Reality is Cruel : Cybershaft new Ultimate OCXO 10M Clocks Shootout OP20 vs OP17

[TLi]

I

I moved from the Cybershaft op17 to a Mutec solely because i now need 4 clock outputs.

When i got the Cybershaft few years back, I recall a conversation with Ken that you would want the phase noise of the clock to suited the inherent embedded noise in the original recording.

So what it meant to me was that it could be counter productive to have anything better than say a op14 -op15 and would be possibly a better match.

My observations are that the Op17 was smoother and had slightly more refinement which was more evident on audiophile recordings.

The mutec was no less enjoyable and gave a gutsier presentation which i actually preferred more for pop rock streaming from tidal.

All this was still without any glare and would say had more presence and brilliance to the highs and a more powerful midband.

This was using the same equipment and clock cable comparing just feeding 1 equipment.

The Mutec was very good, different presentation but no less any worst just after testing a single track and didnt look back or needed to make more comparisions.

I know it is hard to believe, I was just as sceptical as you before I heard OP20?Saying it was a life changing moment is a little over stated, but it did completely changed my perception of a clock.

Before that, same as many audiophiles, I think the external clock is just a mean to synconize all clock words in a system. There is more to that. When the clock reaches certain accuracy and stability, everything fall into place. The effect of clock cables becomes less prominent. The sound is just so correct. I use "correct" not good, because good sound differs from person to person. There is only one correct sound.

It is a pity that not many audiophiles can enjoy OP20 or better. It did change my approach towards my digital system forever.

 

[justubes]

How do you find the switch to the op20 compared to the impact of good clock cables were.

Is the improvement in line with op20 typical in the directions with improvements with good clocking with even less glare, smoother and more dimensional sound?

I havent heard the op20 level, but for myself, i do do like the sound of clocking in between aan ocxo and rubidium most- where it has more brilliance and liviness contributed to the sonic attributes of clocking. Well, some edge to the sound which i am guessing the op20 totally eliminates.

The wrong clock cables with the better clocks can also sound "wrong".

I am not certain how clocking more components will compare to a op20, as i have plans on the Esoteric D1x dual mono dacs which will add the 5th clocked component in my system.

 

[justubes]

Has any Cybershaft users have more feedback tested the few and true 50 ohm cables as the Shunyata clock 50 compared to other cables?

 

[TLi]

How do you find the switch to the op20 compared to the impact of good clock cables were.

Is the improvement in line with op20 typical in the directions with improvements with good clocking with even less glare, smoother and more dimensional sound?

I havent heard the op20 level, but for myself, i do do like the sound of clocking in between aan ocxo and rubidium most- where it has more brilliance and liviness contributed to the sonic attributes of clocking. Well, some edge to the sound which i am guessing the op20 totally eliminates.

The wrong clock cables with the better clocks can also sound "wrong".

I am not certain how clocking more components will compare to a op20, as i have plans on the Esoteric D1x dual mono dacs which will add the 5th clocked component in my system.

Clock and clock cables are doing different things. I also like photography, so I shall use it as an analogy.

Good clock put things in focus. Clock cable is like filter in lens, it changes the picture for the good and bad. If the focus is not exactly right, it never is, filter can alter the picture a little to compensate for it, a little edge enhancement can cover the blurrness in the out of focus area. But when the focus is reasonably correct, everything is clear, the correction brought by filtering is not needed.

Before OP20, I used a few different clocks, each was better than others, the effect of cable was obvious. When I have OP20 is the system, the sound is just right, the difference was so large that you don't need to go back to check. Changing the cables still change the sound but it is not better, just different.

 

[TLi]

Has any Cybershaft users have more feedback tested the few and true 50 ohm cables as the Shunyata clock 50 compared to other cables?

I also bought the 50ohm clock cable from Cybershaft, it is good but as I said above, eventually it is the clock that matters, cable is only secondary.

 

[SCAudiophile]

Did anyone tried the Mutec 10MHz clock ? (OCXO based)

My dCS dealer raves about it. The specs seem to be up with the best:

It is definately much easier to source than the elusive Cyber Shaft.

