Wyred4Sound - Do clocks age?

Hi Erik,

Assuming this is the review you are referring to, the relevant words appear to be on page 3, and are as follows:

Tony's opening comments on owner feedback which Wyred has collected suggest that potential applications include not just 'consumer-grade' mainstream gear but more 'high-end' stuff that's simply a few years old. Sufficient age should imply earlier less accurate types of clocks. I didn't think my Metrum Hex, AURALiC Vega or Aqua Hifi La Voce 2 converters were old enough to apply. Yet. But I'd check just in case.

I would interpret that to not be saying that clock performance deteriorates significantly over time, but rather that recent designs tend to be better performers than older designs.  Meaning that recent designs tend to be better performers than the older designs were even when the older designs were new.

Regards,
-- Al
 

As I might have expected, Geoff, it looks like you're well versed on Einstein's Theory of Special Relativity :-)

Regards,
-- Al
 

I’d be surprised if clocks got all that much better over 2 years though.

I have no particular knowledge of that, Erik, but I wouldn’t be at all surprised if you were right. I was just offering my interpretation of the reviewer’s comment.

Also, when it comes to jitter specs I wonder how much consistency and meaningfulness there is among the different manufacturers with respect to how they are defined. The jitter measurements John Atkinson provides in Stereophile are often specifically stated to be averages taken across a specific bandwidth, and detailed spectral plots based on precisely specified test conditions are often provided as well. But I’d expect that when specs are provided by manufacturers there is lots of opportunity for creative "specmanship." Examples being the bandwidth of the measurement, peak-to-peak vs. peak vs. rms vs. other forms of averaging, measurement at the output of the oscillator vs. the output of the component vs. other circuit points within the component, or just the spec on the oscillator itself as determined under ideal conditions independent of the surrounding circuit and the noise it introduces, etc. etc.

Regards,
-- Al

Do the clocks degrade with time? I have wondered about this as well ...

I have no particular knowledge regarding that question, but it wouldn't surprise me if such concerns arose as a result of flawed extrapolations from the fact that over time digital watches and clocks gain or lose time.  "Flawed" for two reasons:

1)That would occur to some degree even if nothing at all in the watch or clock changes over time, just as a consequence of its initial inaccuracy.

2)If the master oscillator in such a device were to gain or lose the equivalent of say a minute per year, due to some combination of initial inaccuracy and changes in accuracy that may occur over time, that would amount over a period of 10 years to an error of less than 0.002%.  In a digital audio device that would result in a corresponding change in pitch, which would seem unlikely to be perceivable even to those having perfect pitch.  0.002% would mean, for example, that a tone of 1000 Hz would be reproduced as 1000.02 Hz.

What might matter is significant degradation over time in the jitter performance of the device, but FWIW I am not aware of any credible evidence that would raise concern about that possibility.

Regards,
-- Al