Master Clock Talk

[ChJL]

In his, Hans Beekhyzen's video: "defend yourself against marketing gibberish" he at around 5:40 in the clock oscillator section talks about, that in audio one should be concerned about short term accuracy of an oscillator rather than focusing on measures implemented which improve it's performance for long term accuracy.
How do you read producers specs to find out about that fact?

 

[Dandion]

Hello, I am considering adding a clock the my R26 dac without ddc. The Gustard c16 does not have customer feedback even at half cost of the c18? Is the c16 a bad design? For 100$ more cash, I can buy the lhy ock2 but both devices use ocxo clock technology.
Is there a big dirrence in sound performance between c16 and ock 2?

 

[Thetaburn]

Both benefit from the clock and the sound quality is superb.

How would the x26pro benefit from the master clock if it is not plugged in?

 

[MartinWT]

How would the x26pro benefit from the master clock if it is not plugged in?

The clock is fed through the I2S cable, which is why maximum performance from the U18 is only obtained via I2S into the DAC.

 

[joenetic]

The clock is fed through the I2S cable, which is why maximum performance from the U18 is only obtained via I2S into the DAC.

Should X26pro also fed the 10Mhz master clock? Then the whole system are synchronized by the master clock.

 

[JaMo]

Should X26pro also fed the 10Mhz master clock? Then the whole system are synchronized by the master clock.

When You feed the X26Pro with a 10MHz master clock, all inputs benefit (USB, OPT, AES/EBU, SP/DIF, Bluetooth and I2S). With U18 only with 10MHz, only I2s via HDMI benefit.
The U18 does a great job and for most people it may be enough. I feed both.
/Jan

 

[MartinWT]

Should X26pro also fed the 10Mhz master clock? Then the whole system are synchronized by the master clock.

When connected via I2S, the X26 Pro receives the master clock fed into the U18.

I don't use any other inputs.

 

[OCC7N]

(iDSD Pro Signature model)

Wait a minute how did I miss this. How come it is not discussed very much on other threads?
-DARS clock? = Sanford Research PERF10?

Could this work with OCK-2?

 

[gimmeheadroom]

Not sure but @iFi audio should be able to confirm. However, it appears the device can function as a standalone 10 MHz clock.

 

[OCC7N]

Yeah its an amazing thought imo

 

[gimmeheadroom]

In his, Hans Beekhyzen's video: "defend yourself against marketing gibberish" he at around 5:40 in the clock oscillator section talks about, that in audio one should be concerned about short term accuracy of an oscillator rather than focusing on measures implemented which improve it's performance for long term accuracy.
How do you read producers specs to find out about that fact?

I believe I posted this upthread or in a different thread, but:

https://www.mutec-net.com/product_ref_10.php#description

Debunking the "Atomic" Myth​

Our research has shown that the high clock stability of so-called "atomic clocks" based on a rubidium or cesium oscillator is limited to the long-term time domain. This long-term stability refers to the amount the absolute clock frequency (10 MHz, or 10 million cycles per second, in this case) drifts over time. While this long-term stability may be useful for some telecommunications applications, it is essentially irrelevant for digital audio purposes. Ultimately, it is the timing from one sample to another as a digital audio stream is transferred from one device to another that needs to be as precise as possible for best sound quality. Fluctuations in short-term stability are measured as jitter or phase-noise and are central to the audiophile performance of any digital audio device.

In contrast to these "atomic clocks" then, the REF10 is engineered around MUTEC's handcrafted, oven-controlled oscillator (OCXO) made in Germany, featuring highest clock stability in the time domain relevant for audiophile digital audio performance."

