[time-nuts] Method for comparing oscillators

[Magnus Danielson]

Steve Rooke wrote:
> 2009/8/7 Magnus Danielson <magnus at rubidium.dyndns.org>:
>>> For the quality of xtals and sources that some
>>> time-nuts are testing, this is unlikely to be a huge problem but for
>>> lesser sources this is a real factor hence my suggestion for the use
>>> of Hadamard Deviation.
>> Having done the exercise on a fellow time-nuts measurements I beg to differ.
>> It became clear that estimating the drift as a linear static component and
>> then calculate ADEV with raw samples and ADEV with drift compensated samples
>> it became clear that raw samples ADEV was infact drift compensated as it
>> leveled out on the drift value, which is expected.
> 
> Interesting.

I meant to say drift limited. Ah well.

>> You can't make the above assumption unless you know what the drift is and
>> know that the ADEV is above your drift level for the intended tau range. The
>> reason I keep pointing this thing out is that after having it pointed out in
>> several sources relating to how one does real measurements I have done the
>> exercise and been able to remove the limiting drift component.
> 
> That's why we are all here to learn.
> 
>> It's not advanced processing, so just do it rather than argue against it.
> 
> I'm not arguing against it, I was given the impression that it was a
> better tool for oscillators with drift but I now see that this may not
> be the case, thanks.

Good.

I just realized one thing which if correct may be an important point, 
but I need to fiddle with the math a little to make sure I am right 
about it.

Regardless, I don't think ADEV is the god-sent tool for all forms of 
frequency stability analysis, but it is a handy tool. We all need to 
learn what type of issues it addresses and which it does not addresses, 
ways to make values reliable and accurate and ways to handle deviations 
which infect the numbers. There are many real-life aspects to this which 
needs to be handles with dure care. Drift mechanisms is one of them. 
Upper and lower tau-limits is another, number of samples etc. 
Overlapping vs. Naïve Allan Deviation, maybe even more aggressive 
algorithms.

>>> For practical purposes though, the xtals we use
>>> are generally embedded as part of a GPSDO which will compensate for
>>> the drift in the oscillator but cannot practically compensate for a
>>> noisy xtal and HDEV would make comparing one source with another easy.
>> The part of the noise being at taus longer than the loop will be replaced
>> with the GPS-receivers output noise. Measuring loop-locked oscillators isn't
>> the same as stand alone oscillators, so if that is a way to remove drift, it
>> will give you false readings if you beleive you are measuring the oscillator
>> itself.
> 
> I hadn't intended that the oscillator be coupled to the GPS receiver
> for the purposes of this form of measurement for exactly the reasons
> you cite hence the need to account for drift in the oscillator. The
> intent was to characterise oscillators which would be a good candidate
> for a GPSDO design. After all the output from any GPSDO is largely
> dependant on the oscillator itself for it's noise component and all
> the GPS receiver can, and should do, is to keep it loosely steered on
> the correct frequency via a long TC.

Well, OK. Just recall that in the frequency domain, the noise 
frequencies below the loop bandwidth is dominated by the input noise, 
while the oscillator noise is tracked in and suppressed. Above the loop 
bandwidth, the oscillator noise dominates and the reference noise is 
suppressed. This is easilly verified in normal LaPlaceian loop analysis, 
but not as easy to show in sigma-tau plots, but the general tendency is 
there too.

Cheers,
Magnus