MDAC First Listen (Part 00101001)

[Ciunas Audio]

I suspect most PHY's recover the data using a "Clockless" oversampling method like the SPDIF receiver circuits in the ESS DAC's.

Really? OK - it's worth looking into

Even if a "tracking" PLL is used there's very little current involved in tracking a signal.

Well that might be so but doesn't any jitter on the signal below the PLL corner freq make it through unscathed as well as any current draws used by the PLL in dealing with SI issues? I believe I remember reading that SPDIF PLLs can be listened to as they will produce a signal on their PS which correlates to the data being handled - in other words their current draws were measurably correlated to the data being handled

Search Oversampling data recovery or some such - Dominik uses this method to extract the SPDIF in the FDAC FPGA.

Will do but I have a rough idea of the ESS approach & it's ASRC isn't sonically transparent - it overlays the sound like most ASRCs but somewhat better (edit: which has nothing to do with USB clock recovery - sorry)

Edit: I have looked into this & a good treatment of what's normally used in USB clock recovery is given in a 2011 Cypress document here

Anyway, it would seem to be something that JS could explain what he meant?

 

[JohnW]

Really? OK - it's worth looking into Well that might be so but doesn't any jitter on the signal below the PLL corner freq make it through unscathed as well as any current draws used by the PLL in dealing with SI issues? I believe I remember reading that SPDIF PLLs can be listened to as they will produce a signal on their PS which correlates to the data being handled - in other words their current draws were measurably correlated to the data being handled

Your confusing the function of a PLL to recover a Master clock which is then used for audio conversion such as with an SPDIF receiver where the PLL corner frequency matters to the very crude PLL's required to purely recover data (I'm 99% certain most USB PHY's will use Digital oversampling data techniques to recover the data then the need for "Analogue" tracking VCO type PLL circuit - there's just no need for the complexity and silicon area with such a design).

The PLL in the SMSC USB2412 Hub IC is used to multiple the 24MHz input clock to an internal higher clock rate atleast x20, but I suspect higher, its not use to track the input USB data rate.

Will do but I have a rough idea of the ESS approach & it's ASRC isn't sonically transparent - it overlays the sound like most ASRCs but somewhat better

I was referring to oversampling the input SPDIF signal to asynchronously extract the SPDIF data - nothing to do with the ESS's ASRC which can be bypassed.

 

[JohnW]

Edit: I have looked into this & a good treatment of what's normally used in USB clock recovery is given in a 2011 Cypress document here

John,

I guess you did not read the paper? as it has NOTHING to do with USB data recovery - rather how to handle the audio timing requirements of an non Async USB DAC where there is no audio rate feedback to the host USB device - which screams Apple Airplay....

USB packet timing has nothing to do with Audio feedback flow control synchronization discussed in this paper - thankfully we are all using Async DAC's thesedays

 

[Ciunas Audio]

Your confusing the function of a PLL to recover a Master clock which is then used for audio conversion such as with an SPDIF receiver where the PLL corner frequency matters to the very crude PLL's required to purely recover data (I'm 99% certain most USB PHY's will use Digital oversampling data techniques to recover the data then the need for "Analogue" tracking VCO type PLL circuit - there's just no need for the complexity and silicon area with such a design.

No, I'm not really confusing these different issues - I was just making the point that if a PLL was used in USB clock recovery then it would allow any jitter through untouched that was below the corner freq - it was a general point about PLLs & not specific to SPDIF. But if you say that most USB PHYs don't use PLLs then my point is mute

The PLL in the SMSC USB2412 Hub IC is used to multiple the 24MHz input clock to an internal higher clock rate atleast x20, but I suspect higher, its not use to track the input USB data rate.

Yes, I know that PLL is used to multiply up the clock rate

I was referring to oversampling the input SPDIF signal to asynchronously extract the SPDIF data - nothing to do with the ESS's ASRC which can be bypassed.

OK, as I said in my edit - I understand I was wrong here

 

[Ciunas Audio]

John,

I guess you did not read the paper? as it has NOTHING to do with USB data recovery - rather how to handle the audio timing requirements of an non Async USB DAC where there is no audio rate feedback to the host USB device - which screams Apple Airplay....

