You could do the same with your Hi End Audio science fiction lyrics.
What is asserted without proof can be dismissed without proof.
And I would not dare to show mostly unshielded vector board
in that claimed astral sound stage environment.
That may generate traffic on your new blog, but no fame.
The title of this thread contains
"Building a low phase noise/jitter crystal oscillator",
after all.
The only thing that is NOT discussed here is low phase noise/jitter
crystal oscillators. This thread has turned into a solder kit evaluation
mudslinging where the real work done has no more connection to.
What is asserted without proof can be dismissed without proof.
And I would not dare to show mostly unshielded vector board
in that claimed astral sound stage environment.
That may generate traffic on your new blog, but no fame.
The title of this thread contains
"Building a low phase noise/jitter crystal oscillator",
after all.
The only thing that is NOT discussed here is low phase noise/jitter
crystal oscillators. This thread has turned into a solder kit evaluation
mudslinging where the real work done has no more connection to.
Last edited:
Hi Chris,
sorry but I have to disagree.
Audio aside, I don't think the test equipment in your lab uses better oscillators than mostly discussed.
Here we are talking about the best oscillators ever built, the DRIXO oscillator is in class with the Wenzel BT ULN and the Oscilloquartz BVA 8607.
I don't know any instrument which implements such SOTA oscillators.
About perceived sound and measurements, as you know, there was a dedicated thread, but it has been closed after 2000 pages and 20000 posts.
You know my thought, measurements are useful during the development to understand when there is something clearly wrong.
But they are only a part of the process.
In the end we are using audio devices to listen to music and we like to do this in a pleasant way.
Nobody wants to be tortured by a system that tires you out after half an hour of listening to music.
It matters little if the audio devices used have stellar measurements or not, the only thing that really matters is the pleasure of listening to music as close as possible to the real event.
My thought comes from years of experience (I designed my first integrated amplifier in 1990, several thousands unit were sold all around the world) and listening tests.
Just to mention one of these listening experiences
Sound Quality Vs. Measurements
Every time you improve the quality of the clock in the digital to analog conversion the sound quality improves.
I have had exactly the same experience reported by many members on this thread.
Going from a 30 cent oscillator to a Crystek CCHD-957 and finally to the old Driscoll is as if every time you remove an unwanted veil from the music.
Finally, I would close this post quoting planet10 who quoted Floyd Toole:
"Two ears and a brain are massively more analytical and adaptable than an omnidirectional microphone and an analyzer."
Andrea
sorry but I have to disagree.
Audio aside, I don't think the test equipment in your lab uses better oscillators than mostly discussed.
Here we are talking about the best oscillators ever built, the DRIXO oscillator is in class with the Wenzel BT ULN and the Oscilloquartz BVA 8607.
I don't know any instrument which implements such SOTA oscillators.
About perceived sound and measurements, as you know, there was a dedicated thread, but it has been closed after 2000 pages and 20000 posts.
You know my thought, measurements are useful during the development to understand when there is something clearly wrong.
But they are only a part of the process.
In the end we are using audio devices to listen to music and we like to do this in a pleasant way.
Nobody wants to be tortured by a system that tires you out after half an hour of listening to music.
It matters little if the audio devices used have stellar measurements or not, the only thing that really matters is the pleasure of listening to music as close as possible to the real event.
My thought comes from years of experience (I designed my first integrated amplifier in 1990, several thousands unit were sold all around the world) and listening tests.
Just to mention one of these listening experiences
Sound Quality Vs. Measurements
Every time you improve the quality of the clock in the digital to analog conversion the sound quality improves.
I have had exactly the same experience reported by many members on this thread.
Going from a 30 cent oscillator to a Crystek CCHD-957 and finally to the old Driscoll is as if every time you remove an unwanted veil from the music.
Finally, I would close this post quoting planet10 who quoted Floyd Toole:
"Two ears and a brain are massively more analytical and adaptable than an omnidirectional microphone and an analyzer."
Andrea
So let me distil the above...
