Master Clock for your EtherREGEN

[MartinT]

2 hours ago, Iving said:

This has two implications I don't understand!

1) EMI/RFI shielding variability is not manifest on the oscilloscope waveform.

2) SQ is not a function of wave form only. An ethereal alternative?

 

We already know that noise (or lack of) is incredibly important in both the signal and clock.  The 'bad stuff' is in the high MHz and GHz range, well above the signal/clock frequencies and virtually undetectable on an oscilloscope (because it's superimposed on the signal).  You would need a high pass filter to see and measure it.

 

I believe this is the area where variations in sound quality may occur?

[MartinT]

You did even better than me (for a King) of -108dB and -138dB respectively.

 

I'm very happy with it.

[MartinT]

6 minutes ago, R1200CL said:

Did you open it ?

 

To look inside?  I confess that I was in too much of a hurry to get it hooked up.  Now that I've listened to it, I'm even more reluctant to disconnect it for any reason!

 

Also, I would honour Adrian's wishes not to publish internal photos.

[MartinT]

4 minutes ago, R1200CL said:

He hasn’t asked that.  Actually people should know what they’re buying. 
 

You will be surprised what’s inside 😀

 

 

Actually, I doubt it.  More to the point, though, having heard what it does I really don't care what's inside.

[MartinT]

5 minutes ago, GMG said:

@MartinT looks like the itch was to strong on you 😀

Enjoy the new addition 

 

Yes, it was. Not very expensive, though, in the context of my system.

[MartinT]

2 hours ago, R1200CL said:

Interesting POW.
If you’re buying a 911 with a beetle engine, but the HP is stated to be 300, (equal to an original 911), I wonder if you would say the same 😀

 

That's a poor comparison.  If  drove the 911, I would know something was wrong immediately.

 

I've listened to the AfterDark and have only admiration for the sound quality I'm achieving.

 

I've asked you before: have you actually listened to your AfterDark?

[MartinT]

 

32 minutes ago, R1200CL said:

 

I like to see others post their measurements from AfterDark. Mine attached. 
 

I can scan and post mine when I get home.

[MartinT]

35 minutes ago, MarkusBarkus said:

I *think* you may think if a company doesn't directly design and develop every component in its product, there is something bad going on. 
.

.

.

A company may purchase 100 OCXO units, which have the same spec, but upon testing, some might test *better* than others. The best ones may be used in product A$$ while others may be used in product B$.

 

Exactly this.  There are countless OEM arrangements out there, from simple re-badging to module provision.  Using another design does not in any way denigrate the product if it performs to specification.

 

Performance selection and categorisation also happens, which is what I believe AfterDark are doing.

[MartinT]

Attached is my AfterDark certificate.  It appears to be several steps above a King in spec.

AfterDark Giesemann King.PDF

[MartinT]

I don't any more as I now have two clocks, one for each of the ER and Mutec.

[MartinT]

On 7/21/2021 at 10:01 PM, jean-michel6 said:

Have you ever used a clock signal splitter to clock two devices with only one master clock .

 

If yes what were the results ? 
 

Since you have moved now to two separate clock what is the benefit over the splitter ?

 

 

I only used the Mini-Circuits splitter temporarily while awaiting the second clock and only had the system arranged this way for a few days, due to the timing of arriving components.

 

Now, with the two AD clocks each driving one component, there is frankly no comparison.  The King drives the ER while the Emperor Triple Crown drives the Mutec.  Everything that clocks do - finer detail, incredible soundstage width and depth, startling focus, note decay - is enhanced.  Music snaps into space with incredible punch and clarity, yet sounds effortless.  I am so pleased I went this route.

 

I am currently experimenting with more grounding boxes which is the only reason I've gone quiet here.

[MartinT]

1 hour ago, R1200CL said:

@MartinT

I understand you’re using two clocks now. Can the Mutec output 10 MHz? How are you using the Mutec ?

Do you happen to know Mutec’s phase noise numbers?

 

Yes, I'm using an AD King to externally clock the ER, then an AD Emperor Triple Crown to clock the Mutec MC-3+ USB.  I don't know the Mutec's phase noise numbers but the external clock certainly improves its performance.