USB packet timing has nothing to do with Audio feedback flow control synchronization discussed in this paper - thankfully we are all using Async DAC's thesedays

I did scan through the paper & realised that it was about CYpresses approach to USB audio (even going so far as to synthesise the audio clocks from one master clock) but I was looking at it more from the overview that it gives into the then (2011) current mechanisms for clock recovery & the issues involved in various approaches. Cypress should know a thing or two about this area.

As I said however, it's probably best to let JS explain what he meant by his statements

 

[ChrisPa]

Don't respond then - we are all adults and can all form our own opinions.

Many of which will vary.

Pass me the comb.

"Knowledge is fluid" - quote from the professor of physics who lives opposite.

Stop trying to make it absolute..

That would somewhat defeat the object of a forum would it not? However for everyone's sanity that approach may be best

Not in the slightest

With freedoms and rights come responsibility.

"Freedom of speech" does not mean we can say whatever we want whenever we want. It means we are responsible enough to know what to say, how to say it and when to shut up. Possibly even, that we've learnt diplomacy.

A forum is a place for an exchange of ideas, experience and knowledge; a place to listen and learn, and if not learn at least to understand; a place to educate; a place to acknowledge and embrace our differences. As is so so obvious we seem to think that our presence on a forum conveys us the right and ability to be uncensored. That's a shame.

And "knowledge is fluid" really does sum it all up

Enough of my of topic posts now

 

[misterdog]

On a lighter note over in music central here in Sheffield myself and IANSR have just had a most enjoyable evening of live music with all the distorted reflections that live venues can offer.

Firstly we had the delights of Richard Thompson, then a quick flit across the town centre to the Yellow Arch studio to see these guys.

https://www.youtube.com/watch?v=GC0iTIUYCKk

4 hours of live music, runts and all - gotta love it...

To top it all off Houndstooth were selling their new album on vinyl at their merch. store - there is hope for the youth.

Sod the comb.

Sanity restored.

 

[iansr]

Yup it was a very good night. Spoiled somewhat by having to read through so many tedious posts after getting back to the West Riding. The guilty parties know who I mean. And well said Chrispa.

 

[JohnW]

On a lighter note over in music central here in Sheffield myself and IANSR have just had a most enjoyable evening of live music with all the distorted reflections that live venues can offer.

Firstly we had the delights of Richard Thompson, then a quick flit across the town centre to the Yellow Arch studio to see these guys.

https://www.youtube.com/watch?v=GC0iTIUYCKk

4 hours of live music, runts and all - gotta love it...

To top it all off Houndstooth were selling their new album on vinyl at their merch. store - there is hope for the youth.

Sod the comb.

Sanity restored.

Damn, once I complete these designs, I'm also going to get a life!!!

Listening to Houndstooth as I'm working away here on the FDAC PCB....

 

[BE718]

Not in the slightest

With freedoms and rights come responsibility.

"Freedom of speech" does not mean we can say whatever we want whenever we want. It means we are responsible enough to know what to say, how to say it and when to shut up. Possibly even, that we've learnt diplomacy.

A forum is a place for an exchange of ideas, experience and knowledge; a place to listen and learn, and if not learn at least to understand; a place to educate; a place to acknowledge and embrace our differences. As is so so obvious we seem to think that our presence on a forum conveys us the right and ability to be uncensored. That's a shame.

And "knowledge is fluid" really does sum it all up

Enough of my of topic posts now

Sorry but I disagree. To be blunt these issues have only arisen on this thread when a certain individual recently popped up. Someone who has a bit of a reputation from what I have been informed, and who has insulted me in other threads for simply disagreeing with his findings.

The individual seems intent on promoting a certain viewpoint whether it is informed, makes sense or not. So if views come up that I dont agree with, I certainly will challenge them.

So, sorry yes it does defeat the object of a forum if others are supposed to shut up just to appease certain individuals.

One tries to be civil but sometimes ones patience is sorely tested.

Enough said though, I dont want to waste any more of the forum space and peoples time on the issue.

 

[BE718]

Damn, once I complete these designs, I'm also going to get a life!!!