I test and I listen, much experience
Observation is very important so I am careful what I say
My past observations and experience puts me on very solid ground so in the case of Andrea's clocks I do not need to test or listen since it cannot be
Provide me proof and then I will consider these clocks
Back to listening to some sweet music
This would actually be a fitting end to this thread. You had a concept and you have realized it. Now it is shared by many who will use it for years and value the contribution it makes to their digital listening pleasure.
Sadly the thread has been invaded by the measurement obsessed elite.
It always takes the same path. They declare their superiority and self declared right to abuse those they consider their inferior.
Usually we hear "behold my engineering credential, I am superior"
or God forbid "behold my doctorate" or "behold my employment as a designer/academic/tv repairman etc" "I am superior"
Now I shall bash all the inferior ones that do not belong on hallowed diyaudio soil.
It is no different to "behold my pinky white pigmentation / or my Y chromosome / or my political affiliation". End of rational discussion.
Well, that isn't what I am saying at all.
Listening is important, and I also listen to what I am doing - carefully. What I did say is that after a certain point, further improvement is indistinguishable. This confirmed by listening and measurements. Now, I also said that you are welcome to improve things beyond that point - and I attempt to do that with audio circuits every day. Nothing wrong with that at all. But, understand that humans are not very good instruments and are easily lead astray by their wonderful brains. Careful testing with others and changing sources will clearly show this. That is a time consuming process, so I use testing equipment to shorten that cycle and keep me honest. This is also an interest of mine, so I have a lot of experience here. Andrea does as well, and that is fine. I do hold up my hand when he says that improvements past that point where you can't really hear them are audible. I am sorry Andrea, but this is not true.
Now consider this. I am also familiar with the studio recording equipment. They do not use oscillators in ovens. Hmm, what does that mean to you out there? That means the sampling occurs with the same jitter as a normal crystal, or sometimes they use temperature compensated clock modules. A normal ovenized crystal has far lower jitter and greater frequency accuracy. So with that, you have already exceeded the specifications of the encoding process. Andrea, would you not agree with this?
A double oven SC cut crystal (what I use) is the lowest jitter producing oscillator that exists. You can improve the drive circuit and power supply (carefully done in test equipment). I am locked to the GPS system, but that is only for long term measurements. When testing short term effects, I have to go into "hold-over" mode so that there are no step corrections during the measurement. Andrea, can you agree with me on that?
I am not saying I am superior in any way. All I was trying to get across is that yes, I have great experience in this. I do have the test equipment to keep me on track, which most of you do not have access to. Then finally, trying to interject some reason into the conversation. There is a point where further improvement in your clock for decoding does not improve the sound quality. In fact, the jitter is so small your hearing and brain cannot detect it (do the math).
I do recognize that Andrea has improved clocks and I applaud him for that achievement. But, do not oversell that improvement. Be honest, find that point where it doesn't matter (it does exist). People like to have the best, so you will still sell clocks - don't worry. I have zero problem with that. After all, that is what my bench is all about. Doing the best I can. But I do have an issue with claims that improvement of clock jitter and short term stability are clearly audible above the jitter of the encoded material. That is not a logical claim to make. Once you pass the jitter and stability of the encoded material, the dominant error is in the program material itself. A more perfect clock at that point has no effect on the existing errors. So once the dominant error becomes the material, further improvement does nothing but make you feel better and that's about it. Just, let's be honest about it.
-Chris
Listening is important, and I also listen to what I am doing - carefully. What I did say is that after a certain point, further improvement is indistinguishable. This confirmed by listening and measurements. Now, I also said that you are welcome to improve things beyond that point - and I attempt to do that with audio circuits every day. Nothing wrong with that at all. But, understand that humans are not very good instruments and are easily lead astray by their wonderful brains. Careful testing with others and changing sources will clearly show this. That is a time consuming process, so I use testing equipment to shorten that cycle and keep me honest. This is also an interest of mine, so I have a lot of experience here. Andrea does as well, and that is fine. I do hold up my hand when he says that improvements past that point where you can't really hear them are audible. I am sorry Andrea, but this is not true.
Now consider this. I am also familiar with the studio recording equipment. They do not use oscillators in ovens. Hmm, what does that mean to you out there? That means the sampling occurs with the same jitter as a normal crystal, or sometimes they use temperature compensated clock modules. A normal ovenized crystal has far lower jitter and greater frequency accuracy. So with that, you have already exceeded the specifications of the encoding process. Andrea, would you not agree with this?