 

The Mutec does not output 10MHz.

 

My Mutec is modded for use with an external 5V power supply (a Coherent QP-1) and also has other components swapped for improved performance.

[MartinT]

4 hours ago, R1200CL said:

@MartinT

Thanks for replying. How is the Mutec implemented in your chain ?

 

 

[MartinT]

For an inexpensive experiment, you could try a BG7TBL to hear what it's all about.

 

Next up would be the AfterDark units, which I can thoroughly recommend as I run two of them.  Choose your price level and go for it.  Sinewave is fine and they don't need filters.

[MartinT]

On 8/20/2021 at 7:45 AM, Johnnydev said:

 

Did you tried with and without the filters? And what is the difference in SQ.

 

Yes, and the filter made the AD unreliable and worsened sound quality.

[MartinT]

7 hours ago, Johnnydev said:

Thanks, and what DC cable are you using?

 

For that AD, an MCRU copper DC cable. For my better AD feeding the Mutec, a Coherent DC cable.

 

[MartinT]

7 hours ago, Johnnydev said:

sorry, I asked the wrong question, I meant which BNC cable are you using?

 

Hah!  Both are Belden 4694R, very good clock cables.

[MartinT]

I can't edit the old post so here is an updated system diagram.

 

[MartinT]

8 minutes ago, FIndingit said:

Looks like you believe dac is the least important link in the chain. 😁

 

Not at all, but the performance you get from your DAC completely depends on the quality of the datastream you give it.  Reducing phase noise and jitter are paramount.  Thus the reclockers and grounding boxes.

 

Saying that, my new Gustard is an outstanding DAC, very high quality indeed.

[MartinT]

1 hour ago, dylanesque said:

Curious if you found perceptible differences with a King vs Emperor at the Mutec?

 

Yes, I did, as I bought the King first to use with the Mutec.  I then ran my BG7TBL with the ER.  Upgrading the King to Emperor Triple Crown gave noticeably more of what clocks do: soundstage, timing, detail.  I then relegated the King to use with the ER.

[MartinT]

I've already tested splitting the Emperor clock using a Mini-Circuits splitter and feeding 10MHz into the Gustard.  It's very good, but then the Gustard's K2 internal clock is also very good.  I think the AD wins but there is very little in it.

 

The router is on the P10 (but a separate Zone from the power amp) because I'm using all the outputs on the P3, and those components are more critical.  I've tried the router on straight power without the regenerator, but the system sounds better with it on the P10 and being powered with an LPSU.

[MartinT]

@JohnSwenson do you generally advocate sinewave clocks over squarewaves, or did you just use the Minicircuits for the Morion?

 

I ask because I prefer sinewave over squarewave in all my listening tests.

[MartinT]

If it's not too late to make a request, I would love for the EtherREGEN Gen 2 to have the SFP cage facing the optical link for best performance.  I am using my original EtherREGEN in 'reverse' mode as do others so that I can use the ER 'A' side at one end of a link and my Signature Rendu SE at the receiving end.

 

That slight additonal phase noise that @JohnSwenson mentioned when using the ER in reverse would be nice to recover in the Gen 2.

[MartinT]

6 hours ago, GoodEnough said:

Any thoughts anyone?

 

 

I would go with a better LPSU.  If you're going to power both the ER and the external clock, make sure a dual power supply has isolated outputs, or use two power supplies.  You need to avoid shorting the isolation provided by the moat.

[MartinT]

18 hours ago, Superdad said:

Power to the EtherREGEN is to its 'A' side -VE/0-Volt "ground" domain. The 'B' side gets power also, but via an isolating converter.

The "ground" of the external clock connection is in the 'B' side domain.  

 

So if you use a dual-output power supply--where the -VE/0-Volt "ground" is shared by both outputs--to power both an 'A' side and 'B' side device, then you will do a defeating end-run around the EtherREGEN's special and wholly unique active-differential isolation moat.

 

12 hours ago, dbastin said:

One psu to power ER and external clock is not defeating the moat isolation because tharse are all on the same ground domain.

 

 

See Superdad's response above.  If the grounds are common, you will be defeating the moat.