[FredA]

I
Lol, I probably would/could.

Incidentally I don't think it likely the audible effects of the damping I'm hearing are in relation to external vibrations as I'd think it unlikely those would be anywhere near the MHz carrier frequency of the OCXO. I think it more likely the vibration culprits the damping is mitigating are low amplitude high frequency internal vibrations, resonances from the oscillator itself and proximate circuitry. Just a theory of course.

Well I could just leave it there there, and probably should, but in the spirit of fearless exploration / taking one for the team I'll share my latest iteration for your interest & amusement.

It seems OCXOs - or at least this LHY OCXO - are like the Princess and the Pea - audibly sensitive to vibration (incl the particular vibration damping materials used) some 9* (yes 9!) layers remote. No question it is more than just the aggregate mass it is sensitive to.

The four thin black layers in between each of the five cubes are fo.Q SH-22K, a three layer 2mm thick material employing 'constrained layer damping' via a softer more elastic middle layer between firmer outer layers, per fo.Q piezo electric rubber and polymer compounds respectively. Adding the much harder tungsten layers in between means the whole stack is effectively a 17 layer thick (5 + 4*3) constrained layer damping structure.

The combined tungsten & fo.Q sandwich sounds better than the tungsten stack alone - less edgy, more full bodied midrange, easier to listen to.

Now I wish I couldn't hear it (my life would be simpler and I can't now unhear it) but i can hear without any trouble the difference between a narrower 1x2cm fo.Q strip and the 2x2cm square you see in shot.. at the top, the 8th (!) layer. That's how sensitive this OCXO seems to be, a princess indeed.

The tightness, weight, speed and resolution of the bass and overall life and dynamics of the music is now something to behold.. I thought it was good before from the addition of the first tungsten cube then cubes plural, and it was, now it's mental. As is the soundstage which has progressively exploded outwards, particularly with the last two cubes & fo.Q sheets. The midrange progressively lost grain and edges as I went up the layers, smooth as silk by the 5th cube/ 9th layer*. Going down a level or two or eight the bass blooms and loosens, the soundstage and dynamics compress a little then a lot, an unwelcome edginess/glare/shoutiness creeps back in.

All the above is, I remind you, 'just' the internal OCXO of a network switch(!) I have feeding my streamer. I look forward to trying this on my external clock once my main system is reassembled.

To be continued...

* Postscript - so I still had a 1cm cube sitting quietly to one side unused as the stack was precarious enough already, surely. But what the heck I thought just now, in for a penny in for a pound. Sure enough, adding it in as the 6th cube /11th layer takes the sound up another notch - the returns are not yet diminishing. Au contraire, I reckon the delta on this cube/fo.Q layer is at least as great as the preceding one, certainly in terms of midrange refinement. Where does it end, does it end..? I am quitely chortling to myself and shaking my head here, as it is patently ridiculous, precarious and impractical yet it is has brought an awesome sound improvement for comparatively little outlay... What to do.

You could likely achieve similar results applying damping feet underneath and some mass on top. It is a very old trick.

My sw8 is fed with a massive cord, maybe the mass of the cord itself helps to some extent, not just the lower resistance. It made a difference. I have it sitting on a rosewood plate now. Something I got from Afterdark. Can likely absorb some vibrations.

I noticed my Zerozone psus need damping feet to perform at their best. Rubber is to avoid, too elastic, little damping. Grounding boxes seem to like damping feet too.

 

[TheRealDz]

Out of curiosity, what would have a bigger sonic impact for a DAC: a clock connection or I2S connection?

iFi just released a new DAC, and it has a clock connection, but not an I2S connection - which I thought was a bit strange.

 

[mgsu]

iFi just released a new DAC, and it has a clock connection, but not an I2S connection - which I thought was a bit strange.

It is not strange at all. It makes perfectly sense from a technical standpoint.

