In DAC, Anything audible but unable to be measured(so far)?

[SIY]

Some of the old Q Sound recordings could do that. Spooky.

 

[flipflop]

Meier crossfeed is available as a plug in for Foobar. Bauer stereo to binaural which uses some version of crossfeed is available in VLC. There are others. Stereo widening is usually some variation of mid-side processing. I've done that for myself, and sometimes it is better, sometimes it is not. Depends upon the recording.

So far in my experience nothing beats good speakers in a moderately large or large room.

The Bauer thing looks interesting and while I do appreciate your response, I'm still no closer to an answer.
Guess I'll have to make a thread about the topic soon.

 

[restorer-john]

DSP can do a lot of things but miracles are not one of them

I remember hearing the Yamaha DSP-1 in 1985/6 at an audio show and it was absolutely a miracle.

Sonically, the walls and roof disappeared and new acoustic venues were created. Apart from the first time I heard Compact Disc, there has been nothing in high fidelity that impressed me as much, before or since.

from a review back in the day:

 

[Rusty Shackleford]

What an odd posting on Massdrop..

First off, made by someone who you would think did his research, but makes a massive blunder about Amir's credentials when they're all bare for anyone to take a look at.

Second, the link he gives to SABF.. is possibly the worst place on the planet in terms of audioclowns. Even me with a rudimentary understanding of most of this stuff, I've managed to find these retards slipping about. They remind me of some pseudo SJW's (as opposed to actual social justice warriors that do work in fields like human/civil rights). They want to call bs on some measurements, when if that were the case, the company in question would be going after Amir instantly for slandering potentially. So just like idiots today who feel offended on behalf of Jews(when for example a vegan says what the animal agriculture is engaged in today is a literal Holocaust) because their ancestors suffered the Holocaust, these people want to suffer something for Schiit, when Schiit themselves see no error.

Also, this place is riddled with mentally deranged people who think good performance metrics can be faked. The same sort of people who think the THX 789 and Atom are pointless and sterile devices.

So just going by natural human bias (in the same way you would think of people who you might observe to eat shit daily, safe to say you wouldn't want to have dinner with them), I tend to have a bias against listening to what people like the "purr1n" guy on SABF says. These people simply failed far too many litmus tests of rationality and sanity to ever be taken seriously (even if indeed they ever made good points in the future, there comes a time when you stop giving people chances especially if they don't rescind every single thing they've said in the past).

Third, Amir, like any other person has his quirks. He sometimes does stuff that doesn't make sense to me (sometimes because I have no idea of the technical nature because it's something that requires an actual education in the electrical field for example) and sometimes he just does, or says things that don't actually make sense due to typo's or not understanding the gist of something or the contexts around it.

The point being.. who cares? He could be a secret racist for all that matters to me, it has zero impact in the educational and revealing nature of the work he engages. I could think he's dumb for thinking a certain way or whatever, but what does that have anything to do with the content he's provided here (and the reviews + measurements of gear)?

See folks, this is my main problem with audiophool nonsense. And posts like the one on Massdrop (even though the second link leads to a dead-end link).. These people will hark on a person for supposedly starting a cult. A cult of what..? Objectivists? How is this bad exactly?

I see this same idiotic non-critical thinking nonsense in religious vs atheist circles. You tell religious folks you're just not convinced God exists, and you have science the ever progressing handed-down method of attempts at better understanding our world. NOPE, they say because you don't know the Big Bang's actual nature, while they do (that being God, which doesn't actually teach us anything, it only further compounds the issue by presenting us with a bigger problem of understanding a more complex thing like God).

So here's the kicker, so lets say we say fuck-off to doing things scientifically as we cultists here prefer. What's the alternative? Belief in non-evidenced based claims? Are these people LITERALLY retarded? How many times must it be demonstrated to these people that their way of doing things doesn't actually make sense from a critical thinking perspective? Then you'll get the apologetic reasoning "oh but we say just go by your ear, we're not telling you what to believe, just listen to your emotions" and wishy washy nonsense like that.. Except if you give them an inch, they take a mile by then having this sort of mentality leak into scientific deliberations, and before you know it, they'll be making claims like that one stupid fuck (I forget his name, but said he can hear down to -300 db or some such other nonsense, Amir would know his name if he reads this).

