Wall wart (12v power supply) DAC noise?

[Lee Guest]

Apparently, you do think that engineers are fools.

Enjoy your Modi.

lol. I think you misunderstand. I work in engineering. Engineers perfectly understand what they are doing, at least let's hope so. It's the non-engineering people that fail to understand quite a lot of things, and fill the blanks in with opinions rather than facts.

You don't like Modi3s?

 

[E-Stat]

lol. I think you misunderstand. I work in engineering. Engineers perfectly understand what they are doing, at least let's hope so.

Then why do you assume that the engineers of the finest components don't use sound principles for design selection? They would shake their heads at your snide and dismissive tone.

lYou don't like Modi3s?

I'm told its a great bargain for $99.

 

[Lee Guest]

Then why do you assume that the engineers of the finest components don't use sound principles for design selection? They would shake their heads at your snide and dismissive tone.


I'm told its a great bargain for $99.

They work fine. They aren't multi-bit, but they sound fine to me.

 

[Lee Guest]

I suspect the engineers at Ayre, Audio Research, dCS, EMM Labs, Meridian, MSB, Music Fidelity, Simaudio, etc. would just smile at such a comment.

Are aware that the Schiit Yggdrasil uses a linear power supply?

No, I didn't really check on that.
I know that the Modi3 uses a switcher, according to somebody I was talking to at Schiit Audio. It sort of surprised me actually, but with proper design it should be no problem.

We have a bigger problem in Canada. We have to get our trains running. lolol. Glad I don't have to solve THAT problem. A whole lot of upset people.

 

[E-Stat]

It sort of surprised me actually, but with proper design it should be no problem.

Doesn't surprise me in the least. That's what you do when you build to a modest price point.

A challenge not faced with the $2495 Yggdrasil. Nor even with the improved version of the Modi at 2.5 times the price.

 

[Lee Guest]

True enough.

 

[W9TR]

I found it interesting that several internet denizens heard a big difference in sound using a linear supply on the Topping D50. The D50 is a desktop DAC with no internal mains supply. The first thing this linear supply sees is.....you guessed it.....three switchers for +/- 12 Vdc for the analog audio and +3.3 V for the digital circuitry. Hmmmm. I can’t hear a difference running this DAC off a battery or an Apple USB wall wart power supply.

A friend let me borrow a Mytec Brooklyn DAC+ and an external analog power supply. The internal switcher in the Mytec performed quantitatively better than the linear supply. The analog supply had harmonics of the line frequency that showed up in the audio output noise floor and also as sidebands around a decoded fundamental. I couldn’t hear any difference between the two.

Tom

 

[Lee Guest]

Sounds about right. There are times when people hear what they want to, and this includes me. That's why I hate relying on only hearing things, but not measuring them. People say, "Trust your ears". And I do. I trust them to tell me to take some measurements, so that I know what I am hearing is real.

It sort of reminds me of VSauce where Michael might say, "Switching power supplies are always noisier than linear power supplies...........................................................or ARE they ?????????????" lolol

It's one of those "definitely, maybe" things. Linear supplies can be good, and so can switching supplies, and the opposite is also true for both.
I mean, a class D amp is really operating on the same principle as a switching supply, and some people love those things. You never hear those people insisting that we go back to class A or A/B to get rid of the switching noise. lol
(I will admit they are getting a lot better)
It's all in the design.

 

[xdd]

I found it interesting that several internet denizens heard a big difference in sound using a linear supply on the Topping D50. The D50 is a desktop DAC with no internal mains supply. The first thing this linear supply sees is.....you guessed it.....three switchers for +/- 12 Vdc for the analog audio and +3.3 V for the digital circuitry. Hmmmm. I can’t hear a difference running this DAC off a battery or an Apple USB wall wart power supply.

Switchers running from DC (often called DC/DC converters) don't give rise to the CM noise problem that mains-fed switchers do. They generate normal mode noise like any other SMPSU but no CM noise to speak of gets out as they're not in general isolated from the DC that powers them..