It is reported by everyone to be extremely good; I have no direct experience with it but many others do.

FYI on Cybershaft,...after quite a few months away designing and testing, Cybershafts are no longer elusive as they were at one time.

Send an email direct on hasegawa@cybershaft.jp to find out more....

 

[SCAudiophile]

I also bought the 50ohm clock cable from Cybershaft, it is good but as I said above, eventually it is the clock that matters, cable is only secondary.

I've tested that cable and it is very good particularly for the money. The Esoteric Excel 8N-A2000 50-ohm cable which is very rare on the used market is very good indeed but a lot more costly.

Hands-down though after years of using 75-ohm and 50-ohm cables in many setups and testing many others the best to my ears as to their immediate effect are the new CLOCK50 and CLOCK75 at Alpha and Sigma levels from Shunyata. I've spent months with Delta and Alpha and now own Sigma for a long time now and the SIGMA CLOCK50s are the best 50-ohms I've had in my system, again, to my ears and what they've done for my listening experience. I'll leave the rest to industry reviewers who have and will continue to say positive things I'm sure.

 

[Elberoth]

I have borrowed the Mutec Ref-10 clock. So far, I have been only been able to try it on Mutec's own 3+USB SPDIF/USB converter. I will try it next on my SOtM gear (SOtM Switch, sMS-200ultra, SOtM USB reclocker - all with 10M input) and next week on friends Vivaldi stack.

I had opened the clock today and boy - I'm impressed with what I saw. This is german engineering at its best. A real Vorsprung durch Technik:

The Mutec clock starts with an huge EMI AC inlet filter. Its insertion loss for common mode is 70dB @ 1MHz and a whopping 95dB for differential mode @ 1MHz !

That is HUGE. The typical AC filters you sometimes see in hifi components are like 30-40dB.

Then you have a big linear power supply with two separate secondaries (one for the oscillator and one for the logic), 105'C low ESR caps and two stage voltage regulation.

The first voltage regulator is a popular LT1085, which by most designers would probably be called 'good enough'.

But then you have the second stage, which is based on the latest 'wonder regulator', a true creme de la creme among them - the LT3045:

This reagulator was only introduced some 2 years ago. It's specs are trully amazing:

Ultralow RMS Noise: 0.8µVRMS (10Hz to 100kHz)
Ultralow Spot Noise: 2nV/?Hz at 10kHz
Ultrahigh PSRR: 76dB at 1MHz

And this is what supplies their OCXO module. That is probably the lowest noise PSU in any audio component on the market. So no wonder, this clock performs! And how!

The phase noise they got is -116dBc/Hz @ 1Hz !

Supposedly even Mutec's own engineers were surprised with the numbers they got. They even mention that on their website:

(...)We developed a dedicated power source of such incredibly low noise that even our engineers themselves were amazed by the noise figures and the resulting oscillator performance we ultimately achieved. (...)

And I do believe them.

At 10Hz the phase noise is -148dBc/Hz. Now compare that to the rubidium oscillators used in Antelope 10M clock, Jay's Audio 10MHz clock and Esoteric G-0Rb:

Antelope 10M (Spectratime LCR-900 rubidium oscillator): -80dBc/Hz @ 10Hz
Jay's Audio (FE-5680A rubidium oscillator): -100dBc/Hz @ 10Hz
Esoteric G-0Rb (SRS PRS10 rubidium oscillator): -130dBc/Hz @ 10Hz

The performance of these rubidium 'atomic clocks' is just laughable next to this Mutec (and Antelope clock in particular is just a disgrace to this industry and manufacturer) or higher grade CyberShafts.

Now, I think that Mutec uses just a regular off the shelf ultra low noise OCXO modules. I'm almost certain, the numbers they are getting exceed the XO manufacturer's specs, due to better PSU (manufacturers test those XOs using some lab PSU, which I'm almost certain is noiser than what we have here in this Mutec, so the readings they get, are lower).

Now imagine what the performance would be IF Mutec was given the hand picked OCXOs that Cyber Shaft uses in their OP20 clocks ...

