Master Clock for your EtherREGEN

[JohnSwenson]

Please stop asking me to give a detailed breakdown and measurement of a whole bunch of different clocks, it is not going to happen!!!! Do I have to say it again? It is not going to happen. I simply do not have the money to buy a whole bunch of them nor the time to do a full analysis of every clock out there.

 

I can say the AfterDark boxes do NOT use $20 used OCXOs, they are new modules, but I do not know which ones are used in the different models. That's it for the AfterDark clocks.

 

I AM working on my detailed paper on differences between sine and square, 50 ohm and 75 own and different types of cables. NO specific models for any of this, this is generic information. You guys asked for it, but believe me it is just going to make things more complicated! This is subject is NOT cut and dried.

 

I am very busy working on new product designs and doing a whole bunch of clock stuff will delay these by at least year. Not going to happen.

 

John S.

[Superdad]

Hey guys:

It is all going to be okay (at least in our audio world; hard to say the same about the rest of the world given all the awfulness going on).  

But I will confess that I cried when I read John's post above. 

I will explain my reaction and also why I feel responsible.  

 

I cried because if any of you knew John personally as I do, you would know that he is the sweetest, most kind-hearted human--and one for whom it is nearly impossible to make angry or to get under his skin.  So for him to write what he did, with the tone that he did, is really unusual.  

 

I interpret it (and mostly correct I am sure) to mean that:

a) He is feeling rather stressed, be it from how hard he is working on several (still-secret) UpTone products plus a key project for Sonore, and/or from a bit exhaustion (from work, life, and other obligations);

 

b) By my sending him both the BG7TBL and AfterDark clocks to look at and play with I have put him in an uncomfortable situation (for us both really).  He has dissected and examined the circuits of both units and verbally relayed to me a raft of technical comments, some good and some bad--about their design and construction.

Yet, for a range of reasons, I have asked that he not publicly discuss his findings about these clocks.  

 

I feel no obligation to explain my rationale in detail. Yet a couple of obvious reasons are that: 1) These are commercially offered products and it is not appropriate for us to disclose or critique their designs; 2) These are extremely affordable offerings--in a market segment where very high performance is difficult to obtain at low cost. To expect that these are world-beating pieces, rivaling the design prowess of a Mutex REF10 or Cybershaft OP20A would be foolish.  We can all enjoy fast-food sometimes, but do we really want to know all the ingredients or to compare the meal to that from a fine restaurant? 

 

So I will take the blame for this. I was the one who mentioned the AfterDark clock offerings.  I did so as a gesture since Adrian at AfterDark does a very fine job as a dealer for UpTone in Hong Kong/China. And where else will one find well-performing built external 10MHz clocks--with cool carbon-fiber end panels--in the $300~$900 range?

 

But let's move on.

As John mentioned, he has for a while been thinking about the 'paper' he wants to write concerning the many considerations regarding sine wave versus square wave clocks, cabling, impedance matching, and the circuit internals it takes to actually do any of this correctly.  He will get to it at some point, perhaps in the week while waiting for development boards to come back.

 

Trust me: Based on all of what he has conveyed to me verbally on the topic, much is not what it seems and the correct choices are far from black and white!  If there were we--and others--would have come out and stated the answers long ago. Like probably most of you, when making a product purchase decision, I don't really want to understand all the engineering and considerations. I just want to know enough to make a confident, informed choice.  Yet the truth of what goes on in the realm of reference clocks is filled with mystery. And the manufacturers and marketing writers are of little help!

 

Okay, back to my endless battle with e-mail. And I think we just sold-out all of February EtherREGEN production! Dang, have not even had a moment to chase down more parts for the next 250 units...

Cheers,

--Alex C.

[MartinT]

For my part, I am hugely indebted to both Alex and John for all the information and encouragement they have given to those of us who have experimented with an external master clock.  I started this thread really to discover whether a ridiculously inexpensive clock sold for radio hams would work for audio applications.  It does, rather well - job done!

 

I believe quite a few of us have these clocks now and are very happy with them.  Thanks again.

[PYP]

4 hours ago, R1200CL said:

I have several times asked what clocks is in use. 

