[project86]

Project86 just said the Izmo M1 sounds spectacular feeding SPDIF to the D2.  M1 is using a low jitter TCXO as clock.  I don't know the p/n of the clock M1 is using, it may or may not be better than the Crystek I used in your D2 mod.  In any case, TCXOs should have much lower jitter than the clocks Anedio is using in D2.  So one more data point to show that D2 has room for improvement if it is fed with a lower jitter async USB converter.

To be fair to Anedio, I never got to compare the Izmo M1 to the integrated U2 directly. And I can't until I get the USB input working again. Also I was feeding the D2 from a laptop or a modified SB Touch, but now I have the Auraliti PK90. So it's kind of apples to oranges for now.

 

[Viper2001]

Quote:

To be fair to Anedio, I never got to compare the Izmo M1 to the integrated U2 directly. And I can't until I get the USB input working again. Also I was feeding the D2 from a laptop or a modified SB Touch, but now I have the Auraliti PK90. So it's kind of apples to oranges for now.

 
The clock circuit inside U2 is the same design you'll find in SBT.  It is not going to win any low jitter contest...

 

[RedBull]

Quote:

Yeah I tried to find it archived on Google but can't find it either. There's a specific line in there about the D1 matching but not exceeding the Ref 7.1, which is what made me think he held them at roughly the same level. 

 
From my understanding of this sentence, I think I have the impression that Darko like Ref. 7.1 slightly better?
 
"The Eastern Electric Minimax, Anedio D1 and Peachtree iDecco are excellent performers in their ~$1000 price region yet none have the fluency of the Lite or the Audio-gd save for the pricier Lenehan Audio PDX. Broadly speaking, Sabre’d boxes are more caffeinated and upright in their grandeur but not forward leaning. Think ‘music’ versus ‘music’. Some forum dwellers have also complained of the Sabre sound lacking heft and suffering a diminished vocal body. At the extremity of the Sabre flavour there's to transients an airing-cupboard dryness. Despite being overplayed, such criticisms are not without foundation. The audible extent of such foibles largely depends on implementation and system matching. Personally I've always preferred Sabre DACs when sat behind tubes."
 
At that time, did you comparing Reference 7 or 7.1? and if you remember, which version of DSP was inside the loaner program?

 

[qusp]

Quote:

 
The clock circuit inside U2 is the same design you'll find in SBT.  It is not going to win any low jitter contest...

do you have the ability to measure close in phase noise at the level you are trying to improve upon? a couple inches of unshielded, inconsistent impedance wire with no ground plane, will easily undo the lower phase noise of the chd957. you might look at forgetting the silver wire and using u.fl connectors and cables.
 
TCXOs are all about long term phase stability, pretty meaningless for audio. the crysteks are about as good as it gets before going up to stuff like NDK NZ2520S, which have to be custom ordered for audio frequencies $$$. what is much more important at this level is implementation

 

[Viper2001]

Quote:

do you have the ability to measure close in phase noise at the level you are trying to improve upon? a couple inches of unshielded, inconsistent impedance wire with no ground plane, will easily undo the lower phase noise of the chd957. you might look at forgetting the silver wire and using u.fl connectors and cables.
 
TCXOs are all about long term phase stability, pretty meaningless for audio. the crysteks are about as good as it gets before going up to stuff like NDK NZ2520S, which have to be custom ordered for audio frequencies $$$. what is much more important at this level is implementation

No, I don't have phase noise measurement gear.  But the impedance of the circuit trace and IC chip are never going to match anyway, and all that impedance mismatch will bounce around in that short distance many times and died down before affecting the transition point.  The only regret I've is that I should have made that a twisted pair, to reduce EMI.
 
Surely TCXOs, atomic clocks and the likes are for long term stability, in audio, we only care about short term stability.  But most of the time, clocks with long term stability also has decent short term stability too.  So without knowing what TXCO M1 is using, it is hard to say whether it is better or worse than the -957.  But I'm pretty sure it should be better than the Pierce oscillator inside Anedio U2.
 
So what sort of phase noise figure do we get for NDK NZ2520S @ 1Hz, 10Hz and 100Hz, for say a 11.2896MHz XO?

 

[LeoDuran]

Viper,

You just passed the wife's test... LOL.
So I'm playing some "Chet Baker" (You and the Night and the Music), and shes says: "That sounds more like 'real music' now, what's changed?

Seems like the bass it's a little tighter, and mids are nice & clean... There's more separation & deeper soundstage... Great job! (Thanks again).

BTW, I had the D2 playing in a loop for about 48 hours... Just to be sure it's broken in a bit more.

 

[Viper2001]

Leo,
 
Whew, that's great to hear it passed the WAF

 
I'd like to think the mod makes the D2 more musical with better realism.
 
I noticed that the ladies often have better ears than we do.  I often get my wife to listen to my mods, to make sure I'm not off the tangent 

 
Now, only if James is willing to listen to my proposed changes.  I emailed him a few times in the last couple of months, no reply at all.  Somehow I get the impression that he is not appreciative of me modding his designs.  Funny that he did adopt my proposal of the rectifers in D2 when I started discussing about the D2 design with him in May last year.
 
The added parts cost shouldn't be that much more (since some parts are also removed).  That would make for a nice U3/D3 upgrade.
 
The other thing I see on the horizon is DSD over USB.  There is a new standard out there.  But that would require a major architectural change in U2 and D2.  And whether DSD will bring better sound than high-res PCM is still a subject of debate.

