Do you need linear power supply for DACs?

[solderdude]

Most DCDC converters require extra filtering (LC filters) when you want to use them in sensitive circuits. Also these devices often emit common mode spurious signals.

All advantages usually also have some disadvantages.

Care to share what the identifiable artifacts are and which DCDC converters were deemed usable ?
Which brand are you representing ?

 

[signalpath]

Care to share what the identifiable artifacts are

Reminds me of a lab experience from decades ago. We were doing bench listening tests to an audio circuit. We were using an HP bench power supply. We later swapped the HP for a Lambda supply -- identical E and I-limit -- and the tonal character of the DUT was shifted. That opened my eyes to the impact a power supply can have on audio circuits, and has been integral to product testing ever since. Do I understand the objective reasons for these supply-related tonal shifts? Maybe a little (impedance issues - reactance, loop gains, variation in load-switching behavior, filtering artifacts, etc.). But mostly it remains a trial and error process of finding the most accurate systemic sound.

Care to share which DCDC converters were deemed usable?

For full frame 85-264VAC-input SM power supplies with agency and regulatory markings, valid MTBF, deep protection, etc.., there are very few choices beyond the Far East. We've made a couple on our own, but offshore vendors like Meanwell have dozens of top power engineers on staff and do it better than we can.

For spot regulation, SM and LDO, we've used countless types from just about every vendor: AD/LT, TI, ST, etc.. Again, there can be sonic differences traced back to single reg choices (analog paths). We tend to vet a reg choice early in the design process (in situ) so there are no surprises later on. And you can't automatically say one part is always ideal. In one circuit app it could shine -- in another not so much. That's part of the art of analog design.

 

[trl]

RFI exists in SMPS too and may cause nearby opamps to oscillate.

 

[solderdude]

In that case that opamp circuit would already be in a bordering to being unstable operating mode I would guess.

 

[trl]

Of course, especially with a poor PCB layout and poor ground plane. Also, big solid-states opamps might be prone to RF influences, due to their much bigger size.

 

[Vantavimeow]

I bet the higher end topping products have cut down P50 circuitry in them and the idea of the P50 is stop people buying the same components over and over again when it's only the dac chip they intended to replace conceptually.

I would like to see the D50S benched on a Topping P50 psu for presumable SINAD gain personally, especially in the bass frequencies.

 

[solderdude]

Why would bass frequencies ever sound any different from another power supply ?
Why would the distortion and noise differ when another power supply is connected unless the performance of it is really crappy ?

 

[Vantavimeow]

-120 db is the threshold for unwanted noise which the stock PSU exceeds greatly compared to both linear PSUs. If you fix the minor niggles with either PSU they would be vastly better, I think the P50 should be given the benefit of the doubt until proven otherwise.

From my own experience, each time I upgrade my DAC closer to the coveted 120db SINAD, bass is always the main upgrade I notice immediately, followed by dynamics.

For £50 per powered device in the stack and a billet aluminium shell I don't see any reason to dismiss it. Even if it simply solves niche RFI/EMI issues for specific setups. I think it's mostly applicable to desktop users in that regard.

Most people are noting a slight/mild audible improvement, which is supported by the amazon reviews vs it's peers.

 

[solderdude]

You realize that the shown plot is the DC voltage coming out of 3 different power supplies.
This is NOT the same as what comes out of a DAC as DACs usually have internal filtering and the components inside also have a PSSR/PSRR.

Also see post#10 in this thread.

I agree that having a good power supply is a good thing especially if it has low leakage from mains.

Most people are noting a slight/mild audible improvement, which is supported by the amazon reviews vs it's peers.

Blind listening tests or people knowing they have connected a linear power supply ?

 

[Vantavimeow]

I don't believe anyone has blind tested yet, though I'm not defaulting to assuming sheer cognitive bias when topping themselves seem to suggest it as well. I have ordered a P50 and aluminium upgraded remote, not sure exactly when it will arrive yet.

 

[solderdude]

Topping supplies it because there is a market for it.
Amir did not detect/measure any differences.

 

[Vantavimeow]

Topping supplies it because there is a market for it.
Amir did not detect/measure any differences.