And no matter how many times you tell them: "it's okay, you believe whatever you want, that's fine, just don't bleed your nonsense into fields you have no place in, nor do you believe in the basic concepts that science has established, and even the products you use are now based on the total opposite behavior that you now engage in". That's when they get pissed, and start these sorts of threads about the "ASR cult" or whatever such idiotic declaratory statements.

So, in conclusion, even if Amir has some problems, or inconsistencies that random people on the internet might perceive (like ever other normal person on the planet who has their own quirks), why would that even matter? It's like saying "Because Newton actually (and he did) had more writings about religion, that we should disregard all of his scientific contributions". Absolutely ridiculous...

And also, the fact that they attempt to attack the person, and not the science itself (because the half-sensible among them know they get shot down constantly when they battle in that arena), is telling of desperate grasps at straws.

The ONLY portion of the linked Massdrop thread that makes sense was about "getting a second or third opinion" on things. Nothing inherently wrong about skepticism there. The only problem being.. the irony of the second or third opinion those folks would actually need from a neurologist, psychologist, and psychiatrist in all other matters.

(Sorry about the long post folks, I really get fired up when I see links like that, and all I think about is some noob who is emotionally swayed by it, who knows no better, and is led down another direction where voodoo runs rampant. I was almost there when I got into audio, and only stumbled by here accidentally, as I assumed audio was just like computers for example, where benchmarks and validation ran supreme, and not people 'feels').

This post is basically an amalgamation of super offensive things (“retards,” etc.) spouted by someone who’s claiming to be on the intelligent, rational side of things. Do better.

 

[Veri]

The Weiliang E600 is a nice looking TPA6102 amplifier, certainly for the price.
Would get rid of the Muses.

I'd buy one and ship it to Amir, but there's the problem of 110v/220v version... I'd need the EU-compatible one .
Maybe someone else feels like having theirs tested? It does look like a cheap enough balanced amp.

 

[solderdude]

It is a TPA6102 amp so most likely in balanced mode the output R is 20 Ohm.

For the price it is hard to complain about.

 

[Veri]

It is a TPA6102 amp so most likely in balanced mode the output R is 20 Ohm.

For the price it is hard to complain about.

I intend to use it with HD650 balanced so 20 Ohm output is probably still manageable. SE 10ohm R is hopefully also still 'OK'.
They claim pretty great performance on this thing, which is why I'd like to see it verified

Considering it is SE/BAL in/out, if it's anywhere close to Atom's distortion specs that's pretty good value for a 2kg+ enclosure and balanced amp. Its biggest fault would be output impedance, just like DX7/DX7s which can be a problem.. depending on headphones used That is, if there are no big fat distortion surprises I guess I'll find out once it arrives.

 

[JohnYang1997]

I intend to use it with HD650 balanced so 20 Ohm output is probably still manageable. SE 10ohm R is hopefully also still 'OK'.
They claim pretty great performance on this thing, which is why I'd like to see it verified

Considering it is SE/BAL in/out, if it's anywhere close to Atom's distortion specs that's pretty good value for a 2kg+ enclosure and balanced amp. Its biggest fault would be output impedance, just like DX7/DX7s which can be a problem.. depending on headphones used That is, if there are no big fat distortion surprises I guess I'll find out once it arrives.

The distortion figure is directly quoted from the datasheet. I guess it would be around 0.0006% with most loads. And over 0.001% with low impedance load. Also degradation with balanced out. It maybe not bad. Just I don't see the point of it. Balanced, probably true someone what to drive 600ohm headphones. 300ohm wouldn't be as demanding.

 

[Veri]

Just I don't see the point of it.

Considering the THX AAA is very difficult to get, it seems like a decent Chinese alternative for a box offering SE and BAL inputs, outputs. Probably not that high fidelity, but still good value considering it is sold starting $94 for such a big hefty amplifier box. Idk, I'm intrigued lol. I don't really understand how they can manufacture a thing so cheap.

 

[JohnYang1997]

Considering the THX AAA is very difficult to get, it seems like a decent Chinese alternative for a box offering SE and BAL inputs, outputs. Probably not that high fidelity, but still good value considering it is sold starting $94 for such a big hefty amplifier box. Idk, I'm intrigued lol. I don't really understand how they can manufacture a thing so cheap.

One thing to tell is that tpa6120 runs at +-15V, and Chinese people like to run very high voltage. And since 789 is only +-12V you can get some more power with high impedance headphones.

Plus the balanced out should take care of the power supply noise and most common mode distortion. Very nice.

