isolamento galvanico tra USB e DAC. Non basta?? qualche cosa... mi sfugge!!!

[lucadita]

Ciao a tutti,
sto leggendo da spettatore molto curioso il lavoro incredibile che Danielone, Eurodriver e tanti altri stanno compiendo sull'ottimizzazione del NAA sia dal lato software che da quello hardware... persino dall'uso della fibra... (sono spettatore perche' io sono un adepto del lato oscuro e uso squeezelite/LMS)

Io sono un po' duro di comprendonio e mi pongo sempre mille domande.
Una di queste proprio non mi fa piu' dormire di notte.


Premetto che non contesto assolutamente e neanche metto in dubbio che ci possano essere delle differenze tra un NAA alimentato in lineare e collegato in fibra, io vorrei solo discutere con voi QUALI POSSONO ESSERE LE CAUSE DI UN CAMBIO DELLA RESA SONORA se comunque il DAC e' isolato galvanicamente dall'interfaccia USB.
Lo stesso discorso vale per un cubox alimentato in lineare o con il suo alimentatorino switching...
Quindi prego chiunque dall'astenersi dall'iniziare una polemica se ci sono o non ci sono differenze!!
Fidandomi di chi le sente, vorrei solo cercare di capire e discutere di cosa puo' generare queste differenze e perche'
Spero di essermi spiegato.

[audiodan]

Eeeeeh, bella domanda!
Mi viene da citare nuovamente il Bardo "Ci son più cose in cielo e in terra Orazio di quante ne sogni la tua filosofia" perchè io una spiegazione razionale o anche solo tecnica semplicemente non la ho.
Quello che ho è un vecchio paio di orecchie, una vasta discoteca e una grande passione per la musica e per il suo ascolto critico. Accanto a tutto questo ho però maturato una esperienza clinica di ormai....36 anni (ahia, che dolor!) in un campo difficile e assai complessocome l'oncoematologia, lavoro che mi ha permesso di sviluppare una visione della realtà un po' Feuerbachiana, ovvero che la realtà è ciò che ci appare (siamo quello che mangiamo non è distante) e che dall'osservazione della fenomenologia bisogna partire per sviluppare poi la teoria. Il contrario induce all'errore, nel mio lavoro a volte mortale. Per tornare al nostro altrettanto complesso ma per fortuna ben più ameno campo di intervento diciamo che solo la mia insaziabile curiosità, quella della fatidica domanda che ogni giorno o quasi si affaccia alla mia mente ".....ma se faccio questo cosa accade?" mi ha portato a fare cose del tutto prive di basi teoriche ( e che forse chissà se mai le avranno) come la incredibile sequela di operazioni eseguite nei cosiddetti Hypertwikings di cMP2 ( chi ha la curiosità se lo vada a vedere alla fine del manuale di cMP2) ma che avevano riscontri sonici confermati da più parti. Ora, nello specifico, l'effetto di determinate soluzioni è prevedibile ma spesso esistono effetti del tutto inaspettati o addirittura "impossibili", come nel caso di ottimizzazioni HW o SW a monte di un isolamento galvanico efficace. Da quella che diventa un'esperienza diciamo discretamente vasta mi sembra di poter affermare che, al pari che nel campo elettronico, la riproduzione ( e sottolineo riproduzione) di un segnale musicale ha criticità di basso livello che non sono spiegate ne' dalle leggi dell'elettronica ne' da quelle dell'informatica.
Insomma la ricostruzione di un segnale musicale digitalizzato non è solo una questione di 0 e 1 e neanche solo di piani di massa. Ma io non sono alla ricerca di basi teoriche, non faccio il ricercatore, vado " a sentimento" o col mio "naso clinico" provando soluzioni che mi sembrano valide e tralasciando quelle che non mi sconfifferano, ovviamente riferendo dei risultati ottenuti senza alcuna pretesa di enunciare alcuna verità. Ma se trovo una soluzione che abbia un minimo di senso ve la riferisco perchè la valutiate e ne facciate uso se vi va. Insomma l'approccio è quanto di più pragmatico/illogico ci sia (anche se poi una logica di fondo c'è, in realtà). Ma come non si diventa dei grandi chef comprando un libro di cucina, altrettanto non si riesce ad avere una reinterpretazione della musica senza sensibilità e senza l'assunzione del rischio. Anche per curare i pazienti non bastano i libri, serve il senso clinico e l'intuito, sempre moderato da un po' di conoscenza e di razionalità.
Insomma, francamente non ho idea di cosa accada dentro queste macchine del cavolo, ma se gli interventi effettuati, logici o meno, mi fanno ascoltare meglio la musica in fondo chi se ne frega?

