Low cost high quality excellent performing LPSU for audio.

[guiltyboxswapper]

8 minutes ago, plissken said:

LOL. Oh  boy. The power supplies in servers are very well engineered and low noise. Now there are MANY DC rail direct options for servers and network switches were we have house DC in the data center. But this is to save $$ because AC/DC conversion is not a free process.

 

Yawn.  

[guiltyboxswapper]

Just now, plissken said:

I'll offer you the same $8000 to your $1000 that you can't tell the difference when blinded that I've offered to the many...

 

Yawn.

[plissken]

4 minutes ago, guiltyboxswapper said:

 

Yawn.

 

LOL. It's so easy to cleave zealots from their unsubstantiated claims. 

 

Remember claims made without evidence are just as easily dismissed without evidence. There's your door sir.

[The Computer Audiophile]

20 minutes ago, plissken said:

 

I'll offer you the same $8000 to your $1000 that you can't tell the difference when blinded that I've offered to the many...

As part of the separation of objective and subjective content here, made mandatory by posts like yours, I made clear posts like yours don’t belong in this thread. 

[plissken]

Just now, The Computer Audiophile said:

As part of the separation of objective and subjective content here, made mandatory by posts like yours, I made clear posts like yours don’t belong in this thread. 

 

This is all about subjective abilities Chris. It's the entire reason I'm willing to pay out $8000.

[jabbr]

26 minutes ago, guiltyboxswapper said:

 

Many on this forum can tell you otherwise.  Guess we all have "faulty equipment".  I can live with that :-) 

Wouldn't you want the best equipment? Don't you want to configure your equipment correctly?

[guiltyboxswapper]

16 minutes ago, plissken said:

 

LOL. It's so easy to cleave zealots from their unsubstantiated claims. 

 

Remember claims made without evidence are just as easily dismissed without evidence. There's your door sir.

 

There's plenty of observations that those who used an Sonore OpticalModule can hear the difference between simply changing its power input between an SMPS and a decent linear PSU.  That converts copper ethernet directly to fibre.  Guess we're all hearing things........

 

So yes, there's my door indeed - what else is there to say?  

[plissken]

Nothing. My point has already been made... Also you are talking about the PSU on the playback equipment side. Not the server side. DTE vs DCE.

[Superdad]

49 minutes ago, plissken said:

The problem is this is 100% bullshit. There are crap SMPS just as there are crap LPS's.

Ah, the ever profane and incorrect Mark Brown. 

Yes there are a lot of LPS’s whose performance is poor for various reasons—typically high output impedance.

But @airguitaris correct to state that SMPS typically have much higher output noise than a decent LPS.

 

Proof? Let’s take this to the extreme with the example of the absolute lowest output noise SMPS available as far as I know: The Daitron RFS50A: https://www.daitron.com/documents/RFS50A_Catalog_En.pdf

Most modest-current SMPS units will specify about 80mV (80,000 microVolts) of noise. The wonderful Daitron offers 1mV (1,000uV) output noise. That is indeed outstanding for an SMPS, but in 250 unit OEM purchase quantity the thing is still, last I checked, $302.

Building an LPS down into the single-digit milliVolt range is trivial (though as discussed, DC output noise is not the most important criteria for an “audiophile” LPS—or at least not the only one.

However, advanced LDO regs such as the LT3045 make producing supplies with output noise in the single or teens digit microVolt range pretty straightforward.  Of course there are numerous other PS design considerations—and the LT3045 maxes out at 0.7A (most designers pile a bunch in parallel for more current; or choose from the wide world of higher current LDOs still in the microVolt noise range).

[jabbr]

1 hour ago, Superdad said:

Proof? Let’s take this to the extreme with the example of the absolute lowest output noise SMPS available as far as I know: The Daitron RFS50A: https://www.daitron.com/documents/RFS50A_Catalog_En.pdf

Most modest-current SMPS units will specify about 80mV (80,000 microVolts) of noise. The wonderful Daitron offers 1mV (1,000uV) output noise. That is indeed outstanding for an SMPS, but in 250 unit OEM purchase quantity the thing is still, last I checked, $302.

Building an LPS down into the single-digit milliVolt range is trivial (though as discussed, DC output noise is not the most important criteria for an “audiophile” LPS—or at least not the only one.