The DAC is slaved over I2S connection to the DDC/Streamer clock. The DAC is ignoring its own clock when connected over I2S because it is getting the clock signal over I2S from the DDC/Streamer

In case of the IFI DAC, you clock the IFI DAC and feed it with USB which is asynchron and does not carry a clock signal.

 

[TheRealDz]

It is not strange at all. It makes perfectly sense from a technical standpoint.

The DAC is slaved over I2S connection to the DDC/Streamer clock. The DAC is ignoring its own clock when connected over I2S because it is getting the clock signal over I2S from the DDC/Streamer

In case of the IFI DAC, you clock the IFI DAC and feed it with USB which is asynchron and does not carry a clock signal.

So all else equal, you would prefer a DAC with an outbound clock, to a DAC connected to a DDC via I2S?

 

[mgsu]

So all else equal, you would prefer a DAC with an outbound clock, to a DAC connected to a DDC via I2S?

I would prefer a DAC with a clock input because there is no DDC with a clock input on the market.

Feeding your DAC from a DDC you rely on the DDC internal clocks which are cheap Accusilicon or Crystek clocks costing $20, not OXCO what you get in a good external clock.

Edit: Gustard U18 DDC has 10MHz clock input but is not a clean DDC (SNR) as Singser SU-6 which does not have a clock imput.

Best way is to avoid DDC and pick up a DAC with clock input and galvanic isolated USB input. You get the best SQ.

 

[JaMo]

because there is no DDC with a clock input on the market.

Not true. Audio-gd DI20HE/DI20 and Gustard U16 and U18 are examples that have and benefit very well using their EXT XO inputs. I am pretty sure there are more of them out there.
/Jan

 

[mgsu]

Not true. Audio-gd DI20HE/DI20 and Gustard U16 and U18 are examples that have and benefit very well using their EXT XO inputs. I am pretty sure there are more of them out there.
/Jan

thanks, already edited my post

 

[MartinWT]

I noticed my Zerozone psus need damping feet to perform at their best. Rubber is to avoid, too elastic, little damping. Grounding boxes seem to like damping feet too.

I use Black Ravioli footers on my Zero Zone PSUs and all my Quartz Acoustic grounding boxes.

 

[MartinWT]

I would prefer a DAC with a clock input because there is no DDC with a clock input on the market.

Umm, no. The Gustard U18 DDC has a clock input and I feed it with my clock, which goes via I2S to the X26 Pro DAC and feeds that, too.

 

[FredA]

I use Black Ravioli footers on my Zero Zone PSUs and all my Quartz Acoustic grounding boxes.

Dam, these are ridiculously expensive.

 

[camrector]

So all else equal, you would prefer a DAC with an outbound clock, to a DAC connected to a DDC via I2S?

It’s not equal.
If you want to falsely color your signal with “Harmonic spurri” and add jitter then add an external clock.

 

[MartinWT]

Dam, these are ridiculously expensive.

Sadly, they are. However, I have a whole box of different footers I've bought in an attempt to beat the Black Ravioli: no joy. They are simply the best.

I've built up quite the collection of Pads and Big Pads over the years.

 

[TheRealDz]

It’s not equal.
If you want to falsely color your signal with “Harmonic spurri” and add jitter then add an external clock.

I am not sure what you mean here.

 

[camrector]

I am not sure what you mean here.

Despite what people here will tell you, an external clock does not improve performance that then translates to audible improvement.
What it does do is add jitter and harmonics that can change the tonal qualities of the signal. Think like a filter in HQplayer or an EQ, but you can’t control it or turn it off.
Your enternal clock that’s right next to the dac chip or ladder will always perform better than anything external.

Get a good DDC and use I2S.
External Clocks providing sonic improvement is just marketing. Everyone of them adds jitter.

 

[Getafix]

Those cubes are putting unnecessary weight on the circuit board. Solder is not exactly flexible, especially lead free solder and the ubiquitous ceramic smd caps are prone to minuscule flex cracks. They are very hard to see.