[EuroDriver]

My 2 cents is that even well implemented galvanic isolation is unable to reduce the transmission of RF to levels low enough not to affect the analog stages of the DAC. Bruno Putzey said in an interview that the US FCC regulations for RF emission limits were 30 db too high for the hifi audio equipment not to be affected

Here is a quote from John Westlake who used to design DAC's for Audiolab

The RF spectrum from SMPS is pretty horrific, the typically loop bandwidth of an SMPS is typically less then 1KHz (due to the phase shift introduced by the output LC filtering).

The typical SMPS LC filtering will be ineffective at RF due to inter-winding capacitance on the LC filters (and the internal parasitics of the decoupling capacitors etc) - it takes a few moments with a simulator to see the effects of only a few pF on RF coupling & conversely RF isolation.

While RF CAN be attenuated, its not in your typical Consumer grade PSU which have little in the way of screening / filter beyond what is required to pass mandatory CE requirements which is more concerned about inducing external interference then absolute levels of external interference rejection.

SMPS cause a HUGE level of RF interference - as a low level signal designer, I'm left to deal with the effects of these on a daily basis...

As with anything, an SMPS can be well filtered / screened, but its very costly by today's "Consumer" standards, and requires skill of the designer which is not understood by your typical (any) China SMPS vendor.

If you where in my shoes, then you would know how many repetitive times I've had to resolve SMPS RF noise interference for 3rd party's and its only getting worst. Sometimes I feel I'm on repeat - some old problems to the "surprised" client.


Quote:
Originally Posted by Julf View Post
But that kind of RF enters the system directly, not via the power cord.
Unquote

RF is injected via any connector port - the mains AC supply is typically one of the bigger problem areas - the proficiency of crude SMPS (and even cruder LED lighting supply's) and Home network over Mains just increases the number of encountered problems in the field.

In fact I have a Patent on a Modulator technique to reduce RF interference - I spent many years researching and developing high performance SMPS / Class D designs - I have a in-depth working experience of there RF performance.

The thread on Pinkfish Audio is here RFI : how does it affect mains circuitry of hifi equipment? - pink fish media

What I have a big problem to explain is why Primo Ram disk makes a big difference to improve sound quality on HQ Player Desktop on Win 8.1 playing to NAA via optical LAN. The improvement is not due to reduction of RF !

Ramdisk will reduce latency of the PC, and may be, some how, that makes the asynchronous data transfer go more smoothly ? We tend to assume that Asynch transfer with buffers as being perfect. Absolutely "bit perfect" it is, but the timing of the isosynchronus blocks of data arriving at the USB receiver could be causing the USB receiver to run less less smoothly which in turn might cause jitter. I have to say the SQ improvement that Primo Ramdisk makes seems to be different from typical SQ effect of reducing the noise floor. The SQ effect is like getting the sound picture into focus.

I have read this on Apogee's web site

So, what does digital audio affected by jitter sound like? To answer this question, imagine a recording of an orchestra made with a perfectly placed stereo pair of mics. If you listen to the analog output of the mic pres, ideally you’ll have a precise and wide image of the orchestra. If you closed your eyes, you could point to the triangle player, there, behind the second violins.