 

[SCAudiophile]

I have borrowed the Mutec Ref-10 clock. So far, I have been only been able to try it on Mutec's own 3+USB SPDIF/USB converter. I will try it next on my SOtM gear (SOtM Switch, sMS-200ultra, SOtM USB reclocker - all with 10M input) and next week on friends Vivaldi stack.

I had opened the clock today and boy - I'm impressed with what I saw. This is german engineering at its best. A real Vorsprung durch Technik:

The Mutec clock wtarts with an huge EMI AC inlet filter. Its insertion loss for common mode is 70dB @ 1MHz and a whopping 95dB for differential mode @ 1MHz !

That is HUGE. The typical AC filters you sometimes see in hifi components are like 30-40dB.

Then you have a big linear power supply with two separate secondaries (one for the oscillator and one for the logic), 105'C low ESR caps and two stage voltage regulation.

The first voltage regulator is a popular LT1085, which by most designers would probably be called 'good enough'.

But then you have the second stage, which is based on the latest 'wonder regulator', a true creme de la creme among them - the LT3045:

This reagulator was only introduced some 2 years ago. It's specs are trully amazing:

Ultralow RMS Noise: 0.8µVRMS (10Hz to 100kHz)
Ultralow Spot Noise: 2nV/?Hz at 10kHz
Ultrahigh PSRR: 76dB at 1MHz

And this is what supplies their OCXO module. That is probably the lowest noise PSU in any audio component on the market. So no wonder, this clock performs! And how!

The phase noise they got is -116dBc/Hz @ 1Hz !

Supposedly even Mutec's own engineers were surprised with the numbers they got. They even mention that on their website:

(...)We developed a dedicated power source of such incredibly low noise that even our engineers themselves were amazed by the noise figures and the resulting oscillator performance we ultimately achieved. (...)

And I do believe them.

At 10Hz the phase noise is -148dBc/Hz. Now compare that to the rubidium oscillators used in Antelope 10M clock, Jay's Audio 10MHz clock and Esoteric G-0Rb:

Antelope 10M (Spectratime LCR-900 rubidium oscillator): -80dBc/Hz @ 10Hz
Jay's Audio (FE-5680A rubidium oscillator): -100dBc/Hz @ 10Hz
Esoteric G-0Rb (SRS PRS10 rubidium oscillator): -130dBc/Hz @ 10Hz

The performance of these rubidium 'atomic clocks' is just laughable next to this Mutec (and Antelope clock in particular is just a disgrace to this industry and manufacturer) or higher grade CyberShafts.

Now, I think that Mutec uses just a regular off the shelf ultra low noise OCXO modules. I'm almost certain, the numbers they are getting exceed the XO manufacturer's specs, due to better PSU (manufacturers test those XOs using some lab PSU, which I'm almost certain is noiser than what we have here in this Mutec, so the readings they get, are lower).

Now imagine what the performance would be IF Mutec was given the hand picked OCXOs that Cyber Shaft uses in their OP20 clocks ...

Adam,

That's a beautiful clock with excellent design and killer performance specs. Thanks for posting! Have you figured out the jitter calc on it?

 

[justubes]

It is impressive. The clocks are said to be custom item for Mutec.

I have not asked if the -116db phase noise is at TTL at all clock distribution points.

If it was, i would expect some losses at conversion and distribution circuits.

We could then be looking at even better phase noise specs directly at the output of the OCXO itself.

The biggest benefit is i felt was in the design of the power supply and found with mine, the unit was so much less critical of demanding a quality power cord compared and any other equipment i have owned.

 

[Elberoth]

I assume that the numbers both Cyber Shaft and Mutec give are at the OCXO pin outs. So the real world performance will be below that, clock distribution and all.

That said, the Antelope, Jay's Audio and Esoteric's numbers I quoted are also for the clock module only.

The oscillators Mutec uses may be custom made for them (most of those ultra low phase noise OCXOs are made to order anyway) but I don't think they grade them on the basis of actual performance, like Cyber Shaft.