While it is clear many folks are looking for clocks that are less expensive than the Cybershaft OP13, I'm listing the screening criteria Cybershaft uses.  For the Cybershaft OP13 model specifically:      

 

Phase noise offset 1 Hz:  -113 dBC / Hs or less

Allan Deviation:  TAU = 1 sec; less than 5.6E-13(0.00056ppb/s)

 

Each unit includes a Certificate of Calibration for that unit.  Here is mine:  Cybershaft OP13 certificate of calibration.pdf

[soares]

Hi there, I have acquired the following clock  but unfortunately it isn’t working. I followed the instructions from the eR manual but so far no lights on the Ethernet port. Any clues? Tks Jorge

[JohnSwenson]

19 minutes ago, soares said:

Hi there, I have acquired the following clock  but unfortunately it isn’t working. I followed the instructions from the eR manual but so far no lights on the Ethernet port. Any clues? Tks Jorge

What are you powering the clock with? It takes 12V. Does the green light on the box come on?

 

The procedure is to turn off the ER (unplug ITS power, move the little slide switch on the back so the lever is near the BNC jack. Plug the BNC cable in between the clock box and the ER (it doesn't matter which jack on the box you use). Power up the clock box, make sure the green LED on the clock box is lit up. THEN plug the power back into the ER. This should work.

 

John S.

[soares]

22 minutes ago, JohnSwenson said:

What are you powering the clock with? It takes 12V. Does the green light on the box come on?

 

The procedure is to turn off the ER (unplug ITS power, move the little slide switch on the back so the lever is near the BNC jack. Plug the BNC cable in between the clock box and the ER (it doesn't matter which jack on the box you use). Power up the clock box, make sure the green LED on the clock box is lit up. THEN plug the power back into the ER. This should work.

 

John S.

Oh John,

So kind from you... Really thank you!
Yes, I tried with the stock SMPS and also with an LPS 1.2 set to 12 v. The light from the box is green. I also followed the instructions from the eR manual (the same as the ones indicated by you). Could it be a 50 Ohms unit sent by error? Is there a way to verify it? The eR works perfectly. 
Thank you again for any additional input you may have. Cheers Jorge

[JohnSwenson]

11 minutes ago, soares said:

Oh John,

So kind from you... Really thank you!
Yes, I tried with the stock SMPS and also with an LPS 1.2 set to 12 v. The light from the box is green. I also followed the instructions from the eR manual (the same as the ones indicated by you). Could it be a 50 Ohms unit sent by error? Is there a way to verify it? The eR works perfectly. 
Thank you again for any additional input you may have. Cheers Jorge

The impedance is not going to mater as far as working or not is concerned. Have you tried all three outputs? Unfortunately there is no easy way to tell if there is actually output (well there is, but it takes an oscilloscope). If you have a multi-meter you can check continuity of the cable. Probably at least half the time the problems are caused by the cables.

 

John S.

[soares]

Tks again John. You were right. The third output worked. I was really surprised with what I heard. Nothing about timbre but the details were terrific. Wow. It’s a pity that you’re not available to pursue that road... Imagine the success of an etherClock. Just saying. ☺️🤗☺️ However the connection is not stable and after a couple of tracks I must reboot the oR. It is connected with fiber to the eR. Probably the clock needs to stabilize. Let’s see if it improves over the night. Warmest regards. Jorge

[Superdad]

1 hour ago, soares said:

However the connection is not stable and after a couple of tracks I must reboot the oR.

Yes, as previously explained, the EtherREGEN (‘B’ side in particular) will not tolerate even the slightest hiccup in the clock. It will not recover (without reboot) after a glitch even as short as 3uSec. So clock and cable must be perfect.

[MartinT]

Try as short a clock cable as you can manage. What is the cable you are using?

 

Also, make sure the BNC connectors are fully secured both ends.

[Roasty]

I've just sent my sotm sms-200ultraneo back to sotm to install a 75ohm clock input. 

 

With regards to the clock cables.. I have that Geistnote Apogee one from amazon, and another from Ghent Audio (I think the cheap belden one). 

 

Has anyone tried pricier options? I've read on some forums that basically any 75ohm digital coaxial cable with bnc plugs work as clock cables as well. 

[R1200CL]

@Roasty

I think you should request a new clock from HamRadioShop.

Yes, any 75 ohm coax works. But as you probably know, the idea is to keep the cable short.