 

[solderdude]

You can expect DX7s like performance (the amplifier portion).
Nice volpot (motorized), enclosure and has output protection (don't know if it also has DC protection).

 

[JohnYang1997]

Let me share something with you guys. There is a simple mod that can be done with any tpa6120 circuit that solve output impedance issue.
Simple 2 steps
Step 1 replace the Rf to Rf in series with a 560p capacitor.
Step 2 add another resistor across the inverting input and the amplifier output (the other end of the 10 ohm resistor)
As shown in the picture.
NOTE: The total length of Rf + Cf should be as short as possible less than 2cm even 1cm as optimal. Longer total length may cause performance degradation and instability issues.
The Cf should be ceramic type.
Original/stock

New/modified

 

[solderdude]

From the TPA1206A data sheet:

CAUTION Do not place a capacitor in the feedback path. Doing so can cause oscillations.

(bottom page 10, concerns the output resistor)
Is this solution stable with all headphones and PCB layouts ?

For high frequencies Rf becomes 500 Ohm.
The spec sheet clearly shows instability at higher frequencies at 620 Ohm and recommends Rf to be between 1k and 10k.

 

[JohnYang1997]

From the TPA1206A data sheet:



(bottom page 10, concerns the output resistor)
Is this solution stable with all headphones and PCB layouts ?

For high frequencies Rf becomes 500 Ohm.
The spec sheet clearly shows instability at higher frequencies at 620 Ohm and recommends Rf to be between 1k and 10k.

The datasheet meant direct capacitance without the Rf. The circuit I proposed just works. No issue. Fully tested.
Also the direct resistance still works very fine at 220 ohm. So 500ohm is still very stable. You really need to rescale the graph in the datasheet. 2db is very safe.
If you want to play safer use 2kohm for both resistors. But it's not really something that actually matters.

 

[solderdude]

Still the total Rf will be 500 Ohm at 10MHz = not stable acc. to the datasheet.

If I were to implement this I would choose 2x 2k ?

what happens when say a 500pF capacitive load is connected ?
Even with a low impedance load the inductance of the voicecoil will provide almost no resistive load.
With various PCB layouts (The IC is quite sensitive to this and to longer feedback paths)

 

[JohnYang1997]

Still the total Rf will be 500 Ohm at 10MHz = not stable acc. to the datasheet.

If I were to implement this I would choose 2x 2k ?

what happens when say a 500pF capacitive load is connected ?
Even with a low impedance load the inductance of the voicecoil will provide almost no resistive load.

Actually you shouldn't worry about it. Because when you do the capacitive loading simulations for more pushed to the edge circuits, it's going to get much more than 2db. (not really here) 20db is needed to be true instable. More than that is needed to oscillate. Generally 6db below is safe and still has good amount of margin, 2db is very safe.
The Cf is recommended by me from 33p to 1n. 560p is a typical value that I like to use.
Frequency response with different Cf from 33p to 1n.(right to left) (no capacitive loading)

Capacitive loading from 33p to 10n. (right to left)(Cf = 560p)

 

[JohnYang1997]

Another scenario shows some peaking at high frequency. If the
peaking is 2dB, the amplifier is stable as there is still 45 degrees of phase margin. As the peaking increases, the
phase margin shrinks, causing the amplifier and the system to approach instability.

2db is considered stable in the datasheet.
Only thing that needed to do is stability analysis/simulation which is in the last post. No issue at all.

 

[solderdude]

How did you obtain a spice model for the TPA ?
Did you include PCB traces and layout in the model ?
Above 10MHz this becomes a thing so to speak.
Why not add a small inductor in the output path just to be sure ?

 

[JohnYang1997]

How did you obtain a spice model for the TPA ?
Did you include PCB traces and layout in the model ?
Above 10MHz this becomes a thing so to speak.
Why not add a small inductor in the output path just to be sure ?

Ths6012 is the same chip as tpa6120.
Pcb traces isn't a concern. You can take off 10° of phase margin for rule of thumb. I have pushed a circuit with tpa6120 to 13° phase margin and it's still stable. I did add capacitor across inveting input and ground as well as inductance everywhere before. The circuit above is just for simple demonstration.

 

[JohnYang1997]

Also i feel this is getting a little bit ridiculous to me. I just wanted to share a tested modification here. This was firstly done long time ago. If a simple modification such as this needs this kind of back and forth, I don't really want to share stuff on this forum anymore.