Amir's Music Server Build Thread

[Mivera]

Link returns a 404 error.



Link shows plenty of eye candy but no explanation for how it might be necessary.

http://jlsounds.com/i2soverusb.html

 

[Opus111]

That flip flop is perhaps guaranteeing a cleaner transition between states before being fed to the DAC, but it's not a clocking operation - effectively it's a buffer/filter in its operation.

The only way I can at present think that an extra FF brings anything to the party is if the clock's been sent across an isolation barrier and picked up some jitter in the process. But that's just a band-aid for shoddy implementation, with async the DAC needs the clock sitting right next-door.

 

[Opus111]

http://jlsounds.com/i2soverusb.html

Again no explanation for why in an engineering sense 'reclocking' performs any useful function. Perhaps they've not put the clock right next to the DAC to begin with? If so, see my post directly above this.

 

[Mivera]

The only way I can at present think that an extra FF brings anything to the party is if the clock's been sent across an isolation barrier and picked up some jitter in the process. But that's just a band-aid for shoddy implementation, with async the DAC needs the clock sitting right next-door.

Well I don't need the concept proven to me. I've heard the results first hand. It's allowed me to use any computer with the same sound results.

 

[Opus111]

There are others here reading besides you - you're apparently misleading them with your claims.

 

[Mivera]

There are others here reading besides you - you're apparently misleading them with your claims.

I suppose any DAC designers reading can try for themselves then. Me personally I'm done with USB anyways.

 

[Opus111]

Its great you got a result that you can listen without worrying about the source component, that's the main thing.

 

[Mivera]

Its great you got a result that you can listen without worrying about the source component, that's the main thing.

We will see how effective Sonny's approach is when a client of mind tries a Sonore Microrendu powered by Ifi Ipower in place of a NUC 3815 powered by cheap SMPS.

 

[fas42]

The only way I can at present think that an extra FF brings anything to the party is if the clock's been sent across an isolation barrier and picked up some jitter in the process. But that's just a band-aid for shoddy implementation, with async the DAC needs the clock sitting right next-door.

Yes, the clock is next to the DAC - but another signal with transitions is that of the data itself, the DIN pin adjacent to the CLK pin. What is the actual level of "jitter", noise on that DIN waveform, how clean is it? Yes, should mean very little, but depending on the chip it may be significant - so, an option is to do some grooming of the quality of that signal line - may help, or may not ...

 

[Opus111]

Ah I see your point, I'd call that something like 'de-noising' rather than de-jittering. Over on DIYA some of the guys into TDA1541s have been using I2S attenuators for many years to condition the generic logic levels to suit the specific thresholds of the chip.

 

[Phison Audio - Sonny]

I use reclocking as the isolation barrier adds time distortion to the transfer from one zone to another. It is documented in the datasheets. Time distortion in the way that the delay from level shift on one side to another is not uniform. The DAC in PCM mode uses the master clock for internal digital filtering. DSD jumps over this part, so it needs to be clean from the beginning before entering the DAC

 

[Mivera]

I use reclocking as the isolation barrier adds time distortion to the transfer from one zone to another. It is documented in the datasheets. Time distortion in the way that the delay from level shift on one side to another is not uniform. The DAC in PCM mode uses the master clock for internal digital filtering. DSD jumps over this part, so it needs to be clean from the beginning before entering the DAC

Isn't time distortion another term for Jitter?

 

[Phison Audio - Sonny]

Depends on how you look at it.

Analog devices solution is made up of two coils that has a attached logic that detects changes in the field because the coils has no static state. So this logic try to predict when to switch. The will switch in 100% of all cases, but there can be a small delay.

Then we have jitter introduced in mismatched transmission lines,
Then we have noise on the supply voltage messing up the the transition level.
The noise on the supply voltage is due to currents flowing when changing state or level. Best way to deal with this is ceramics capacitors dealing with DC to GHz not audiophile caps

 

[Mivera]

Depends on how you look at it.

Analog devices solution is made up of two coils that has a attached logic that detects changes in the field because the coils has no static state. So this logic try to predict when to switch. The will switch in 100% of all cases, but there can be a small delay.

Then we have jitter introduced in mismatched transmission lines,
Then we have noise on the supply voltage messing up the the transition level.
The noise on the supply voltage is due to currents flowing when changing state or level. Best way to deal with this is ceramics capacitors dealing with DC to GHz not audiophile caps

Well I can vouch for your reclocking circuit working good. The Exasound DAC I had used galvanic isolation, as well as a very large FIFO buffer for the Async USB. They boasted up and down about being completely isolated from connected computers, but I could still hear difference in the source computer. Same with the Resonessence Mirus. The Amanero passes noise through like crazy. It's the worst I've tried. Your Mirand DAC is quite sensitive to the source used as well. However the PD2 was much less sensitive. Chris will be doing a bunch of experimenting with 2 of my streamers, a Sonore Microrendu, as well as a few power supplies.

 

[dctom]

Hi Amirm

thank you for your response to my question about galvanic isolation on wtb forum.
tried to reply on wtb and cannot find the thread. The link on my email alert does not work for some reason.

best
Dave

 

[amirm]

My pleasure Dave. Good to see you here anyway .

As to the thread on WBF, it seems gone. I will create it again when I have time.

 

[dctom]

Thanks Amirm

 

[dallasjustice]

Amir,

I found a great power supply for your router. It's on sale for only $750!
http://www.computeraudiophile.com/f...mponents/paul-hynes-sr5-12v-28885/#post552111

 

[FrantzM]

Amir,

I found a great power supply for your router. It's on sale for only $750!
http://www.computeraudiophile.com/f...mponents/paul-hynes-sr5-12v-28885/#post552111

And based on the pictures it really output 12 Volts!!

 

[amirm]

Amir,

I found a great power supply for your router. It's on sale for only $750!
http://www.computeraudiophile.com/f...mponents/paul-hynes-sr5-12v-28885/#post552111

I am actually created a test setup for power supplies! Will test the ones I have now before testing others.