[time-nuts] ADEV vs MDEV
Bruce Griffiths
bruce.griffiths at xtra.co.nz
Sat Feb 6 23:59:54 UTC 2010
WarrenS wrote:
>
> And still another way of looking at the problem is:
> One wants to know the Average freq over each of the 1 sec Tau sample
> intervals,
> and do it in such a way that there is no dead spots and no overlapping
> information. (within the stated bandwidth)
> Can be done with Phase (which tends to have limited resolution at low
> tau)
> BUT Easy enough to do with a properly set up "Tight Phase-Lock Loop".
Stein doesnt believe so, particularly for long Tau (greater than a few
times the PLL inverse bandwidth).
The relevant paper is on the Symmetricom site.
> Now send that raw data to "PLOTTER" and tell it that it is Freq data
> and let it do its thing.
> NO phase need be involved at the raw data, and I don't think 'Plotter'
> is first turning Freq Data into phase to get ADEV Freq data.
> but that is a Ulrich question.
>
>> "However its easier just to use the sampled phases in the alternative
>> formula."
> Are we off topic again?
> For me, It is just as easy to have "Plotter" use either formula, It
> does not complain about which is harder.
>
> BTW, using a ADEV program I wrote in excel. When using Freq data, it
> was a little easer than using Phase data.
> Both give the same answer, If there is enough accurate Phase
> resolution, i.e 0.1 pS. for low taus faster than 1 sec.
>
>> "One cannot use EFC samples spaced at intervals of Tau directly ..."
> Right, I hope all know that. Needs to be integrated (averaged) Freq
> over the sampled Tau period,
> NOT instantaneous freq at some random points along the way.
> The NIST VtoF converter did that integration and so will a simple RC
> filter with oversampling and a PC.
>
> Still don't see where it needs to reconstruct the Phase evolution.
What do you think that your DVM is attempting to do?
There is (hopefully) no significant difference between the DVM output
and a sequence of phase samples.
The only concern is (with either method of reconstructing the phase
evolution) the effect of integrating small errors.
The tight PLL requires that an integrator (that introduces noise) of
some form is needed to produce the average frequency samples (or
equivalent phase samples).
Average frequency is proportional to the phase difference between start
and end of the averaging time.
So the phase differences and the average frequency samples are
equivalent to within a known scale factor.
> On the contrary the phase is used to reconstruct the Average Freq.
> AND one of the BIG problems is that is very hard to do accurately if a
> Phase TI is being used at Taus below 100 ms. (10Hz)
>
> Just so things do not get too far off the original topic, here is a
> reminder:
>>>>> "I would appreciate any comments or observations on the SIMPLEST
>>>>> scheme for making stability measurements at 1e-13 in one sec."
>
>
> ws
Bruce
>
> **************
> ----- Original Message ----- From: "Bruce Griffiths"
> <bruce.griffiths at xtra.co.nz>
> To: "Discussion of precise time and frequency measurement"
> <time-nuts at febo.com>
> Sent: Saturday, February 06, 2010 2:24 PM
> Subject: Re: [time-nuts] ADEV vs MDEV
>
>
>> Another way of looking at the problem is:
>> One has to reconstruct the phase evolution with time by integrating
>> the instantaneous frequency.
>> Then if the resultant phase evolution is sampled every Tau seconds
>> and the first differences taken and divided by Tau the result is a
>> sequence of average frequency samples required by the AVAR formula.
>> However its easier just to use the sampled phases in the alternative
>> formula.
>>
>> With the tight PLL method one has a sequence of frequency samples
>> averaged over an interval on the order of the inverse PLL loop
>> bandwidth.
>> One then has to use these samples to reconstruct the phase evolution
>> over time.
>>
>> One cannot use EFC samples spaced at intervals of Tau directly in the
>> ADEV formula which requires a sequence of frequency averages over an
>> interval of Tau.
>> If one ignores this requirement the resultant stability measure is
>> not ADEV.