 

I don't know about that, but my Mellanox MSN2700 switch has its own PSU ith AC power inputs -- there isn't the possibility of swapping out a PSU.

 

The unit obviously operates at the extreme low noise tolerances of 100 Gbe and newer units, also with an AC power cord go to 200,400 and soon 800 Gbe. You can't even measure the noise on those lines without equipment that no one here has. Mellanox does almost $1 Billion in yearly sales and was acquired by NVidia for $ 7 Billion. They have resources.

 

At $30K + a switch they can afford to put whatever the best PSU is in the switch. Its not purely LPS vs SMPS and your/our beloved LT3045 chips can be used to smooth out any SMPS jaggies before they hit the ASIC. You use a similar technique in some of your PSUs...

 

This equipment is rather far away from my audio system, and within my Equi-tech protected audio zone, only LPS are allowed. Similarly in my basement I plug AC cables into my servers which have ATX supplies and don't give this another thought.

 

Subjective experience? Who here has tested intel vs Solarflare vs several speeds and generations of Mellanox for "SQ"? Who has compared 10gbe vs 40gbe vs 100gbe switches for "SQ"? (not directed at you @Superdad just a general statement for anyone reading the thread)

[plissken]

50 minutes ago, Superdad said:

Ah, the ever profane and incorrect Mark Brown.

 

How so? You didn't even answer my post (which lets call a spade a spade is your normal operating procedure). So what is the analog volume implementation in my DC-1 DAC again? So I wouldn't go throwing the incorrect rock at other glass houses. It won't end well for you based on the posts I can quote.

 

I simply said there are crap LPS just like there are crap SMPS's. Including the Meanwell you had to finally admit to having egg on your face when you vaunted $302 in part's let it all through Alex Crespi.

 

Well built DAC's won't show any of this on the output where you have two well built power supplies. One being SMPS and the other Linear.

 

BTW I'm not in disagreement that each offers advantages.

[Miska]

17 hours ago, Superdad said:

100 milliVolts p-p DC output noise? No thanks.  (Just to give the lay folk perspective, the output noise of the Linear Technology LT3045 is a few microVolts; 100 milliVolts = 100,000 microVolts.)

 

For example powering a cheap Pro-Ject Pre Box S2 Digital DAC through it's USB interface, from an UP Board NAA powered by the said medical SMPS PSU gives less than 8 µV of output noise from the DAC analog outputs.  While powering the DAC from the supplied wall-wart through the separate power connector instead doubles this noise. to about 17 µV.

 

 

This given that the said UP Board NAA is running an open PCB without any casing. Highest noise components also in this case are 50 Hz harmonics.

 

So yeah, I continue happily powering my NAAs with that SMPS PSU.

 

Or at least I want to see some DAC analog output measurement data.

 

[m5sime]

To provide an answer to the OP - I exchanged quite a few emails with Nick of Longdog audio in the UK (selling through MCRU). I even sent him my Chinese 12v 5A rated PSU I bought from there (he measured it as a favour) to compare with his own.. Under load my Chinese linear PSU was a bit noisy (still not bad in his assessment) The specs of his PSUs are very good. Not bonkers money either.. I would suggest starting there.. 

[jabbr]

10 hours ago, Miska said:

 

For example powering a cheap Pro-Ject Pre Box S2 Digital DAC through it's USB interface, from an UP Board NAA powered by the said medical SMPS PSU gives less than 8 µV of output noise from the DAC analog outputs.  While powering the DAC from the supplied wall-wart through the separate power connector instead doubles this noise. to about 17 µV.

 

 

This given that the said UP Board NAA is running an open PCB without any casing. Highest noise components also in this case are 50 Hz harmonics.

 

So yeah, I continue happily powering my NAAs with that SMPS PSU.

 

Or at least I want to see some DAC analog output measurement data.

 

 

Great point. I power my NAA (NUC) via an "el cheapo" LPS. One could compare that to the SMPS and see if there is any difference. The LT3045 is quoted at 0.8 uV RMS noise and that's just for the chip! So the NAA and DAC hard do say, as they also have their onboard power circuits ... but rather than quoting theoretical noise from spec sheets, the output noise at the DAC would seem the relevant measure. I will say however that the Pro-Ject S2D does to my ear have some sensitivity to the USB power source, moreso than the iFi iDSD.

