[time-nuts] ADEV vs MDEV
bruce.griffiths at xtra.co.nz
Sun Feb 7 01:40:10 UTC 2010
To optimise the performance and extend the range of Tau for which the
method produces valid results, the PLL bandwidth should be changed for
each value of Tau.
In practice you may be able to use the same PLL bandwidth for a range of
The only question is how wide a tau range is usable for any given PLL
> I've never heard of a 1e-13 at 1sec HP 10811, so it may be MORE than
> hard to find.
> (again not so hard to fine one at 1e-12 and 0.1 sec)
> Agree, a tight PLL is Not as flexible as a heterodyne or a DMTV, and
> has other limitations.
> Always those darn tradeoffs when you want simple and low cost.
> One trick I've done using the Tight PLL method, if the reference does
> NOT have a EFC or it is already used elsewhere such as a GPSDO.
> That is to put the feedback on the Device under test, assuming it has
> a unused EFC input.
> Get same simple block and results, Just need to correct for the Tuning
> gain of the tested Osc.
> ----- Original Message ----- From: "Bob Camp" <lists at cq.nu>
> To: "Discussion of precise time and frequency measurement"
> <time-nuts at febo.com>
> Sent: Saturday, February 06, 2010 4:26 PM
> Subject: Re: [time-nuts] ADEV vs MDEV
> I believe the statement:
> "Both systems are equally limited by the reference oscillator"
> was part of the same paragraph as the comment on 10811 short term
> Neither system, no matter how well set up will get below the stability
> of the reference oscillator.
> I have indeed read a lot of threads here. I've also tested a *lot* of
> oscillators. Finding a 10811 that consistently does <=1.0x10-^13 at 1
> second is *not* an easy task.
> Far more to the point - the tight loop requires a voltage controlled
> reference. Weather it's a 10811 or something else, it needs voltage
> control. The heterodyne approach does not. You do need to get luck
> with your frequencies if the heterodyne reference is not tunable.
> Something like a 10811 is indeed needed in a tight lock system.
> On Feb 6, 2010, at 6:55 PM, WarrenS wrote:
>>> "An ADEV noise floor of 1E-13 isn't likely when using an HP10811A as
>>> the VCXO for example."
>> How quickly one forgets and gets lost on these long topics.
>>>> "If you accept that the measurement is going to limited by the
>>>> Reference Osc,
>>>> for Low COST and SIMPLE, Can't beat a simple analog version of
>>>> NIST's "Tight Phase-Lock Loop" "
>> And which method are you saying is NOT limited by the Reference Osc??
>> Correct, not going to get to 1e-13 at one sec with a HP10811A,
>> nor likely with any other Ref Osc that most Freq nuts have.
>> SO Seems like that is GOOD enough noise floor limit to use for a "low
>> cost & simple" configuration.
>> A well setup "Tight Phase-Lock Loop" method will go below that..
>> and a good HP 10811A can go below 1e-12 at 0.1 sec. (at a bandwidth
>> of 30 Hz)
>> Bruce Griffiths said:
>>> The noise of the OCXO used as a VCXO will limit the noise floor.
>>> An ADEV noise floor of 1E-13 isnt likely when using an HP10811A as the
>>> VCXO for example.
>> ***** Original Topic *************
>>>> 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