A double oven SC cut crystal (what I use) is the lowest jitter producing oscillator that exists. You can improve the drive circuit and power supply (carefully done in test equipment). I am locked to the GPS system, but that is only for long term measurements. When testing short term effects, I have to go into "hold-over" mode so that there are no step corrections during the measurement. Andrea, can you agree with me on that?
I am not saying I am superior in any way. All I was trying to get across is that yes, I have great experience in this. I do have the test equipment to keep me on track, which most of you do not have access to. Then finally, trying to interject some reason into the conversation. There is a point where further improvement in your clock for decoding does not improve the sound quality. In fact, the jitter is so small your hearing and brain cannot detect it (do the math).
I do recognize that Andrea has improved clocks and I applaud him for that achievement. But, do not oversell that improvement. Be honest, find that point where it doesn't matter (it does exist). People like to have the best, so you will still sell clocks - don't worry. I have zero problem with that. After all, that is what my bench is all about. Doing the best I can. But I do have an issue with claims that improvement of clock jitter and short term stability are clearly audible above the jitter of the encoded material. That is not a logical claim to make. Once you pass the jitter and stability of the encoded material, the dominant error is in the program material itself. A more perfect clock at that point has no effect on the existing errors. So once the dominant error becomes the material, further improvement does nothing but make you feel better and that's about it. Just, let's be honest about it.
-Chris
While logic is present I'd like to pass the following... maybe better clocks in the recording chain would improve the whole party even more?? I mean, if you have a bland drink with some taste for strawberry, adding more strawberries would increase the strawberry sent, no? It's not necessarily a on/off effect? Then I would like to state that I am indeed sceptical to the real hearing impact and cant help to think that it is placebo / await-effect acting...
...but as usual.. you cant disagree with someones experience....
//
...but as usual.. you cant disagree with someones experience....
//
Hi Chris,
audio aside, your double oven SC cut crystal oscillator cannot be the lowest jitter producing oscillator that exists unless it was the Wenzel BT ULN or the Oscilloquartz BVA 8607 (or one of its successors).
If you are using a MTI Milliren or a Morion oscillator they are very good oscillator but far from the above listed (and also far from the DRIXO oscillator).
We are using a pair of MTI 260 as the reference for the phase noise measurements with the Timepod, from the plots I have published the DRIXO oscillator outperforms them by at least 10-15 dB.
About the measurements you know that at least in part we disagree.
You know we have a suitable lab, I have listed our gears.
I own a QA401 which is almost the same of your audio analyzer, I usually make measurements with such gear but the results are not my Bible.
At the end I use my ears to evaluate the results and if I get tired after a few minutes listening to music it does mean there is something wrong, regardless of the measurements.
As I said several times listening to music has to be a pleasure for me and not a torture.
Andrea
audio aside, your double oven SC cut crystal oscillator cannot be the lowest jitter producing oscillator that exists unless it was the Wenzel BT ULN or the Oscilloquartz BVA 8607 (or one of its successors).
If you are using a MTI Milliren or a Morion oscillator they are very good oscillator but far from the above listed (and also far from the DRIXO oscillator).
We are using a pair of MTI 260 as the reference for the phase noise measurements with the Timepod, from the plots I have published the DRIXO oscillator outperforms them by at least 10-15 dB.
About the measurements you know that at least in part we disagree.
You know we have a suitable lab, I have listed our gears.
I own a QA401 which is almost the same of your audio analyzer, I usually make measurements with such gear but the results are not my Bible.
At the end I use my ears to evaluate the results and if I get tired after a few minutes listening to music it does mean there is something wrong, regardless of the measurements.
As I said several times listening to music has to be a pleasure for me and not a torture.
Andrea
But when everyone who gloms this clock into their DAC, with varying skills ALL report the same things in the same audiophile mag type gushing words you have to wonder if you aren't believing it will make this difference so you hear it?
No not really, no. But I didn't expect you to get that.
But do carry on by all means...
