Timing is everything

[microstrip]

(...) But even if I’m oversimplifying, I do think that my simple numbers demonstrate that the higher the sample rate, the more critical timing will become.

Curious that I would expect the opposite if trying to simplify - for the same jitter, as you are using more samples at an higher sample rate for a transition, and jitter is random and centered around a fixed time, I would expect some kind of cancellation due to later filtering. But may be I am wrong!

 

[microstrip]

Not necessarily. A different cable can be simply allowing more pickup of RF interference and other nasties, which then disturbs the DAC operation. (...)
Frank

And these interferences just introduce jitter. Remember no other second order "magic" effects are allowed here!

 

[fas42]

Curious that I would expect the opposite if trying to simplify - for the same jitter, as you are using more samples at an higher sample rate for a transition, and jitter is random and centered around a fixed time, I would expect some kind of cancellation due to later filtering. But may be I am wrong!

Again, I would echo Don, if jitter were purely random in nature then it would have minimal deleterious audible effect. But once some sort of pattern is induced into the jitter for whatever reason, say a poorly designed power supply struggling to drive the CD reader cleanly, then all bets are off ...

Frank

 

[Vincent Kars]

Curious that I would expect the opposite if trying to simplify - for the same jitter, as you are using more samples at an higher sample rate for a transition, and jitter is random and centered around a fixed time, I would expect some kind of cancellation due to later filtering. But may be I am wrong!

Don’t know as this is beyond what I as a typical armchair engineer can grasp.

I have a very simple line of reasoning.
I used the term “intrinsic” jitter.
This is simply the best possible value one can get as this is what the clock can manage.
For me it is evident that the higher the sample rate, the more the magnitude of the time step approaches the value of the intrinsic jitter.

It’s about time DonH writes a “Jitter for armchair engineers”

 

[fas42]

And these interferences just introduce jitter. Remember no other second order "magic" effects are allowed here!

Perfect overlap there! Yes, which is first, the chicken or the egg?! No "magic" has to come into it, everything ultimately is distortion, that is, a signal somewhere in the circuit is not behaving in a textbook manner. It could be called jitter, RF spurs, TIM: they are all variations on a theme, electronics misbehaving. The clear thing is not to try and address the solving a problem by correcting something which is a by the way in the whole issue, but the true, underlying causes ...

Frank

 

[fas42]

This is simply the best possible value one can get as this is what the clock can manage.
For me it is evident that the higher the sample rate, the more the magnitude of the time step approaches the value of the intrinsic jitter.

Clocks in themselves are not the problem: pay the money, you can get a clock as "perfect" as you want. No, the big headache is that as soon as you take the output of that clock somewhere in the circuit it starts getting messed up. Unless the designers take absolutely scrupulous care in actually hooking up the clock to where it's needed, all the money spent on getting a high quality clock has been completely wasted! It's not the components, it's the implementation: sloppy design and assembly will ruin the usefulness of the best components ...

Frank

 

[Phelonious Ponk]

TIM: they are all variations on a theme, electronics misbehaving.

Now there's a gross over-simplification a guy like me can really appreciate.

Tim

 

[fas42]

Now there's a gross over-simplification a guy like me can really appreciate.

Tim

What I was trying to say is that audible problems are the result of electronics in various ways not performing as the designer intended, a bit like being aware that the world is in a financial mess because greed and accumulation of wealth is generally rewarded in our culture; over-simplifying is saying that the precise misbehaviour that is called jitter is the root cause, something like saying that the last decision by the Fed is where it's all gone wrong ...

Frank

 

[microstrip]

(...) It’s about time DonH writes a “Jitter for armchair engineers”

Vincent,

Don did it sometime ago:

http://www.whatsbestforum.com/showthread.php?1322-Jitter-101