Thanks Nenon, that makes sense. I guess experimenting with several setups will tell.

But it looks like this would a good starting point:

- SBooster 19V/1.75A => HDPlex 800W DC-ATX => motherboard

- HDPlex 200W 12v/10A => CPU

- SR4 12V/2A => USB (iFi micro iUSB3.0)

 

Or for the latter, even better (for additional power cleansing) : 

- SR4 12V/2A => MPAudio DLS-HPULN 10.5V/5A(2A) => USB (iFi micro iUSB3.0)

 

Alternatively, looking at your last remark (multiple rails of same PS):

- HDPlex 200W 19v/10A => HDPlex 800W DC-ATX => motherboard

- HDPlex 200W 12v/10A => CPU

- SR4 12V/2A => MPAudio DLS-HPULN 10.5V/5A(2A) => USB (iFi micro iUSB3.0)

 

[bodiebill]

That sounds like a S.M.A.R.T. approach @austinpop !

 

Am I right to suppose that if the Kill-a-watt meter (to be had for EUR 15) measures a maximum of 24W AC (!) power on the 12V CPU line, then the LPSU should supply 2A as an absolute minimum?

 

What makes this measurement difficult? Sounds rather easy to me.

[bodiebill]

16 hours ago, drjimwillie said:

@bodiebill is the 400 W Seasonic currently feeding the 400W HDPlex DC – ATX?

What is the amperage limit of the 12V rail or output feeding the CPU/EPS?

 

I checked the manual: 33A for the 12V rail.

[bodiebill]

7 hours ago, austinpop said:

 

First of all, as @bobfa rightly points out, these numbers are AC watts drawn by the PSU, not the DC watts being delivered to the server. I neglected to mention this measurement is only meant to serve as a rough guide.

 

My point about difficulty was addressed at people who are trying to size a PSU for a system that they haven't built yet. Even if you assemble the system, you still need to find temporary PSUs to run ATX and EPS. I don't know - that seems pretty inconvenient to me.

 

I see your point re inconvenience.

But with an existing system (such as my audio PC) the kill-a-watt meter approach seems like a good sanity check.

 

So I guess with the conversion loss the measured AC power will be higher than the actual DC power supplied. That means that if we calculate amperage required from the LPSU based on the AC power, we will in fact have some headroom as actually less is needed. Correct?

[bodiebill]

23 minutes ago, mourip said:

Given the current (no pun intended) thread here regarding power supplies I have a question... or more of an observation.

 

I have always believed in the importance of well constructed, stiff, over-built power supplies. When modifying or building analog preamps and amps I have always gone for overspec'd transformers, big caps, tight regulation and well thought out grounding. The payoff is usually better accuracy, a more realistic sound stage and especially improved dynamics.

 

For music servers and endpoints I am certainly a fan of linear supplies, good grounding technique, and isolation however once one has met the peak current requirements to avoid sag during high computational needs I really do not understand when someone talks about "improved dynamics" of the actual music. It seems to me that once a good digital stream is produced then the downstream analog components starting with the DACs analog output would determine music dynamics. I can certainly understand how an under-spec'd supply could cause audible issues but I just do not understand how, once the peak needs have been reached, that an over-spec'd supply would improve perceived music dynamics.

 

I am interested in how this might work. Perhaps the subjective experience of improved dynamics comes from finally meeting the actual needs of the server and not from surpassing its real needs?

 

Thanks!

 

Interesting questions/observation. Not that I have an immediate answer, but in the discussions above it is not always clear to me when we are talking server or endpoint, or maybe better: high-lifting PC (such as HQPlayer) or Spartan audio streamer (such as AL, Euphony, NAA). I think this makes a big difference. I do not think there is one PS-related 'truth' for both.

[bodiebill]

20 minutes ago, afrancois said:

In my opinion if the server is powerful enough and behind a fiber connection it doesn't matter how optimized the memory, power supply, etc, is.

 

That partly concurs with my recent experience: I hardly heard a difference powering my HQPlayer server (far away from and fiber-connected to my AL+NAA endpoint) with a nonlinear Seasonic 400W PS vs a HDPlex 200W linear PS. I have chosen for the former as it is more powerful and generates less heat. An ever so slight preference for the SQ with the Seasonic could be the result of this 🙂 

[bodiebill]

1 hour ago, austinpop said:

Yes, please publish your findings here when you get them, along with build details and the exact workload (OS, music player, if HQP - what settings) you used. One useful piece of info that will be useful to others will be the relative consumption between ATX and EPS (CPU). 

 

Will do. I just ordered a watt meter...

 

1 hour ago, austinpop said:

Yes, again - this is not meant to be an exact science. Typically, your PSU should have a capacity several multiples of your observed consumption. But it will tell you, for example the relative weight of ATX and EPS, so you can decide how/where to allocate your resources into the capacity of each rail.

 

Also be sure to observe other infrequent, but intensive workloads, like: library scanning, OS and software updates, for example.

 

And finally, of course, a reminder to look at both peak and steady state.

 

In any case, the wattage required by the PC (or CPU, ATX) will be lower than the AC wattage measured, and similar for the amperage. And that is better than the other way around.

 

And thanks for the pointers; I will try to do a thorough test.