What I’m wondering, is if aJSSG360 version, is expected to have a positive effect. I know Ghent can make from Belden 4694R upon request. (Or could). 

[R1200CL]

One interesting observation is that the clocks we get from Ham is sine wave that’s converted to square wave. At least that’s my understanding. 
 

So I’m asking myself why he isn’t using pure square wave clocks ?

https://www.ebay.com/itm/MORION-OCXO-MV102-10MHz-squarewave-Crystal-Oscillator-12V-EFC/143856696386?hash=item217e86fc42:g:vp8AAOSwkz1eYVRS

 

https://www.ebay.com/itm/DAPU-O55F-ASDD-10MHz-OCXO-Crystal-Oscillator-12V-EFC/143045716607?epid=28026910829&hash=item214e306a7f:g:a8YAAOSwW4hceKjd
 

Maybe something to ask Ham about. 
(I just tried via eBay). 
 

Maybe data sheet will indicate that sine version is better ?

[Confused]

5 hours ago, Roasty said:

I've just sent my sotm sms-200ultraneo back to sotm to install a 75ohm clock input. 

 

With regards to the clock cables.. I have that Geistnote Apogee one from amazon, and another from Ghent Audio (I think the cheap belden one). 

 

Has anyone tried pricier options? I've read on some forums that basically any 75ohm digital coaxial cable with bnc plugs work as clock cables as well. 

Not a direct answer to your question, but I while ago I started a thread with the aim of trying to pull together recommendations for clock cables.  See link below.  The thread never quite got the interest that I had hoped that it would, but there are a few interesting tips there.  And if anyone else has some recent clock cable experience, feel free to post per the thread linked below!

 

 

[AfterDark.]

This is called SMA connector. The recommended cable is HUBER+SUHNER SMA Shielding (Made in Swiss) 

 

 

[AfterDark.]

2 hours ago, R1200CL said:

One interesting observation is that the clocks we get from Ham is sine wave that’s converted to square wave. At least that’s my understanding. 
 

So I’m asking myself why he isn’t using pure square wave clocks ?

https://www.ebay.com/itm/MORION-OCXO-MV102-10MHz-squarewave-Crystal-Oscillator-12V-EFC/143856696386?hash=item217e86fc42:g:vp8AAOSwkz1eYVRS

 

https://www.ebay.com/itm/DAPU-O55F-ASDD-10MHz-OCXO-Crystal-Oscillator-12V-EFC/143045716607?epid=28026910829&hash=item214e306a7f:g:a8YAAOSwW4hceKjd
 

Maybe something to ask Ham about. 
(I just tried via eBay). 
 

Maybe data sheet will indicate that sine version is better ?

 

The quality of clock is determinate by Phase Noise and Allan Derivation on frequency accuracy, which can measured on sine wave. We cannot measure the phase noise with square wave peak, it is just too broad. 

 

We suggested to get a better performance of sine wave OCXO clock, it is because if lousy sine to square wave convertor, it makes worse sound. 

 

It is much easier to convert sine wave to square wave by simple amplify the sine wave and limit the peaks to get the square wave. 

 

The EtherREGEN and Mutec MC3-USB can accept the sine wave because of build in convertor, so they can accept both square and sine wave without compatible or performance issue.  

[soares]

11 hours ago, Superdad said:

Yes, as previously explained, the EtherREGEN (‘B’ side in particular) will not tolerate even the slightest hiccup in the clock. It will not recover (without reboot) after a glitch even as short as 3uSec. So clock and cable must be perfect.

Thank you  so much John. Just to say that after 24 hours, everything seems to be ok and I am quite happy with the results. 

[renan]

On 1/15/2021 at 4:22 PM, Ponkbutler said:

I look forward to hearing how you found the different clocks changed things, Renan.

Hi Ponkbutler,

 

Got time to make some testing on DCS Network bridge, Etheregen and Mutec Ref10 but unfortunately not all together since the post office was retaining my Mutec while I had the ehterregen home for an aftenoon.