As far as specs, I guess most companies will not list it because they're repurposing GPS clocks. On the Mutec page https://www.mutec-net.com/product_ref_10.php#data

it shows:

Clock Generation​

• Type: 10.000 MHz ultra-low noise oven-controlled crystal oscillator
• FQ stability when shipped: < +/-0.01 ppm
• FQ stability vs. temperature range: < +/-0.01 ppm within -20 °C to +70 °C (-4 °F to +158 °F)
• Short term stability (Allan deviation) at Tau = 1 s: 1 x 10-12 (typically)
• Aging after 30 days operation:
  < +/-0.0002 ppm (per day),
  < +/-0.03 ppm (1st year),
  < +/-0.2 ppm (10 years)
• Warm-up time at +25 °C (+77 °F) : <5 min

 

[ChJL]

I believe I posted this upthread or in a different thread, but:

https://www.mutec-net.com/product_ref_10.php#description

Debunking the "Atomic" Myth​

Our research has shown that the high clock stability of so-called "atomic clocks" based on a rubidium or cesium oscillator is limited to the long-term time domain. This long-term stability refers to the amount the absolute clock frequency (10 MHz, or 10 million cycles per second, in this case) drifts over time. While this long-term stability may be useful for some telecommunications applications, it is essentially irrelevant for digital audio purposes. Ultimately, it is the timing from one sample to another as a digital audio stream is transferred from one device to another that needs to be as precise as possible for best sound quality. Fluctuations in short-term stability are measured as jitter or phase-noise and are central to the audiophile performance of any digital audio device.

In contrast to these "atomic clocks" then, the REF10 is engineered around MUTEC's handcrafted, oven-controlled oscillator (OCXO) made in Germany, featuring highest clock stability in the time domain relevant for audiophile digital audio performance."

As far as specs, I guess most companies will not list it because they're repurposing GPS clocks. On the Mutec page https://www.mutec-net.com/product_ref_10.php#data

it shows:

Clock Generation​

• Type: 10.000 MHz ultra-low noise oven-controlled crystal oscillator
• FQ stability when shipped: < +/-0.01 ppm
• FQ stability vs. temperature range: < +/-0.01 ppm within -20 °C to +70 °C (-4 °F to +158 °F)
• Short term stability (Allan deviation) at Tau = 1 s: 1 x 10-12 (typically)
• Aging after 30 days operation:
  < +/-0.0002 ppm (per day),
  < +/-0.03 ppm (1st year),
  < +/-0.2 ppm (10 years)
• Warm-up time at +25 °C (+77 °F) : <5 min

Great, I knew I've come across that topic before. Maybe here or probably on the mutec page...

Anyway, thank you for taking time and posting it again!

Am I right in assuming when looking at specs that the phase noise @10mhz is not so important for audio than the one @1mhz?

And FQ stability should be as low as possible because that will naturally decrease over the years?
(C 18 <+-50ppb vs REF10 <+-0.01ppm) meaning REF 10 deviates less with a factor of 5?

And is Allan deviation what Hans was referring to? Again the number being as little as possible... measured per second.

Gustard are also stating FQ accuracy (besides FQ stability), again the smaller the better right?

 

[gimmeheadroom]

Great, I knew I've come across that topic before. Maybe here or probably on the mutec page...

Anyway, thank you for taking time and posting it again!

Am I right in assuming when looking at specs that the phase noise @10mhz is not so important for audio than the one @1mhz?

And FQ stability should be as low as possible because that will naturally decrease over the years?
(C 18 <+-50ppb vs REF10 <+-0.01ppm) meaning REF 10 deviates less with a factor of 5?

And is Allan deviation what Hans was referring to? Again the number being as little as possible... measured per second.

Gustard are also stating FQ accuracy (besides FQ stability), again the smaller the better right?

I will let the more knowledgeable guys speak to your questions since what little I know is from reading their posts.

 

[MartinWT]

Am I right in assuming when looking at specs that the phase noise @10mhz is not so important for audio than the one @1mhz?

You have an extraneous 'm' in there. The argument is over whether phase noise measured at 1Hz or 10Hz is more important. I recently read arguments that 10Hz is more significant for sound quality,

 

[ChJL]

I see Hz not MHz!

Just trying to get 3 or 4 specs to look for and a simple range from low, middle, highly accurate - oscillator.

I find it also interesting that Hans Beekhyzen mentioned that really good ones cost 10s! of dollars. I don't see all that many parts in a master clock for thousands of $ retail. But I know very little about production of components...