Listening to Houndstooth as I'm working away here on the FDAC PCB....

Do you sleep John? Currently listening to Amadou Balake.

Anyway, I think I know the answer to this question, but I am going to ask anyway. Signed up for the detox, but I'm not so keen on an external box.

Is it a realistic proposition for it to be internally housed inside an RF screening can? Or is this simply not physically practical or a serious compromise in terms of potential RF leakage etc?

 

[ChrisPa]

BE718

look in the mirror

This feels like a crusade. And there always has to be one more answer

 

[BE718]

BE718

look in the mirror

This feels like a crusade. And there always has to be one more answer

My comments for the most part have been patient, measured and accurate. Usually returned with distinct argumentativeness in this and other threads, therefore difficult to avoid escalation. There is no crusade.

I have no intention of discussing further. Please take it up with Tony for moderation if you feel you need to.

 

[ChrisPa]

My comments for the most part have been patient, measured and accurate. Usually returned with distinct argumentativeness therefore difficult to avoid escalation. There is no crusade.

I have no intention of discussing further. Please take it up with Tony for moderation if you feel you need to.

This isn't about how wars start it's about how wars end.
It takes two to tango. In the forum tango the dancers lead each other round the floor.

I believe people in life should be self-moderating. There have been enough hints from other posters in this thread.

There have been many opportunities to decide when to bite the tongue. There have been many earlier references along the lines of "saying no more"
and then more had been said.

 

[LarsS]

@John,

It's probably been mentioned before, memory is short. As for the digital board, what type of digital inputs have you considered for the FDAC?

 

[BE718]

This isn't about how wars start it's about how wars end.
It takes two to tango. In the forum tango the dancers lead each other round the floor.

I believe people in life should be self-moderating. There have been enough hints from other posters in this thread.

There have been many opportunities to decide when to bite the tongue. There have been many earlier references along the lines of "saying no more"
and then more had been said.

Chris,

I have just stated I do not want to discuss further. Yet you are the one persisting. Perhaps you should take some of your own advice?

 

[BE718]

@John,

It's probably been mentioned before, memory is short. As for the digital board, what type of digital inputs have you considered for the FDAC?

Further on this, John do you know how the back panel is going to be configured, connection types etc, as I think I saw reference to going back to large xlrs?

 

[sprint]

John

You have had the ReGen and JB for a while now and have undertaken a number of evaluation tests and measurements to understand their workings and processes, which have been very interesting to read and understand.

Taking all of that into account have you connected either of them into an audio system and heard any improvements/changes in the SQ as reported/claimed by the manufactures, Hi Fi mags and various individuals?

 

[JohnW]

Signed up for the detox, but I'm not so keen on an external box.

Is it a realistic proposition for it to be internally housed inside an RF screening can? Or is this simply not physically practical or a serious compromise in terms of potential RF leakage etc?

Mounting the Detox internally brings the issue if RF isolation and the need for a separated power supply - its technically correct to remove unwanted RF / LF crud before it enters the DAC enclosure.

I'm sure with the experience gained from the Detox that some of the ideas will filter down to the FDAC digital PCB.

 

[adamdea]

I suspect most PHY's recover the data using a "Clockless" oversampling method like the SPDIF receiver circuits in the ESS DAC's. Even if a "tracking" PLL is used there's very little current involved in tracking a signal.

Search Oversampling Clockless data recovery or some such - Dominik uses this method to extract the SPDIF in the FDAC FPGA.

http://www.researchgate.net/publica...Add-Drop_FIFO_for_USB2.0_High-Speed_Interface

When you oversample the input Data, SI Rise & fall times absolutely has no effect on internal circuit conditions, you just need the "Eye Opening" wide enough to recover the data, nothing internally has to "work harder" under poor SI conditions. With USB 2.0 there is no dynamic element unlike USB 3.0 which used a quite intelligent "active" receiver compensation circuits (like HDMI).

This is very interesting. That's how I thought data was extracted in the modern world.
I'm not quite sure what mechanism it leaves for interface jitter to affect the conversion clock. I can see why you are sceptical about the "swenson hypothesis" for transmission.