Cybershaft gets a batch of say -115db OCXOs and then measures the actual performance of each unit. Some will be -115dB, but one in say 50, will be -120dB (and some may be below spec as well).

This is why their -120dB clock is 5x more expensive than the base model, even though if you open them up - they look exactly the same. The -120dB clock will have the hand picked, best performing OCXO from the batch.

 

[Elberoth]

So far, the Mutec Ref-10 sounds great with Mutecs own SPDIF/USB converter, model 3+USB. Adding clock brings the ussual tricks of increased resolution and greater smoothness.

My friend has borrowed the Ref-10 clock to try on his Vivaldi stack, where it replaced the CyberShaft's older OCXO Premium (which has a phase noise of -110dBc/Hz) and his initial reaction is very positive. The difference between the Mutec Ref-10 and CyberShaft OCXO Premium is much greater than the CyberShaft OCXO Premium and no 10MHz external clock.

He is using the Shunyata Sigma 75Ohm digital clock cable.

 

[oldmustang]

That's funny, I was just talking to a technical friend of mine and he told me about a new chip from Micro Semi that's become available for civilian use:
https://www.microsemi.com/product-d...-references/3824-chip-scale-atomic-clock-csac

Have you heard of this? Is it in use for audiophile digital equipment?

Looking at the spec sheet for the CSAC on the Micro Semi website the phase noise specs are mediocre compared to any of the Cybershaft units.

Just sayin’

Steve Z

 

[Elberoth]

I have googled that Microsemi miniature atomic clock - it costs $6,587.50 and performance wise is even behind the SRS PRS-10 rubidium oscilator (which costs $1400).

Its main selling points are low power consumpion and small size - not the price or performance.

 

[Barry2013]

Any more news on your friend's experience with the Mutec Ref10?

 

[jackson_k]

Guys got the MUTEC + REF 10 ,,, the first impression is that ,, um.. well it sounded very good, as it should be, but the difference really came in when I too the REF10off ! ,,, I connected the G2 directly to the DAC and that when the air, soundstage just kinda fell off the ground,, its pretty crazy ! , but I am also trying to figure out if the power pc on MUTEC and REF10 will make more difference Than the CLOCK BNC cable or another way around, any tips will be highly appreciated ! thank you !

 

[SCAudiophile]

I assume that the numbers both Cyber Shaft and Mutec give are at the OCXO pin outs. So the real world performance will be below that, clock distribution and all.

That said, the Antelope, Jay's Audio and Esoteric's numbers I quoted are also for the clock module only.

The oscillators Mutec uses may be custom made for them (most of those ultra low phase noise OCXOs are made to order anyway) but I don't think they grade them on the basis of actual performance, like Cyber Shaft.

Cybershaft gets a batch of say -115db OCXOs and then measures the actual performance of each unit. Some will be -115dB, but one in say 50, will be -120dB (and some may be below spec as well).

This is why their -120dB clock is 5x more expensive than the base model, even though if you open them up - they look exactly the same. The -120dB clock will have the hand picked, best performing OCXO from the batch.

Adam,

I cannot speak for all the test results on Cybershaft's website by OP level but I recently sent my prototype/custom build that I've had for quite a while now back to a month ago to have the new power supply with 102SSC lead wiring and the 50/75 ohm selectable output model (3 outputs now) and the micro-calibration dial added (in short, he converted my prototype to his latest production level) and here are the test results and 1/10/100 that I specifically asked him to measure at the BNC outputs, not just the OCXO module PIN outs;

Here's how things stack up from early prototype I paid to have built and today now that it's been taken to Kenji's latest production level.

1st results column is the prototype I purchased in 2017
2nd results column the unit recently taken from prototype to current production levels with new power supply, 102SSC wiring,
new distribution board with 50/75-ohm selectors, etc..as published on his website (Limited2 build level)




This is the same OP module I've had for a long time now; it was tested and did not need recalibration or replacement.

This underlines at least to me in great detail what you and others have been saying for a long time; it's more than just
how the OCXO module tests out on its pin-outs, it's about the total implementation, power supply, et. al.

