R-Core Vs Toroidal Vs Wall Wart Transformers for DAC's?

[Infinity!]

Hey guys, I often read that certain power supplies in a DAC can either make it great or make it crap?! Is this true? I have NO idea which transformers are the best, but I do know that people look down on wall warts and often praise toroidal transformers (Example: Parasound solely using toroidal transformers in their DAC's). Why is this? How do Toroidal transformers compare to R-Cores? Any and all help is extremely appreciated. Thanks!

 

[BeerCan]

As I understand it toroidals have lower magnetic leakage but higher inter winding capacitance. That makes them prone to slightly higher "noise" in the 60hz range. They can also be lighter for a given power compared to split bobbin transformers.

OTOH split bobbin transformers have higher magnetic leakage but lower inter winding capacitance. So they have less noise but are bigger and require better sheilding to work well.

Someone more knowledgeable may correct me but I think that is the jist of it.

 

[Infinity!]

As I understand it toroidals have lower magnetic leakage but higher inter winding capacitance. That makes them prone to slightly higher "noise" in the 60hz range. They can also be lighter for a given power compared to split bobbin transformers.

OTOH split bobbin transformers have higher magnetic leakage but lower inter winding capacitance. So they have less noise but are bigger and require better sheilding to work well.

Someone more knowledgeable may correct me but I think that is the jist of it.

So are split bobbin transformers the best for dacs then? I just dont know this subject very much at all.

 

[whoaru99]

It's seldom the exact nature of the transformer, or any other individual component, than about the implementation within the total design. Don't be distracted by shiny objects.

 

[robgmn]

It's seldom the exact nature of the transformer, or any other individual component, than about the implementation within the total design. Don't be distracted by shiny objects.

And avoid shoddy objects.

I've used some wall warts that were electrically noisy and even physically noisy (components buzzing), and others that were dead silent on both accounts.

Cost can often be an indicator of what to expect.

 

[BeerCan]

It's seldom the exact nature of the transformer, or any other individual component, than about the implementation within the total design. Don't be distracted by shiny objects.

Agreed. A competent designer will make either work well.

 

[whoaru99]

Maybe my comment about being distracted by shiny objects could be misunderstood. Just so it might be no misunderstanding, it has nothing to do with price/cost, but everything to do with marketing hype and urban legend, so to speak.

Toroidal transformers are a good example. There was a time it was implied, and to some degree still is as evidenced by this thread, that unless the gear has a toroidal transformer it can't possibly be top tier or perform well. That's simply not true. Hence, do not be distracted by shiny objects.

 

[Infinity!]

Well, the three DAC's I'm looking at are Parasound Zdac, Grant Fidelity Tubedac-11, and Schiit Bifrost with Uber Analog and USB. I have really been leaning towards the Grand Fidelity because i would love having the option for a tube pre out as well as a regular solid state out. I've always wanted to hear a tube dac. The Parasound and G.F. have toroidals while the Schiit Bifrost has a split bobbin if I'm correct? Just not sure!

 

[pete_mac]

For dacs and source gear in particular, I'd rate them as follows in terms of suitability (from best to least):

R-core
EI core
Toroidal

R-core have characteristics that are best suited for low level source applications.

 

[BeerCan]

you have way more things to worry about than the transformer IMO
Just some examples
If you plan on using USB the Tubedac uses the TE7022 chip, so for me that would be a no go. There are much better usb implementations these days.

Same with the Parasound it uses the TAS1020B , its better but meh. It also upsamples everything if that matters to you.

Birfrost uses the CM6631A usb, the best of the bunch if you ask me but still not xmos I also like the AKM chip better but that is a matter of taste.

If it were my money (its not) it would be no contest and I would buy Schitt

 

[gadget73]

If the power supply is properly designed, frankly it just doesn't matter that much. You're looking for clean DC voltage at the end of the supply chain. You can get there with most any sort of transformer. The filter network will play more into it than what shape the transformer core is.

 

[davidflas]

For dacs and source gear in particular, I'd rate them as follows in terms of suitability (from best to least):

R-core
EI core
Toroidal

R-core have characteristics that are best suited for low level source applications.

I know practically nothing about the components inside my gear, but for strictly educational purposes, which category do the transformers inside my Gungnir fall into?

 

[BeerCan]

EI Core?

 

[restorer-john]

Nelson Pass used a laptop power supply in his amp camp designs because: "We assume that most of the participants in this project do not have the skills to safely connect the components which make up an amplifier power supply to the AC power line."

And it was cheap, very cheap.

Apparently, its contribution was 100uV to the noise floor of his amp. Wow, that's hifi- not.

No DAC manufacturer aspiring to anything greater than telephone quality audio would remotely consider a low/mid frequency SMPS like a 15-30kHz SMPS laptop supply.

I personally hate SMPS supplies, due to their frequent spectacular failures, poor regulation, noise (did I say noise) and the damage the myriad of non-PFC SMPSs are doing to the mains waveform integrity.

A decent DAC needs a decently designed PSU. EI, toroid, R or C core are all fine if designed properly and screened effectively. SMPSs are cheap junk and have no place in high fidelity gear designed to last longer than its warranty period. The most common repairs in the last decade have revolved around SMPS failures in consumer electronics for me. Everything from washing machines to monitors, DVD players, class D amp supplies, computers PSUs, you name it- the SMPS and the over-stressed electrolytic capacitors all cooking themselves from the inside out, poorly designed with little or no ventilation.

 

[bacobits]

Amazing how you can buy a $1795 Chord DAC QuteHD at the time (I didn't pay that) and have these SMPS. AND they are rated for constant duty?
I got rid of all of them including the one on my Regen and use all LPS's.