Smaller amounts of jitter start to cloud the precision of the stereo image – you can’t pinpoint sources as clearly. As greater amounts of jitter degrade the system, the stereo image starts to shrink – your orchestra isn’t so wide. With increasing jitter, changes in timbre occur that accentuate ugly harmonics.

What they describe seems to have similarities to what I have heard with the improvement that Primo Ramdisk made to HQ Player decktop. So maybe it's jitter reduction
WS2012 and AO pay a lot of attention to reducing latency

Apolgies for this thinking whilst writing post

[bibo01]

...

What I have a big problem to explain is why Primo Ram disk makes a big difference to improve sound quality on HQ Player Desktop on Win 8.1 playing to NAA via optical LAN. The improvement is not due to reduction of RF !

Ramdisk will reduce latency of the PC, and may be, some how, that makes the asynchronous data transfer go more smoothly ? We tend to assume that Asynch transfer with buffers as being perfect. Absolutely "bit perfect" it is, but the timing of the isosynchronus blocks of data arriving at the USB receiver could be causing the USB receiver to run less less smoothly which in turn might cause jitter. I have to say the SQ improvement that Primo Ramdisk makes seems to be different from typical SQ effect of reducing the noise floor. The SQ effect is like getting the sound picture into focus.

I have read this on Apogee's web site



What they describe seems to have similarities to what I have heard with the improvement that Primo Ramdisk made to HQ Player decktop. So maybe it's jitter reduction
WS2012 and AO pay a lot of attention to reducing latency

Apolgies for this thinking whilst writing post

From a latency point of view, if that is the cause for the difference in sound quality, then a RealTime OS could have a positive impact.
As a matter of fact, in the coming weeks I will be testing one. Coincidence?! Maybe not.

[bigtube]

QUALI POSSONO ESSERE LE CAUSE DI UN CAMBIO DELLA RESA SONORA se comunque il DAC e' isolato galvanicamente dall'interfaccia USB.
.

ti rispondo in breve: è irrilevante per chiunque abbia la mente fresca ( cioè privata dai condizionamenti psicologici e culturali a cui siamo soggetti)
Bisogna far fare le prove agli abitanti delle foreste del Borneo....chissà quanto si divertirebbero.
Noi invece siamo prigionieri dei nostri maledetti pregiudizi ( di cui non abbiamo colpa ...il cervello funziona cosi in genere ....cerca sempre una soluzione a disposizione che non costi troppa fatica ).
L'unica ribellione è essere scettici e dare spazio alla immaginazione, quella risorsa che da sempre ci avvicina sempre di piu' ad appropiarci della verita'.

secondo me nella percezione del fluire musicale l'unico vero elemento cui rivolgere l'attenzione è il tempo , il timing. I disturbi e le distorsioni non illudono
la ricerca del cervello nel riconoscere il fluido muoversi del flusso musicale come evento che avviene davvero in quel momento. Ogni minimissima discrasia temporale il cervello la decodifica immediatamente e tutto cio' che da allora in poi si dipana ci appare irrevocabilmente discrasica e quindi priva di verità o
veridicità.....subito ci annoia perchè il cervello ha colto il salto temporale. questa è la mia opinione infatti non mi fanno paura le distorsioni i disturbi udibili
e non udibili ( perchè al cervello non importa quell'aspetto .....di disturbi ce n'è dovunque e sempre anche quando si ascolta la musica dal vivo!!!)

[DacPassion]

...una risposta "mossiana" Giovanni :-)

[EuroDriver]

From a latency point of view, if that is the cause for the difference in sound quality, then a RealTime OS could have a positive impact.
As a matter of fact, in the coming weeks I will be testing one. Coincidence?! Maybe not.

Real time OS, hmmmmm

What defines a real time OS ?

How is it clocked ?

[bibo01]

Real time OS, hmmmmm

What defines a real time OS ?

How is it clocked ?

"A real-time operating system (RTOS) is an operating system (OS) intended to serve real-time application process data as it comes in, typically without buffering delays. Processing time requirements (including any OS delay) are measured in tenths of seconds or shorter.