 

So I have this clock

https://fr.aliexpress.com/item/4001307541392.html?src=google&albch=shopping&acnt=248-630-5778&isdl=y&slnk=&plac=&mtctp=&albbt=Google_7_shopping&aff_platform=google&aff_short_key=UneMJZVf&&albagn=888888&isSmbAutoCall=false&needSmbHouyi=false&albcp=10191220526&albag=107473525328&trgt=709881906465&crea=fr4001307541392&netw=u&device=c&albpg=709881906465&albpd=fr4001307541392&gclid=CjwKCAiAxp-ABhALEiwAXm6IyX36yrH6JfiPDW4KCWrkR07Xp9S6x_cllUR7XFUAgOK6ULVE3QMxmBoChbgQAvD_BwE&gclsrc=aw.ds

 

which is quite different from the one from Martin T since it is suppose to replace a Mutec Mc3 Smart clock allowing output at 44 and 48Hzs multiple.

 

Now I have an aqvox se and a friend bring me an etheregen for an afternoon. 

Results of the listening. Yes the Fos-3 clock (mine is not DAPU bue FE-180) improve the listening on etheregen. Surprisingly clocking both Etherregen and  DCS Network bridge was not very good. The clock on only one apparatus is better. May be a pb of impedance adaptation  between them??

 

3 days after while I have no more the etheregen I receive my Mutec Ref10. So the only comparison i've done was Fe-180 clock providing 10MHz and 44KHzs vs Mutec ref10 providing 10MHzs in the aliexpress clock on my DCS Network bridge. There is a huge difference and sound improvement by adding the Ref10. No a 20 times improvement like the difference in price but let say a "3"times improvement if i have to quantify. Much more impact for low frequency that I did not feel with the FE-180 and more air than with FE-180.

Well to say it the FE-180 is a very good improvmeent but the Ref10 is even better. This will not surprise a lot of people.

 

Just one question left does my aliexpress clock is a real "same quality "replacement for the Mutec MC3 smart clock or not??

[AfterDark.]

39 minutes ago, renan said:

Hi Ponkbutler,

 

Got time to make some testing on DCS Network bridge, Etheregen and Mutec Ref10 but unfortunately not all together since the post office was retaining my Mutec while I had the ehterregen home for an aftenoon.

 

So I have this clock

https://fr.aliexpress.com/item/4001307541392.html?src=google&albch=shopping&acnt=248-630-5778&isdl=y&slnk=&plac=&mtctp=&albbt=Google_7_shopping&aff_platform=google&aff_short_key=UneMJZVf&&albagn=888888&isSmbAutoCall=false&needSmbHouyi=false&albcp=10191220526&albag=107473525328&trgt=709881906465&crea=fr4001307541392&netw=u&device=c&albpg=709881906465&albpd=fr4001307541392&gclid=CjwKCAiAxp-ABhALEiwAXm6IyX36yrH6JfiPDW4KCWrkR07Xp9S6x_cllUR7XFUAgOK6ULVE3QMxmBoChbgQAvD_BwE&gclsrc=aw.ds

 

which is quite different from the one from Martin T since it is suppose to replace a Mutec Mc3 Smart clock allowing output at 44 and 48Hzs multiple.

 

Now I have an aqvox se and a friend bring me an etheregen for an afternoon. 

Results of the listening. Yes the Fos-3 clock (mine is not DAPU bue FE-180) improve the listening on etheregen. Surprisingly clocking both Etherregen and  DCS Network bridge was not very good. The clock on only one apparatus is better. May be a pb of impedance adaptation  between them??

 

3 days after while I have no more the etheregen I receive my Mutec Ref10. So the only comparison i've done was Fe-180 clock providing 10MHz and 44KHzs vs Mutec ref10 providing 10MHzs in the aliexpress clock on my DCS Network bridge. There is a huge difference and sound improvement by adding the Ref10. No a 20 times improvement like the difference in price but let say a "3"times improvement if i have to quantify. Much more impact for low frequency that I did not feel with the FE-180 and more air than with FE-180.

Well to say it the FE-180 is a very good improvmeent but the Ref10 is even better. This will not surprise a lot of people.

 

Just one question left does my aliexpress clock is a real "same quality "replacement for the Mutec MC3 smart clock or not??

 

With the aliexpress clock, it actually sounded worse in Mutec MC3-USB than the plain (without the aliexpress clock). The sound is congested towards the middle with narrower width. Mid image came forward. Details become messy (using selectable square wave). If choose sine wave, it is smoother but lacked dynamic range. 

 

There is no comparison vs Mutec Ref10. 

[R1200CL]

@AfterDark.