 

[audioguy1958]

Looking at the spec sheet for the CSAC on the Micro Semi website the phase noise specs are mediocre compared to any of the Cybershaft units.

Just sayin’

Steve Z

You're playing fun with specs. You don't know what that particular clock can achieve with a SOTA power supply, etc. implementation. I have no doubt the Cybershaft unit is high quality, as well as the dCS.

When I look at a product (like a DAC or reference clock) I want to know how it works at the lowest level of detail and decide whether its worth my time to build one myself. I stopped playing fan boy a long time ago after I started looking at the guts of audio products. Many big name brands use beautiful casework to hide a shit implementation. Just sayin...

 

[audioguy1958]

I have borrowed the Mutec Ref-10 clock. So far, I have been only been able to try it on Mutec's own 3+USB SPDIF/USB converter. I will try it next on my SOtM gear (SOtM Switch, sMS-200ultra, SOtM USB reclocker - all with 10M input) and next week on friends Vivaldi stack.

I had opened the clock today and boy - I'm impressed with what I saw. This is german engineering at its best. A real Vorsprung durch Technik:

The Mutec clock starts with an huge EMI AC inlet filter. Its insertion loss for common mode is 70dB @ 1MHz and a whopping 95dB for differential mode @ 1MHz !

That is HUGE. The typical AC filters you sometimes see in hifi components are like 30-40dB.

Then you have a big linear power supply with two separate secondaries (one for the oscillator and one for the logic), 105'C low ESR caps and two stage voltage regulation.

The first voltage regulator is a popular LT1085, which by most designers would probably be called 'good enough'.

But then you have the second stage, which is based on the latest 'wonder regulator', a true creme de la creme among them - the LT3045:

This reagulator was only introduced some 2 years ago. It's specs are trully amazing:

Ultralow RMS Noise: 0.8µVRMS (10Hz to 100kHz)
Ultralow Spot Noise: 2nV/?Hz at 10kHz
Ultrahigh PSRR: 76dB at 1MHz

And this is what supplies their OCXO module. That is probably the lowest noise PSU in any audio component on the market. So no wonder, this clock performs! And how!

The phase noise they got is -116dBc/Hz @ 1Hz !

Supposedly even Mutec's own engineers were surprised with the numbers they got. They even mention that on their website:

(...)We developed a dedicated power source of such incredibly low noise that even our engineers themselves were amazed by the noise figures and the resulting oscillator performance we ultimately achieved. (...)

And I do believe them.

At 10Hz the phase noise is -148dBc/Hz. Now compare that to the rubidium oscillators used in Antelope 10M clock, Jay's Audio 10MHz clock and Esoteric G-0Rb:

Antelope 10M (Spectratime LCR-900 rubidium oscillator): -80dBc/Hz @ 10Hz
Jay's Audio (FE-5680A rubidium oscillator): -100dBc/Hz @ 10Hz
Esoteric G-0Rb (SRS PRS10 rubidium oscillator): -130dBc/Hz @ 10Hz

The performance of these rubidium 'atomic clocks' is just laughable next to this Mutec (and Antelope clock in particular is just a disgrace to this industry and manufacturer) or higher grade CyberShafts.

Now, I think that Mutec uses just a regular off the shelf ultra low noise OCXO modules. I'm almost certain, the numbers they are getting exceed the XO manufacturer's specs, due to better PSU (manufacturers test those XOs using some lab PSU, which I'm almost certain is noiser than what we have here in this Mutec, so the readings they get, are lower).

Now imagine what the performance would be IF Mutec was given the hand picked OCXOs that Cyber Shaft uses in their OP20 clocks ...

That does look impressive. Surface mount devices/short distances reduce the noise even more in high precision instruments. Parts selection is critical in a real SOTA product. Low quality parts equals bad performance.

 

[BobbinHobbin]

Any seen the new clock from Titan Audio Labs, the Horae?

http://www.titansaudiolab.com/en/products/Horae/Horae_jiexi.html

Specs seems to be very good. Maybe on par with the Cybershaft 017/020 & Mutec Ref-10?