//
Understood, and duly noted. And 25 years later, you decided to turn into a DIY aficionado and volunteer your designs for the benefit of the DIYAudio community. Nice.
I don't think there's anybody after fame here. After revenue, some DIYers and their hobby partners, no doubt about.
Chris, it sounds like you are claiming that your personal hearing sets the standard for what is humanly audible for any person on earth. What if some people notice small differences that your brain ignores? Why isn't that another possibility besides your opinion expressed above?
Last edited:
Hi Mark,
Sorry, that is not my intention at all. However, it is the listening tests done by many I am referring to. I have slightly better than average ability, but by no means the best.
I have learned that human perception has limits short of what we can measure with the proper equipment and setup. So, while I listen to what I am doing, I also measure it to see if they are in agreement. As I have said so many times, this keeps me honest and on the right path. It is so easy for your mind to play tricks on you. I also don't assess my own work, I let others do that.
-Chris
Sorry, that is not my intention at all. However, it is the listening tests done by many I am referring to. I have slightly better than average ability, but by no means the best.
I have learned that human perception has limits short of what we can measure with the proper equipment and setup. So, while I listen to what I am doing, I also measure it to see if they are in agreement. As I have said so many times, this keeps me honest and on the right path. It is so easy for your mind to play tricks on you. I also don't assess my own work, I let others do that.
-Chris
As usual you have understood less than zero.
30 years ago I already worked as IT consultant and not as audio professionals.
A friend, who was the italian importer of Naim Audio, did ask me and my IT collegue Carlo (RIP) if we would design an integrated ampliefier for his company.
We did the design as a personal favor since the person who asked us was a school friend of Carlo.
We did the job in our free time, it took more than one year.
I got a Naim CD3 as the payment for the design.
The amplifier (Audiogram MB1) was then sold over the world and got fine review from the most important audio magazine (Italian and brithish).
STOP.
We were not audio professionals.
Just like me and Roberto we are not audio professionals, and yes we like to design audio devices and also we like to share them with the audio community.
But we're not Santa Claus either, therefore we cannot afford to give away expensive items.
Bill, For my AK4499 dac with well optimized Crystek clocks and with an external clock input, not much improvement was heard by me and an audio designer friend when we tried one of Andrea's clocks at 22MHz using quick A/B switching between clocks. But there was a little bit of improvement with Andrea's clock. Using one of Andrea's 11MHz clocks and his doubler, there was some further small-ish improvement. The "squarer" was powered by my clock PCB.
Other dac designs may give other results of course.
The above said, and so far as I can tell, very few people in the forum have figured out how to get really good performance out of Crystek clocks. Diyiggy gives a good hint about which bypass caps help. Other details matter too: proper power supply design, clock output loading and buffering, keeping clocks running at all times, PCB layout, etc. Because of that I don't find it hard to believe that most people will hear a significant improvement from Andrea's clocks if not simply because they are much more well-implemented clocks than most Crystek-based designs are.
Other dac designs may give other results of course.
The above said, and so far as I can tell, very few people in the forum have figured out how to get really good performance out of Crystek clocks. Diyiggy gives a good hint about which bypass caps help. Other details matter too: proper power supply design, clock output loading and buffering, keeping clocks running at all times, PCB layout, etc. Because of that I don't find it hard to believe that most people will hear a significant improvement from Andrea's clocks if not simply because they are much more well-implemented clocks than most Crystek-based designs are.
Last edited:
Hi Mark,
Mostly with the HP 5372A and looking at it with a spectrum analyzer with the window very small and looking at the skirt near the bottom. With the 5372A the observation time is very long and I use a box over the oscillator to block air currents.
So, how do you do this if I may ask?
Mostly with the HP 5372A and looking at it with a spectrum analyzer with the window very small and looking at the skirt near the bottom. With the 5372A the observation time is very long and I use a box over the oscillator to block air currents.
So, how do you do this if I may ask?
All I have for that is listening tests. Having measurement instrumentation for it would be nice, but I'm a hobbyist when it comes to audio design and the proper gear is expensive. My friend is a professional audio designer but he doesn't do dac design, so he doesn't have dac-specific test gear I could borrow.
Last edited:
- Status
- Not open for further replies.