Is it possible to have internal pictures of your clock, and some explanations of the variants?

Is it an option for you to offer a pure DC version ?

 

What is your point of view in suppling dual clock outputs ?

[ambre]

On 1/18/2021 at 3:34 PM, ambre said:

 

1.     Master clock for EtherRegen:

a)     Master clock for EtherRegen together with appropriate linear supply. 

b)     This version has following specs; Phase noise -130dBC/Hz @10Hz. In a Square wave or Sin-wave edition. 

c)     Basic Audiophilestyle version with Square- wave will cost approx. HK$ 3600,00 or the Sin-wave (version Queen) also

         HK$ 3600.

d)     Including 1.0 mtr GIESEMANN CLOCK CABLE + 10% OFF DISCOUNT CODE: AUDIOPHILESTYLE.COM + 75OHM

        ADAPTOR.

e)     The Giesemann Linear Power Supply at 12V 35W 2.2A, Dual 12 outputs. With CARBON FIBER + SCHURTER GOLD

        FUSE & FURUTECH NCF will cost HK$ 1200,00

f)      Resume; two options for Master Clock for EtherREGEN (Sine and Square wave)

 

Hi Guys,

Just ordered the Giesemann Master Clock with Emperor Crown OXCO with seems  superior to entry level of Cybershaft.

Including an appropriate Supply LPS ( Carbon Fiber, Gold fuse, Furatech NCF and 0,8 mtr DC cable)  with all further add-ons Clock Cable (Huber + Suhner & BNC plugs)  and Cybershaft 75-> 50 Ohm adaptor.   AND A NICE GROUPS BUY DISCOUNT FOR " AUDIOPHILESTYLE"  USERS.

 

I give the Giesemann a fair chance. Ps. I also ordered the BG7TBL 10Mhz Ocxo 3 Channel Square Wave for around € 90.00. And I can compare everything with a Mutec Ref 10 and maybe a SOTM too.

 

For your information. A well-known person whose name cannot be further mentioned uses the entrance Cybershaft (Sin wave) to his satisfaction. The Emperor Crown from Giesemann for the approx. same price is much better in terms of specifications according to Afterdark. For the all-in price of approximately € 840,00 this seems a good purchase. Especially compared to the established brands in this segment.

 

 

[Ponkbutler]

2 hours ago, ambre said:

Hi Guys,

Just ordered the Giesemann Master Clock with Emperor Crown OXCO with seems  superior to entry level of Cybershaft.

Including an appropriate Supply LPS ( Carbon Fiber, Gold fuse, Furatech NCF and 0,8 mtr DC cable)  with all further add-ons Clock Cable (Huber + Suhner & BNC plugs)  and Cybershaft 75-> 50 Ohm adaptor.   AND A NICE GROUPS BUY DISCOUNT FOR " AUDIOPHILESTYLE"  USERS.

 

I give the Giesemann a fair chance. Ps. I also ordered the BG7TBL 10Mhz Ocxo 3 Channel Square Wave for around € 90.00. And I can compare everything with a Mutec Ref 10 and maybe a SOTM too.

 

That sounds very exciting and your impressions will I am sure be helpful to many of us.

[renan]

3 hours ago, AfterDark. said:

 

With the aliexpress clock, it actually sounded worse in Mutec MC3-USB than the plain (without the aliexpress clock). The sound is congested towards the middle with narrower width. Mid image came forward. Details become messy (using selectable square wave). If choose sine wave, it is smoother but lacked dynamic range. 

 

There is no comparison vs Mutec Ref10. 

So you think the clock in the MC3-USB is better then the DAPU ocxo clock or its implementation. 

[AfterDark.]

53 minutes ago, renan said:

So you think the clock in the MC3-USB is better then the DAPU ocxo clock or its implementation. 

 

Mutec MC3-USB is very sensitive to quality OCXO clock, on the other way, it require better performance/quality clock to perform well in related to the re-clock functions for 10M Optimization. When tested with classical music, it will show the differences and improvement when you use the high quality OCXO. So, it is worth to have better OCXO for Mutec MC3-USB, to match the price range to work with. 

 

We have perform varies test against most of the OCXO clock in the market, so we suggested to get the Phase Noise graph @1Hz & @10Hz (reference points) and Allan Derivation measurement when comparing high quality OCXO. It can shows the performance in written form.