>>
>> Bruce
>>
>> Bruce Griffiths wrote:
>>> The tight PLL method doesn't directly produce the average frequency
>>> over Tau.
>>> As explained in (see snapshot of relevant section):
>>> NIST special Publication 1065 Handbook of Frequency Stability
>>> Analysis <http://tf.nist.gov/timefreq/general/pdf/2220.pdf>
>>> the average frequency deviations for averaging time Tau are needed
>>> for the calculation.
>>> You need to sample at a sufficiently high rate to avoid aliasing and
>>> average (ie integrate) the individual EFC samples.
>>>
>>> If one uses phase measures then the fluctuations in the frequency
>>> averages can easily and directly calculated from the difference
>>> between the phase measured at time intervals separated by Tau.
>>>
>>> Bruce
>>>
>>> WarrenS wrote:
>>>> Bruce said:
>>>>
>>>>> Thus NIST and others quietly dropped this method several decades ago.
>>>> Could it be another reason?
>>>> I'll bet that was after they wanted to do better than 1e14
>>>> resolution AND had unlimited amounts of time and Money,
>>>> Something most time Nuts are not blessed with. I Never said it was
>>>> the BEST way.
>>>> JUST given the goal, which was 1e13 in one second, there is not a
>>>> simpler and cheaper way to do it.
>>>> And nothing you said counter that point.
>>>>
>>>>
>>>>> The frequency measures need to be integrated (either implicitly or
>>>>> explicitly) to produce phase measures which can then be used to
>>>>> calculate ADEV, MDEV etc.
>>>>
>>>> Well ONE of us certainly has something backward.
>>>> To calculate ADEV, MDEV etc. YOU need Freq Differences.
>>>> The first thing that happens when phase is used is that it is
>>>> turned into Freq by taking the difference between each sample.
>>>> Integrated Freq data, which is what "Tight Phase-Lock Loop Method"
>>>> gives you directly (no Phase conversion needed),
>>>> Need not FIRST turned into Phase so that it can then be turned back
>>>> into Freq.
>>>> BUT in any case there is no difference in the noise, for a given
>>>> bandwidth, If you don't run out of digits and You have enough
>>>> resolution.
>>>> The "Tight Phase-Lock Loop Method" can EASY get sub pS resolution,
>>>> which is better than most other ways.
>>>> AND don't need filters and slue rate control and multistage
>>>> limiters and on & on to do it, an RC works fine to replace all the
>>>> stuff.
>>>>
>>>> ws
>>>>
>>>> *****************
>>>>
>>>> ----- Original Message ----- From: "Bruce Griffiths"
>>>> <bruce.griffiths at xtra.co.nz>
>>>> To: "Discussion of precise time and frequency measurement"
>>>> <time-nuts at febo.com>
>>>> Sent: Saturday, February 06, 2010 12:11 PM
>>>> Subject: Re: [time-nuts] ADEV vs MDEV
>>>>
>>>>
>>>>> Sounds good but you still haven't found its Achilles heel:
>>>>>
>>>>> The frequency measures need to be integrated (either implicitly or
>>>>> explicitly) to produce phase measures which can then be used to
>>>>> calculate ADEV, MDEV etc.
>>>>> The major problem is that integration amplifies the small errors
>>>>> that are inevitably present.
>>>>> In practice (except for very noisy sources) the technique isnt
>>>>> particularly useful for Tau more than a few times the inverse PLL
>>>>> bandwidth.
>>>>>
>>>>> Thus NIST and others quietly dropped this method several decades ago.
>>>>> This is alluded to in Steins recent paper availble on the
>>>>> Symmetricom website:
>>>>>
>>>>> *The Allan Variance – Challenges and Opportunities*
>>>>>
>>>>>
>>>>> Bruce
>>>>>
>>>>> WarrenS wrote:
>>>>>> Peat said:
>>>>>>
>>>>>>> I would appreciate any comments or observations on the topic of
>>>>>>> apparatus with demonstrated stability measurements.