 

[whoaru99]

Nelson Pass used a laptop power supply in his amp camp designs because: "We assume that most of the participants in this project do not have the skills to safely connect the components which make up an amplifier power supply to the AC power line."

And it was cheap, very cheap.

Apparently, its contribution was 100uV to the noise floor of his amp. Wow, that's hifi- not.

No DAC manufacturer aspiring to anything greater than telephone quality audio would remotely consider a low/mid frequency SMPS like a 15-30kHz SMPS laptop supply.

I personally hate SMPS supplies, due to their frequent spectacular failures, poor regulation, noise (did I say noise) and the damage the myriad of non-PFC SMPSs are doing to the mains waveform integrity.

A decent DAC needs a decently designed PSU. EI, toroid, R or C core are all fine if designed properly and screened effectively. SMPSs are cheap junk and have no place in high fidelity gear designed to last longer than its warranty period. The most common repairs in the last decade have revolved around SMPS failures in consumer electronics for me. Everything from washing machines to monitors, DVD players, class D amp supplies, computers PSUs, you name it- the SMPS and the over-stressed electrolytic capacitors all cooking themselves from the inside out, poorly designed with little or no ventilation.

Poorly designed and cheaply built is one thing, but I am inclined to disagree about SMPS in general because I have many examples of more than a decade of trouble free operation of SMPS supplies.

Granted, they're mostly in amps costing >$1000 each, not a $5 power brick...but, that's part of my point.

 

[Infinity!]

Well, I know this discussion is old, but I have finally had the chance to listen to both the Schiit Bifrost Multibit and the Grant Fidelity Tubedac-11, and the G.F. wins hands down!

I know the specs and components of the Schiit are better, but the sound of the G.F. kicks it's butt. The soundstage/imagining is better along with being warmer (but not TOO warm). Even the regular "dac out" that completely bypasses the tube, sounds better than the Schiit Bifrost IMO. Of course, this is in my system (Infinity RS IIb's/Kappa 9's) that uses electromagnetic tweeters and mid-ranges, so the warmer sound does compliment them nicely.

 

[SoNic67]

E-Core can have better electrical insulation between high voltage and low voltage side, if they are provided with a copper shield between layers. Almost nobody makes E-I cores with coils on separate legs, because they will leak more magnetic field, so that is not an advantage.
However, they will "spill" more magnetic field around them and in between them. Sometimes the separated digital and analog transformers are angled at 90 degrees to compensate for that and sometimes are shielded with a magnetic steel enclosure.
Those shielded electrically (copper foil between layers) and magnetically (steel enclosure) are the best in my view. Sometimes they are triple shielded electrically - copper foils between core-primary, primary-secondary, secondary-outer core. Sometimes the primary is wounded last to minimize the capacitive coupling.

Toroidal transformers don't "leak" magnetic field so they don't need a magnetic shield. However they seldom have electrical shield between coils, so some degree of electric "leakage" will happen. Proper design and grounding the equipment (three prong cable and outlet) can take care of that.

Wall-Warts are seldom good - only if they purely transformer type and if they are (over-)sized properly. But they seldom are... because of costs and third party suppliers. They seldom have the copper shield between coils, so there will be some electric coupling. That's why the bad reputation. Why they are used? Well the idea is to take the magnetic influence out of DAC enclosure, instead of shielding it inside. Hmm... but then you have just one transformer for both digital and analog stages. I don't care what the design is, that's never providing a "clean" enough power for the analog section.
Since almost never they have the ground pin, they are not grounded and they will "push" any electrical stray currents towards the DAC and from there to any other equipment that is really grounded (amplifier), via analog cable shields. Stray currents will try to reach the ground no matter what and going there trough the analog ground shields... that's bad.

Note that lots of new Wall-Warts are just Switched-Mode Power Supplies. Lots of electrical noise, some magnetic noise too, and what's worse it is in a high-frequency range. That propagates easier, it can travel on the surface of shielded cables and is hard to ground properly - every wire has an inductance that will make a higher impedance (resistance equivalent in AC) for the high frequency signals. If for 50-60 Hz the ground wire presents a 1 ohm impedance, for a 30kHz parasite signal that impedance can be 300-600 ohms or even more. Now that "ground" is not really a ground anymore, so the parasitic HF signal cannot be bleed out that easy.

Example of the "good stuff":

And a pic for the shielding stuff that I have talked about:

 

[Infinity!]

Well, I know this discussion is old, but I have finally had the chance to listen to both the Schiit Bifrost Multibit and the Grant Fidelity Tubedac-11, and the G.F. wins hands down!

I know the specs and components of the Schiit are better, but the sound of the G.F. kicks it's butt. The soundstage/imagining is better along with being warmer (but not TOO warm). Even the regular "dac out" that completely bypasses the tube, sounds better than the Schiit Bifrost IMO. Of course, this is in my system (Infinity RS IIb's/Kappa 9's) that uses electromagnetic tweeters and mid-ranges, so the warmer sound does compliment them nicely.

I COMPLETELY retract this comment... the Schiit is infinitely better!! After hearing the Schiit Bifrost Multibit again (having a few months of burn in time), it has far surpassed the Grant Fidelity Tube Dac 11! The g.f. dac is so sterile now in comparison to the Schiit. Bass isnt as good, highs are harsher, and overall just not that great. Don't know what happened in those few months, but its night and day after re-listening to them. Just wanted to clear this up!

 

[SoNic67]

IMO your brain adjusted to normal, good sound
I am not kidding, brain will compensate for what he hears to a certain degree, to match a more "natural" sound.

PS: It's harder on rock music, because most of the time only piano (and even that is a maybe) and drums are "real", concerts are over-amplified. Other genre are easier to adapt, like symphonic music, especially if the brain gets exposed to life concerts a lot (to have a baseline).