 

[W9TR]

Since most (all?) mains switchers start with a full wave bridge and capacitor filter I don’t understand the distinction you are making. I guess if the DC-DC converter is not isolated then the noise will be reflected to the supply lead and not the ground, although my experience is that the grounds around any smps need to be very low impedance or you get a lot of switcher noise on the ground, which can then get into your unbalanced interconnects and other equipment.

https://www.analog.com/media/en/technical-documentation/data-sheets/3798fa.pdf

Either way some type of CM and DM filtering is needed before the switcher or you’ll get interference on the mains.

Are linear power suppliers switchers?
No, but they can act like it sometimes.
If a linear supply is a standard transformer followed by full wave bridge rectifier and capacitor design (no choke) then very high currents will flow through the rectifier and transformer for a short time when the capacitors are charging every half cycle. This noise is really hard to filter, which is why I see it on the outputs of most supposedly quiet linear power supplies I’ve used. There are a bunch of methods to get rid of this noise but they add cost and complexity.

 

[E-Stat]

You never hear those people insisting that we go back to class A or A/B to get rid of the switching noise.

Go back? I've never left class AB for the best overall performance.

As one of "those people" I can tell you that switchers fall short at the top end in terms of natural extension and transparency to these ears. They can, however, have a very nice midrange. I chose a Hypex Ncore amp in the garage system for its power and overall sound, but it would never replace the VTLs in the main system.

 

[xdd]

Since most (all?) mains switchers start with a full wave bridge and capacitor filter I don’t understand the distinction you are making. I guess if the DC-DC converter is not isolated then the noise will be reflected to the supply lead and not the ground, although my experience is that the grounds around any smps need to be very low impedance or you get a lot of switcher noise on the ground, which can then get into your unbalanced interconnects and other equipment.

Yes both examples work as DC/DC converters in effect (maybe PFC input converters are an exception), its the isolation in the mains-fed case which means there's no path for the currents which are flowing through the parasitic elements of the transformer to return. Here's an article which dives into the nitty-gritty : https://www.eetimes.com/power-tip-4...mode-emissions-in-isolated-switchers-part-1/#

Good point about linear power supplies not being completely immune to noise. In my career I had one linear power supply fail EMC testing due to my omission of RC snubbers across the rectifier bridge diodes.

 

[W9TR]

Great article- clarified the situation for me. Thanks!

 

[Lee Guest]

Go back? I've never left class AB for the best overall performance.

As one of "those people" I can tell you that switchers fall short at the top end in terms of natural extension and transparency to these ears. They can, however, have a very nice midrange. I chose a Hypex Ncore amp in the garage system for its power and overall sound, but it would never replace the VTLs in the main system.

Yes, I pretty well stick to A/B amps too. Class D has come a long way, but (for me) it is simply easier to use A/B amps.

 

[Lee Guest]

Then why do you assume that the engineers of the finest components don't use sound principles for design selection? They would shake their heads at your snide and dismissive tone.


I'm told its a great bargain for $99.

I think you mistook my statement. What I meant was, NOBODY in engineering goes by anything except known principles and measurements, at least let's hope so. I was presenting a situation which would never happen, where engineering people go by opinion and here-say. It's just an inside joke in engineering, where things get said that are obviously the opposite to what really happens.
Engineering always involves necessary compromises. It is experience that teaches you what compromises you can make and still achieve the original design goal.

 

[E-Stat]

I think you mistook my statement.

I think I understood fully your shallow attempt at humor in post #38.

Which has absolutely nothing to do with why the companies I cited use linears. Their only *compromise* is added cost.

 

[Lee Guest]

.

 

[Lee Guest]

I think I understood fully your shallow attempt at humor in post #38.

Which has absolutely nothing to do with why the companies I cited use linears. Their only *compromise* is added cost.

I think I will find another audio board. WAAAAAYYY too tense here. This is supposed to be fun, and I am tired of walking on eggshells so that people aren't offended. You could lighten up a bit, just some advice.

 

[E-Stat]

I think I will find another audio board. WAAAAAYYY too tense here. This is supposed to be fun, and I am tired of walking on eggshells so that people aren't offended. You could lighten up a bit, just some advice.

If snide humor that has no basis in fact is your thing, I suggest Hydrogen Audio.

 

[Lee Guest]

If snide humour that has no basis in fact is your thing, I suggest Hydrogen Audio.

You are clearly not hear to learn, and sarcasm to complete strangers seems to make you feel better. No worries. I don't need this near as much as you do.
Cheers