>>>>>>> My motivation is to discover the SIMPLEST scheme for making
>>>>>>> stability measurements at the 1E-13 in 1s performance level.
>>>>>>>
>>>>>>
>>>>>> If you accept that the measurement is going to limited by the
>>>>>> Reference Osc,
>>>>>> for Low COST and SIMPLE, with the ability to measure ADEVs at
>>>>>> that level,
>>>>>> Can't beat a simple analog version of NIST's "Tight Phase-Lock
>>>>>> Loop Method of measuring Freq stability".
>>>>>> http://tf.nist.gov/phase/Properties/one.htm#oneone Fig 1.7
>>>>>>
>>>>>>
>>>>>> By replacing the "Voltage to freq converter, Freq counter&
>>>>>> Printer with a Radio shack type PC data logging DVM,
>>>>>> It can be up and running from scratch in under an Hr, with no
>>>>>> high end test equipment needed.
>>>>>> If you want performance that exceeds the best of most DMTD at low
>>>>>> Tau it takes a little more work
>>>>>> and a higher speed oversampling ADC data logger and a good offset
>>>>>> voltage.
>>>>>>
>>>>>> I must add this is not a popular solution (Or a general Purpose
>>>>>> one) but
>>>>>> IF you know analog and have a GOOD osc with EFC to use for the
>>>>>> reference,
>>>>>> as far as I've been able to determine it is the BEST SIMPLE
>>>>>> answer that allows High performance.
>>>>>> Limited by My HP10811 Ref OSC, I'm getting better than 1e-12 in
>>>>>> 0.1 sec (at 30 Hz Bandwidth)
>>>>>>
>>>>>> Basic modified NIST Block Diag attached:
>>>>>> The NIST paper sums it up quite nicely:
>>>>>> 'It is not difficult to achieve a sensitivity of a part in e14
>>>>>> per Hz resolution
>>>>>> so one has excellent precision capabilities with this system.'
>>>>>>
>>>>>> This does not address your other question of ADEV vs MDEV,
>>>>>> What I've described is just a simple way to get the Low cost,
>>>>>> GOOD Raw data.
>>>>>> What you then do with that Data is a different subject.
>>>>>>
>>>>>> You can run the raw data thru one of the many ADEV programs out
>>>>>> there, 'Plotter' being my choice.
>>>>>>
>>>>>>
>>>>>> Have fun
>>>>>> ws
>>>>>>
>>>>>> *************
>>>>>>
>>>>>> [time-nuts] ADEV vs MDEV
>>>>>> Pete Rawson peterawson at earthlink.net
>>>>>> Sat Feb 6 03:59:18 UTC 2010
>>>>>>
>>>>>> Efforts are underway to develop a low cost DMTD apparatus with
>>>>>> demonstrated stability measurements of 1E-13 in 1s. It seems that
>>>>>> existing TI counters can reach this goal in 10s. (using MDEV
>>>>>> estimate
>>>>>> or 100+s. using ADEV estimate). The question is; does the MDEV tool
>>>>>> provide an appropriate measure of stability in this time range,
>>>>>> or is
>>>>>> the ADEV estimate a more correct answer?
>>>>>>
>>>>>> The TI performance I'm referring to is the 20-25 ps, single shot TI,
>>>>>> typical for theHP5370A/B, the SR620 or the CNT81/91. I have data
>>>>>> from my CNT81showing MDEV< 1E-13 in 10s. and I believe the
>>>>>> other counters behave similarly.
>>>>>>
>>>>>> I would appreciate any comments or observations on this topic.
>>>>>> My motivation is to discover the simplest scheme for making
>>>>>> stability measurements at this performance level; this is NOT
>>>>>> even close to the state-of-the-art, but can still be useful.
>>>>>>
>>>>>> Pete Rawson
>>>>>>
More information about the time-nuts
mailing list