 

Certainly at the NAA -> DAC end, the PSU does have a reasonable mechanism for affecting SQ, unlike PSUs in my switches where not only I can't swap the PSU but I wouldn't even dream of hacking a PSU  .... it was hard enough to explain to that very nice Israeli engineer that I was putting the switch in my house ... I wouldn't dare go back and explain that I borked the switch by hacking the PSU .... 

[Miska]

8 minutes ago, jabbr said:

I power my NAA (NUC) via an "el cheapo" LPS. One could compare that to the SMPS and see if there is any difference.

 

NUC takes in 19V by default and runs it through it's switching regulators to produce for example the USB VBUS. One reason I prefer NAA boards that take in 5V and feed it through to the USB only through the current limiter chip (sorta mandatory part of USB).

 

[asdf1000]

2 hours ago, Miska said:

One reason I prefer NAA boards that take in 5V and feed it through to the USB only through the current limiter chip (sorta mandatory part of USB).

 

Another great thing about 5Vdc endpoints is they can be powered by USB powerbank!

 

This completely removes the biggest issue of SMPS - leakage current / ground current / ground loop etc.

 

A bigger issue than their output ripple (depending which DAC designer you talk to).

[The Computer Audiophile]

It would be cool to measure my Pi that’s only powered over Ethernet. 

[jabbr]

3 hours ago, Miska said:

 

NUC takes in 19V by default and runs it through it's switching regulators to produce for example the USB VBUS. One reason I prefer NAA boards that take in 5V and feed it through to the USB only through the current limiter chip (sorta mandatory part of USB).

 

These details might be important!

 

Ive been using NUC for the 5ghz Wi-Fi but the up core might be worth a try even though it’s 2.4ghz 

[plissken]

I wonder what a SMPS amp puts out and if that wouldn't swamp a low voltage power supply. Same for even class AB amps.

[Miska]

6 hours ago, asdf1000 said:

Another great thing about 5Vdc endpoints is they can be powered by USB powerbank!

 

This completely removes the biggest issue of SMPS - leakage current / ground current / ground loop etc.

 

A bigger issue than their output ripple (depending which DAC designer you talk to).

 

Well, the medical PSU I'm using doesn't have leakage currents and since it's floating it doesn't have that ground current.

 

That ripple is left-over from the switching DC/DC converter. Usually it is higher frequency than the 50/60 Hz mains hum and thus easier and cheaper to smooth out.

 

One challenge with the LPSUs is to find ones with high enough current output. The medical SMPS I'm using outputs 5V 6A (30W). Which is of course still very low power, compared to the switching PSUs for example for CPUs and GPUs that can put out 100+A at 1.2V supply voltage...

 

[asdf1000]

3 minutes ago, Miska said:

Well, the medical PSU I'm using doesn't have leakage currents and since it's floating it doesn't have that ground current.

 

 

Is that measuring the same leakage current @JohnSwenson writes about ?

[asdf1000]

9 minutes ago, Miska said:

Well, the medical PSU I'm using doesn't have leakage currents and since it's floating it doesn't have that ground current.

 

It has low leakage, not no leakage currents. So you will still have (low) dirty ground currents / ground RF...

 

Whereas a battery / USB powerbank with actually have zero , as it is not connected to mains AC power at all.

 

 

[Miska]

On 1/31/2021 at 9:09 PM, Superdad said:

If the PS that is the source of the leakage has an electrical path to something that is grounded (such as a DAC, preamp, poweramp etc), maybe an interconnect, USB cable, Ethernet cable etc, the leakage current will create a voltage across the impedance of the cable, frequently the "ground wire" or shield of the cable.

 

By the way, if your ethernet cable is grounded, you are certainly doing it wrong!

 

[Miska]

10 minutes ago, asdf1000 said:

It has low leakage, not no leakage currents. So you will still have (low) dirty ground currents / ground RF...

 

Well, it doesn't have ground leakage because it doesn't have ground!

 

 

[asdf1000]

43 minutes ago, Miska said:

 

Well, it doesn't have ground leakage because it doesn't have ground!

 

 

 

Noted but it's leakage currents will find it's way through USB DAC via USB ground most likely?

 

Whereas battery power supply will not have any leakage current at all - of any kind.

 

But then how well USB DAC deals with higher ripple of a powerbank, that's a different question of course.