16.9344MHz LOW JITTER CLOCK KIT...

Status
Not open for further replies.
Jump to Latest
#21
Capacitor of voltage regulator

If we change capacitor of voltage regulator, also phase noise characteristic was change.

I choice SANYO OS caps.
 

Attachments

  • 3c.jpg
    3c.jpg
    71.7 KB · Views: 1,966
#23
OSCON's are good, aren't they...

They can feed the noise from Vcc to the ground (common) plane, and they do excellent job.

Now, please consider what will happen if you lower the noise OF THE COMMON (GROUND PLANE) and you keep the OSCONS?

This is why I recommended to ground the oscillator to the SACD PCB ground plane with the least amount of noise.

Regards,
Boky
 
#27
Re: XO3 is good C/N but +/-5ppm deviation

nagaesan said:
Let's stop minding clock deviation +/- 1ppm or not.
Click to expand...


parts per million deviation is not that important; the important thing to know (or measure if you have the equipment that can do it) is jitter performance at around 10Hz.

5 PPM means that you will either LOOSE or GAIN for example 5 seconds in approximately 11 days' period (Wikipedia), and this is not important for us at all - AS LONG AS THOSE seconds "keep coming" at exactly the same intervals.



Boky
 
#29
Yes, the amount of jitter at around 10Hz will directly affect the timing of our clock signal, and our ears are very sensitive (actually, our brain!) to these variations.

The lower the power supply noise - the lower the jitter - hence - lower disturbance and more accurate timing of our clock signal.

Some prefer sine wave output, some square wave... it is academic...but I am sure many would discuss this topic to the Nth degree.

Boky
 
#30
Let's stop minding clock deviation +/- 1ppm or not.

This is pre-production model for commentator & reputator sound check.

SACD decorder module frequency is 33.868800MHz.
But this was 33.868000MHz clock.


Because, even if 800Hz (23.62ppm) slipped, people did not notice it.
 

Attachments

  • cl.jpg
    cl.jpg
    61.5 KB · Views: 1,563
#32
Re: Jitter = phase Noise

nagaesan said:
Dear oscillator researchers

clock deviation ..... +/-1ppm, = not jitter
clock purity.............C/N, phase noise,-130dBc@100Hz = Jitter

I measured several crystal oscillator's Phase Noise with hp E5501B.
Attached file is LcAudio XO-3 jitter.
It is very nice jiter than usual SPXO and TCXO oscillators.


Crystal oscillator produces undesirable phase noise near the output frequency.
�@Phase noise is measured in the frequency domain, and is expressed as a ratio of signal power to noise power measured in a 1 Hz,10Hz,100Hz,1000Hz, bandwidth at a given offset from the desired signal.
Click to expand...


MAX2620 is better.
 
#35
I have oscillators in my DACs marked:

TCXO-230
xxx MHz
SCTF 03

So they are from Shenzhen SCTF Electronics.
I cannot find a meaning for the -230 ending but the sheet I can find for DIP-14 TCXO starts at -70 dB jitter at 10 Hz. 1.5-2 ppm over time.

10Hz 100Hz 1kHz 10kHz 100kHz
-70 -95 -120 -145 -150 dBc/Hz

Could be better?

What is the meaning of the -230?
Are better noise spec oscillators available? If so: where?
Or do I need a `full` (expensive) clock?

What is the relation between the clocking of a transport and that of the DAC?
I.e.: does the DAC clock matter more or less than the transport clock?
 
#36
Gents,
We at vacuum State have been doing some extreme jitter experiments on various clock designs, and have found that the powerssupply of the clock oscillator is VASTLY more important than anyone has ever realised.

The Chinese clock that started this thread proudly says it uses LM317's as regulators - well we found they are not quiet enough by a HUGE amount in our tests, and now have filtering beyond anything anyone would consider sane or reasonable, but the sound keeps getting better and better.

We now market two clocks using this technology, and there are success stories on our website from end users. The TerraFirma Lite is not cheap, but would be of interest to serious experimentors.

Regards, Allen (Vacuum State)
 
#37
Allen Wright said:
Gents,
We at vacuum State have been doing some extreme jitter experiments on various clock designs, and have found that the powerssupply of the clock oscillator is VASTLY more important than anyone has ever realised.

The Chinese clock that started this thread proudly says it uses LM317's as regulators - well we found they are not quiet enough by a HUGE amount in our tests, and now have filtering beyond anything anyone would consider sane or reasonable, but the sound keeps getting better and better.

We now market two clocks using this technology, and there are success stories on our website from end users. The TerraFirma Lite is not cheap, but would be of interest to serious experimentors.

Regards, Allen (Vacuum State)
Click to expand...


As you are still using the Tent module it can't be better than that.
You supply looks primitive to my eyes.
🙁
 
#39
Tico wrote:

>As you are still using the Tent module it can't be better than that<

Why not? The TENT with our powersupply is HUGELY improved over the TENT with other powersupplies. And even cheaper oscillators with our powersupply sound pretty much equal to the TENT.

>You supply looks primitive to my eyes.<

Primitive? You can tell circuit topology, circuit concepts and circuit values from looking at a photo? Opinions don't count for much with me. What do you like then?

Regards, Allen (Vacuum State)
 
#40
Re: Re: MAX2620 is better?

QSerraTico_Tico said:


Page 6 of the datasheet shows SINGLE SIDEBAND PHASE NOISE 10dB lower than the LCaudio. And this is with a cermic resonator so I guess it will be much better with a good crystal.
Click to expand...

Don't get too excited, the Offset frequency on that grapph, is KHz
not Hz away from carrier. 🙂

So we are talking -105dBc/Hz at 10KHz not 10Hz.

This is to be expected it is a cheap chip solution.

cheers

T
 
Status
Not open for further replies.
Top Bottom