He didn't test the P50 however, filtering is a band aid solution that can be circumvented with better engineering. It's not essential but it's definitely the best option available for relatively little expense.

I would like to see a zoomed in graph in the audible frequency range.

 

[solderdude]

What would make the P50 a different linear power supply so that it has magic powers over other linear power supplies ?
What would a zoomed in graph show and what measurement(s) should be zoomed in ?

Why would 'filtering' be a band aid and what filtering are you talking about ?
Filtering is often an integral part of the engineering.

 

[BDWoody]

Most people are noting a slight/mild audible improvement, which is supported by the amazon reviews vs it's peers.

Wife in the kitchen probably heard it too...
When 'most people' start doing 'sonic' reviews based on double blind testing, they will start to mean something. They will also go away, or turn to meaningful elements like connectivity, etc, because only then would 'most people' realize they really aren't hearing anything...in terms of audible differences...even 'slight/mild' ones.

 

[Vantavimeow]

So that the audible spectrum portion is decently legible? He conceded it was slightly cleaner and at this level a few db here and there is a reasonable gain. The P50 doesn't have a gimmicky additions that /may/ have introduced issues amir complained about. I trust the data, but at the end of the day it's still subject to interpretation rather than REEEE clean power placebo/bad vs the most noisy source of power available. Half the battle in the past few years was isolating the DAC from the massive EMI etc that computers produce, something schiit seemed to have great trouble with if I recall correctly.

No one needs to justify thinking USB power is jank, it's common knowledge in this circle.

 

[Vantavimeow]

The sheer arrogance surrounding this hobby is so off putting, it's a wonder audiophile isn't an anagram for see you next tuesday. Though outsiders wouldn't have to think twice before coming to that rightful conclusion.

 

[solderdude]

So that the audible spectrum portion is decently legible? He conceded it was slightly cleaner and at this level a few db here and there is a reasonable gain.

It is true that -130dB and -124 is a massive improvement (6dB is a doubling of signal) yet both are inaudible even when your speakers would be reaching 130dB SPL.
And if you mean zooming in on the 0dB part then you can expect a 0.001dB increase or so. Why would that be audible

Half the battle in the past few years was isolating the DAC from the massive EMI etc that computers produce, something schiit seemed to have great trouble with if I recall correctly.

It still is, mostly common mode noise related, not so much the noise on the +5V line except perhaps for some really poor design dongles perhaps.

Schiit simply did not know how to design a decent USB receiver. They learned over time.

There is a huge difference between common mode noise and differential mode noise. What Amir tested is differential mode noise and noise on the output of 1 or 2 DACs. He did not measure common mode noises nor what influence that might have on some DACs in some specific situations.

That does not 'improve' sound quality, tighten bass or smooth highs but in such case the improvement consists of the absence of weird sounds, ticks, dropouts and or sounds that should not be there.

 

[solderdude]

The sheer arrogance surrounding this hobby is so off putting, it's a wonder audiophile isn't an anagram for see you next tuesday. Though outsiders wouldn't have to think twice before coming to that rightful conclusion.

Agreed there is a lot of sheer arrogance and ignorance among 'audiophiles' in this hobby. Most of them making unsubstantiated claims based on what they heard without proper investigation on their part (science based).

Outsiders (those that aren't into this hobby) often have no idea about it except that some stuff is screamingly expensive and they swallow whatever the salesman (very little woman) and magazines/internet tells them because they 'know' this stuff.

 

[BDWoody]

The sheer arrogance surrounding this hobby

Agreed. With a virtually endless stream of nonsense from the audio press constantly bombarding people with what amounts to dis-information, there is no wonder why this is so.

When people hear more punch or slam or whatever word of the day from one power supply, USB cable, speaker cable lifters, Ethernet switch, fuse, audio rack, and on and on, vs another, it is not based on the physics of the real world as we know it. That seems pretty arrogant to me...

For those that claim it is audible but not measurable, then simply demonstrate it under controlled circumstances, and let's send the science to understand it. Without that needed demonstration, with controls, it's just back to 'cuz I say so.'

 

[Veri]

The sheer arrogance surrounding this hobby is so off putting

That's very true!