 

[DaveBSC]

Quote:

Leo,
 
Whew, that's great to hear it passed the WAF

 
I'd like to think the mod makes the D2 more musical with better realism.
 
I noticed that the ladies often have better ears than we do.  I often get my wife to listen to my mods, to make sure I'm not off the tangent 

 
Now, only if James is willing to listen to my proposed changes.  I emailed him a few times in the last couple of months, no reply at all.  Somehow I get the impression that he is not appreciative of me modding his designs.  Funny that he did adopt my proposal of the rectifers in D2 when I started discussing about the D2 design with him in May last year.
 
The added parts cost shouldn't be that much more (since some parts are also removed).  That would make for a nice U3/D3 upgrade.
 
The other thing I see on the horizon is DSD over USB.  There is a new standard out there.  But that would require a major architectural change in U2 and D2.  And whether DSD will bring better sound than high-res PCM is still a subject of debate.

 
I'd like to think I had a small hand in helping to create Empirical's Short Block filter. I mentioned to Steve a long while back that there wasn't much out there besides the SoTM card, pretty much everything else couldn't pass 24/192 or didn't provide any actual ground filtering. I asked about what Empirical could do with USB filtering, and lo and behold...

 

[Viper2001]

Quote:

 
I'd like to think I had a small hand in helping to create Empirical's Short Block filter. I mentioned to Steve a long while back that there wasn't much out there besides the SoTM card, pretty much everything else couldn't pass 24/192 or didn't provide any actual ground filtering. I asked about what Empirical could do with USB filtering, and lo and behold...

 
So are you using the short block filter?  How does it improve the sound?

 

[qusp]

seriously guys, look into the fifo and GB thread
 
it makes all this stuff fairly meaningless, groundbreaking, cheap (I doubt you could even build one yourself for this price) and awesome. if you have the ability to hook up clocks in your dac and only need 2 channel, this thing is pretty much endgame. you can then use optical for full galvanic isolation and done. optical and i2s input into this is the same to these ears, as it should be. i'm using it with my ESS and running it as master clock too. this way I can use lowest DPLL settings, which is a good indication of the jitter. the combination of the clock used plus the buffer on the clock board is the sum total of the jitter in the system, in my case less than 1ps
 
I have no affiliation other than being an early adopter. think I mentioned it to you before too Dave
 
the quality of the design, layout, software, support are impeccable, there are megabuck dacs that dont have the results this thing produces.
 
btw DSD over USB is more than on the horizon amanero and GB (if youre quick, we are up to 100 units)

 

[Viper2001]

I read the FIFO thread on DIY Audio the other day.  The work Ian did is so pro, it is a rarity.
 
The concept is not new though.  PS Audio Genesis Audio Lens, and even SqueezeBox has the same idea.  Just think of SqueezeBox as a Wi-Fi DAC, with a FIFO in between.  The parts where the SqueezeBox failed are that they did not reclock the output with a Picogate flip-flop and did not use a good clock.  If they did, the sound would be really up there.
 
Anedio U2 did the reclock with Picogate flip-flop trick, except they did not have good clock.  This is why out of all things in the D2, I only modded the clock in U2.  Obviously, lots of manual labor was involved and not some automatic assembly like Ian's FIFO board to reduce cost.
 
So what clock are you using with Ian's FIFO?  CCHD-957?
 
I know there are hardware that does DSD over USB, Mytek 192 Digital being one commonly available one.  The questions are implementation and software.  Whether the hardware is going to sound good depends on the implementation and the clock.  As for software, how many native DSD encoded music are out there?  Doing SACD rips is no easy task, one needs to have the original PS3 with hacks in order to do that.  And even that, many SACDs were not recorded in DSD to begin with.
 
Finally, don't expect the CCHD-957 to sound superb automatically with just a low noise or high PSRR supply. There are tricks involved, as I've been playing with this XO since Nov last year

 

[DaveBSC]

Quote:

 
So are you using the short block filter?  How does it improve the sound?

I am not, but the beta testers have had very positive results. It's a common mode choke combined with some other filtering. The primary purpose is to stop ground noise from the PC dead, before it gets to the converter or DAC. It also severs the 5V Vbus line, so the U2, presumably D2 and most other converters and DACs are out. The Legato should work though.

 

[Sapientiam]

Quote:

I am not, but the beta testers have had very positive results. It's a common mode choke combined with some other filtering. The primary purpose is to stop ground noise from the PC dead, before it gets to the converter or DAC.

 
Perhaps 'stop... dead' is just hyperbole here. Common mode noise from a SMPSU comes from a very high source impedance (typically 100s of kohms at lowish freqs) so stopping it 'dead' would require a lot of chokes in series to get sufficient impedance. That's not to say that CM chokes don't work - they do a great job, but no way do they kill the noise.

 

[DaveBSC]

Quote:

 
Perhaps 'stop... dead' is just hyperbole here. Common mode noise from a SMPSU comes from a very high source impedance (typically 100s of kohms at lowish freqs) so stopping it 'dead' would require a lot of chokes in series to get sufficient impedance. That's not to say that CM chokes don't work - they do a great job, but no way do they kill the noise.

 
Perhaps. To be fair to Empirical they don't claim to eliminate common mode noise entirely, just greatly reduce it.

 

[Orangecrush]

Received my Aqvox power supply. After a couple of days I honestly can not hear any difference. So either the Anedio D2 USB is really well designed or the SBT USB power is pretty clean.

I am using DH Labs newly designed USB cable.

Note, I have no cans but by my two channel system is pretty revealing.