A key characteristic of an RTOS is the level of its consistency concerning the amount of time it takes to accept and complete an application's task; the variability is jitter.[1] A hard real-time operating system has less jitter than a soft real-time operating system. The chief design goal is not high throughput, but rather a guarantee of a soft or hard performance category. An RTOS that can usually or generally meet a deadline is a soft real-time OS, but if it can meet a deadline deterministically it is a hard real-time OS.[2]

An RTOS has an advanced algorithm for scheduling. Scheduler flexibility enables a wider, computer-system orchestration of process priorities, but a real-time OS is more frequently dedicated to a narrow set of applications. Key factors in a real-time OS are minimal interrupt latency and minimal thread switching latency; a real-time OS is valued more for how quickly or how predictably it can respond than for the amount of work it can perform in a given period of time." (from wikipedia)

There is an interesting explanation paper at National Instruments on RTOS:
What is a Real-Time Operating System (RTOS)?

In a RTOS it's possible to assign FIFO priority to IRQs that relate to audio and system clock, and to applications. For example, rtc0 (real time clock) could have 90, USB for DAC 85, LAN 80 and App 75.

[EuroDriver]

"A real-time operating system (RTOS) is an operating system (OS) intended to serve real-time application process data as it comes in, typically without buffering delays. Processing time requirements (including any OS delay) are measured in tenths of seconds or shorter.

A key characteristic of an RTOS is the level of its consistency concerning the amount of time it takes to accept and complete an application's task; the variability is jitter.[1] A hard real-time operating system has less jitter than a soft real-time operating system. The chief design goal is not high throughput, but rather a guarantee of a soft or hard performance category. An RTOS that can usually or generally meet a deadline is a soft real-time OS, but if it can meet a deadline deterministically it is a hard real-time OS.[2]

An RTOS has an advanced algorithm for scheduling. Scheduler flexibility enables a wider, computer-system orchestration of process priorities, but a real-time OS is more frequently dedicated to a narrow set of applications. Key factors in a real-time OS are minimal interrupt latency and minimal thread switching latency; a real-time OS is valued more for how quickly or how predictably it can respond than for the amount of work it can perform in a given period of time." (from wikipedia)

There is an interesting explanation paper at National Instruments on RTOS:
What is a Real-Time Operating System (RTOS)?

In a RTOS it's possible to assign FIFO priority to IRQs that relate to audio and system clock, and to applications. For example, rtc0 (real time clock) could have 90, USB for DAC 85, LAN 80 and App 75.

Many thanks for this info !

[lucadita]

My 2 cents is that even well implemented galvanic isolation is unable to reduce the transmission of RF to levels low enough not to affect the analog stages of the DAC. Bruno Putzey said in an interview that the US FCC regulations for RF emission limits were 30 db too high for the hifi audio equipment not to be affected

thank you Eurodriver. I totally agree with the the quoted text from John Westlake about switching power supplying without a system to attenuate RF interferences! ...but here we "suppose to have" a system to decouple signals coming from the pc from signals going in to the DAC.

what I'm trying to understand is the effectiveness of galvanic isolators in making the system more or less sensitive to EMI and his impact on Sound Quality.
why a linear power supply can do the difference in a NAA when the I2SoverUSB board has galvanic isolation?
why a cable like Corning optical USB 3.0 cable can make the difference?
Is the galvanic isolation implemented in board like JLSounds I2SOverUSB a good an sufficient "wall" to shield from EMI?
Boxing the I2SoverUSB in a EMI proof chassis can increase the Sound Quality?

What I have a big problem to explain is why Primo Ram disk makes a big difference to improve sound quality on HQ Player Desktop on Win 8.1 playing to NAA via optical LAN. The improvement is not due to reduction of RF !

this is a mystery for me, and I'd like, for now , to focus on hardware. :-) The influence of software is another step in my understanding in what can change the Sound Quality of a system :-)