John Curl's Blowtorch preamplifier

[andy_c]

Diamond was the guy's name, if memory serves.

Here's a web site that sells products based on Dr. Diamond's ideas 🙂.

 

[MJL21193]

I love tests, does one have to comply to all items or will a couple of them do ?

I have all bases covered.

 

[SY]

Wow, he's a versatile quack!

 

[Joshua_G]

For example:
http://www.golledge.co.uk/pdf/products/ocxos/hcd660.pdf

Thank you.
However this one goes up to 20 MHz, while I need 80 MHz.

 

[Joshua_G]

For example, I found the power supply noise vs oscillator noise very useful.

Thank you, John.

There is no doubt in my mind about the critical importance of very low power supply noise. However, I'm looking into other realms as well.

 

[Joshua_G]

I am interested in some of the Rubidium sources since they are programmable to the clocks we need and claim low phase noise. But I need to know more about them. Since they are available on e-bay for under $200 I may just explore it after CES.

Can you please post a link?
Thank you.

 

[1audio]

his spelling isn't great but here you go: http://cgi.ebay.com/Supper-Low-JITT...ryZ22636QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

It seems they are programmable for different frequencies and this guy figured out how to do it.

 

[Joshua_G]

Thank you, Demian.

 

[x-pro]

Thank you.
However this one goes up to 20 MHz, while I need 80 MHz.

http://www.golledge.com/pdf/products/ocxos/hcd210.pdf

This one goes to a 100 MHz.

Cheers

Alex

 

[gerhard]

John:
The questions about Cesium and Rubidium are both legitimate. I looked into them as well. And Teac seems to think its important. A Hydrogen Maser would be really over the top. But not necessarily better for audio. But is would give someone serious bragging rights in the $$$=good audio world.
Understanding the ways noise affects the output of a crystal oscillator (or any of the other sources) is more fundamental. I am interested in some of the Rubidium sources since they are programmable to the clocks we need and claim low phase noise. But I need to know more about them. Since they are available on e-bay for under $200 I may just explore it after CES. Still lots cheaper and faster than getting a Wenzel or Oscilloquartz oscillator. The harder part is analyzing the jitter spectrum and still harder is making sure that you aren't listening to a problem unrelated to what you were testing.

An (active) Hydrogen maser would at least not have to rely on a crystal oscillator (at least in concept).
Cesiums & Rbs & passive masers need a high quality crystal oscillator, which is responsible for all behavior with time constants of less than a few minutes. The Cs, Rb.. corrects only the long time trend. The same holds true for GPS disciplined oscillators. The usual
1pps-based synchronization will extend the influence of the crystal oscillator to the low hour range.
Cs & Rbs will provide better holdover in case the GPS reference is lost; but in all cases the short term stability aka phase noise depends only on the crystal, and only short term stability is of interest here.
(It does not matter if a performance lasts 100 ns longer.)

ebay Rb sources may have used up their rubidium reserve; this cannot be fixed at the $100 level. Nevertheless they will contain a quality oscillator but the burden is on you to generate the exact frequencies needed without messing up the phase noise.

MTI 260 is a nice osc series, too, or the venerable HP10811.

Don't even think about trying to characterize a Wenzel, MTI or Oscilloquartz. You would need a cooled whispering gallery sapphire
or other, better references, or multiple references in the same league and cross correlation. A bottomless pit.

regards, Gerhard

 

[gerhard]

Actually, I hear nuances that some much younger men don't hear.

That condition has a name: tinnitus.


regards, Gerhard

 

[Juergen Knoop]

I wonder how many of these clock mods are only audible due to the shift in clock frequency and other unwanted effects. 😕
regards

 

[gerhard]

However Jitter is usually measured in nanoseconds or picoseconds. and then usually expressed in rms values. Last I heard 10 pS was considered the target but lower values would be better.

... and without measurement bandwidth the numbers are weapons grade nonsense.


other than that, 100% consent.

regards, Gerhard

 

[1audio]

Agreed- marketing weapons to sell to the unassuming. Getting deterministic jitter info and some type of peak info is hard. And it can all be a distraction from looking at the analog output of the card. An instability or noise in the power supply for the DAC could easily swamp all of the low jitter stuff. All of this makes an analog preamp seem easy, what you see is really what you get, not an indirect reflection of it.

 

[john curl]

What I like to do, is to learn new things. Especially about audio design. Usually, I know most of the analog, beyond the usual, BUT digital is an area that I am not as well versed in. When I can learn something new in this area, it gives me a perk. Of course, I have most of the usual textbooks on digital audio, and have designed, on paper, at least, a complete CD playback system, but that is nothing to what I could know about the subject. Anyone else out there with some REAL information, and not just the 'party line' from Sony, etc.?

 

[scott wurcer]

Anyone else out there with some REAL information, and not just the 'party line' from Sony, etc.? [/B]

This is supposed to be the blowtorch thread, I would just sit back and wait this jitter stuff seems to have become a real pee-peeing contest.

How does the master clock jitter relate to analog output on a typical 24/96 sigma-delta converter or the just plain performance limits of the DAC chips?

 

[SY]

Hmmm, didn't a co-worker of your have a nice JAES paper about that?

 

[scott wurcer]

Hmmm, didn't a co-worker of your have a nice JAES paper about that?

Hmmm you mean Bob Adams? I don't remember. In any case this atomic clock stuff has to be overkill.

 

[syn08]

In any case this atomic clock stuff has to be overkill.

No. When using an atomic clock, I can hear the quarks hissing.

 

[tvi]

If you are interested in atomic clocks you might want to contact JohnW , he seems to have some experience.

From that other forum

I use 2x FTS4050 Cesium Atomic clocks - phase locked to modified Stanford Research Systems PRS10 Rubidium.
The 2x Cesium's provide the long term accuracy to Stratum 1 Performance, while the SRS PRS10 provides better short term Phase Noise.
The PRS10 was modified for lower LF noise by improving the internal regulator voltage references. John

Regards
James